The Economic Effects of Micronutrient Deficiencies:
The case of iodine.

by
Dimitra Politi
B. A., Athens University of Economics and Business, 2002
M. A., Brown University, 2004

Submitted in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy in the
Department of Economics at Brown University

Providence, Rhode Island
May 2009

c Copyright 2009 by Dimitra Politi

This dissertation by Dimitra Politi is accepted in its present form by
the Department of Economics as satisfying the dissertation requirement
for the degree of Doctor of Philosophy.

Date
David N. Weil, Director

Recommended to the Graduate Council

Date
Kenneth Chay, Reader

Date
Yona Rubinstein, Reader

Approved by the Graduate Council

Date
Sheila Bonde
Dean of the Graduate School

iii

Vita
The author was born in Rethymno, on the island of Crete, Greece, on August 11th, 1980. She
graduated from The 2nd High School of Rethymno in 1998, and moved to Athens, Greece, where she
attended Athens University of Economics and Business (AUEB). During the 2000-2001 academic year
she was an Erasmus Exchange Student in Economics at Université de Paris I - Panthéon - Sorbonne in
Paris, France. She returned to Athens in 2001, and in 2002 she graduated with a B.A. from AUEB’s
Department of International and European Economic Studies, where she majored in International
Economics and Finance. In August 2003 she moved to Providence, RI, and began her doctoral studies
in the Department of Economics at Brown University. After completing the requirements for her
Master’s degree in Economics in 2004, she began working on the link between health and economic
growth under the guidance of Prof. David N. Weil. She also assumed teaching responsibilities as a
primary instructor and a teaching assistant for various undergraduate economics courses. In 2006
she completed The Teaching Certificate I Program: The Sheridan Teaching Seminar, awarded by
Brown University’s Sheridan Center for Teaching and Learning. This dissertation represents the
culmination of her efforts to complete the requirements for the Degree of Doctor of Philosophy in
the Department of Economics at Brown University.

iv

Acknowledgements
This long journey would not have been possible had it not been for the support of many people,
only a few of whom I can mention here.
First and foremost I would like to thank my main adviser, Professor David N. Weil. David has
been my greatest teacher and the person who introduced me to the advanced study of the link
between health and wealth. I am still amazed by David’s clear-cut way of thinking about economic
issues, and by his ability to get to the essence of problems directly, bypassing any sidetracks. Apart
from being an excellent academic adviser, though, David has also been a true friend. I will always
appreciate the fact that his office door was open whenever I needed a counsel or just a little (or lots
of) encouragement.
Many thanks also go to my committee members, Professors Kenneth Chay and Yona Rubinstein,
whose suggestions greatly improved not only this dissertation, but also my quality as a researcher.
Ken’s scientific meticulousness is unparalleled, and I am greatly indebted for all the time and effort
he spent advising me on even the finest points of my work. Yona taught me to pay attention to
the story told by what we don’t observe in the data, and that is an extremely valuable lesson for
any empirical economist. He also helped me think outside the box in my dissertation and about
economics in general, and for this I am thankful.
Furthermore, I would like to thank Anna Aizer, Andrew Foster, Pedro Dal Bo, Sophocles Mavroeidis, and Herakles Polemarchakis for valuable comments and advice over the course of my graduate
studies. I also benefited from discussions with my co-author, James Feyrer, at Dartmouth College;
besides learning from him, working with him has also been a lot of fun.
At all stages of my research, my fellow graduate students and participants in the informal lunches
at the Department of Economics at Brown University provided me with lots of feedback and suggestions in a friendly atmosphere. I thank them for their help in shaping my dissertation, but also for
making my journey much more enjoyable; they have been excellent colleagues and friends in more
ways than one.
This project would have been impossible without the help of Professor Doctor Hans Bürgi, who
provided me with hard-to-find historical data and sources, and answered my many questions on the
Swiss Iodization Campaign, a subject I knew little about when I first embarked on this project.
Many thanks also go to Mr. Dominik Ullmann at the Swiss Federal Statistical Office, who helped
me gain access to the Swiss Census Data.
v

During my research I benefited greatly from financial and technical support from the Population Studies Training Center (PSTC) at Brown University. In particular, many thanks go to Paul
Koussa, Director of Computing and Information Services at the PSTC, who was always there when
I was technologically challenged. Eric Scantlebury, the Systems Manager in the Economics Department, has also been of great help. And, of course, I need to thank the Economics Department’s
Student Affairs Manager, Angelica Vargas, whose help and support in all practical matters have
been invaluable. Angelica has been a time-saving machine and a good friend ever since the first day
I arrived in Providence, and I will miss her warm smile.
Before coming to Brown, I had the opportunity to have great undergraduate advisers, who
undeniably left their mark on my training and thinking as a social scientist. To this day, Professor
Euclid Tsakalotos taught me most of what I know about the workings of the economy at large. It
was his thought-provoking lectures which fueled my curiosity and led me to pursue a graduate degree
in Economics. Professor Panos Hatzipanayotou advised me during my preparatory year, before I
came to Brown, and his help in getting me ready for Graduate School was invaluable.
My friends here at Brown, back in Greece, and elsewhere in the world supported me throughout
my journey. They believed in me even when I could not, and made me feel right at home, even
when home was far away. Chiara Valenzano has been a sister to me. Andreas Kakridis read reports
of my progress, proofread my thesis, and has been a great friend altogether. Singing with Apoichos
has been a privilege and a great way to relax and unwind when the going got tough. Members of
the Hellenic Community at Brown have been a natural support group.
Finally, I am thankful for my family. My dear grandmother, who was with me for most of this
journey, gave me her love and affection, and she provided an attentive audience whenever I visited
with her. Thanks for being there, Granny, I miss you every day. My mother is, undoubtedly, the
best editor I could ever have. My parents, sister, and extended family offered me unconditional love
and support in every step of the way, and for this I am forever grateful. There is no way I could
have completed this journey without them.

vi

To my family.

Contents
List of Tables

x

List of Figures

xii

1 The Swiss Iodization Campaign and schooling outcomes

1

1.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

1.2

Related literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4

1.3

Background on Iodine Deficiency Disorders . . . . . . . . . . . . . . . . . . . . . . .

4

1.4

The campaign for salt iodization in Switzerland . . . . . . . . . . . . . . . . . . . . .

7

1.5

The 1970 Swiss Census and Bircher’s monograph . . . . . . . . . . . . . . . . . . . .

9

1.6

Iodized Salt and the decline in deaf-mutism . . . . . . . . . . . . . . . . . . . . . . .

15

1.7

A first look at the data:using 1922 as the cutoff date for treatment . . . . . . . . . .

17

1.8

A second look at the data:The penetration of iodized salt in each canton . . . . . . .

20

1.8.1

A closer look into cantonal iodized salt sales . . . . . . . . . . . . . . . . . . .

20

1.8.2

An iodine response function . . . . . . . . . . . . . . . . . . . . . . . . . . . .

26

1.8.3

Using iodized salt sales to identify the effect of iodization . . . . . . . . . . .

27

Fuzzy regression discontinuity and jumps in sales of iodized salt . . . . . . . . . . . .

29

1.10 Interpretation of coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35

1.9

1.11 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

36

1.A Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

38

2 The Swiss Iodization Campaign and occupational outcomes

40

2.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

40

2.2

Iodine Deficiency Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42

2.3

Swiss Iodization Campaign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

2.4

Bircher’s monograph and Iodized Salt Sales . . . . . . . . . . . . . . . . . . . . . . .

44

2.5

The 1970 Swiss Census and Dictionary of Occupational Titles . . . . . . . . . . . . .

45

2.6

Identification using geographical and temporal variation . . . . . . . . . . . . . . . .

48

2.6.1

Iodized salt sales and birth location . . . . . . . . . . . . . . . . . . . . . . .

50

2.6.2

Jumps in iodized salt sales; a Fuzzy Regression Discontinuity Approach . . .

50

viii

2.7

The effect of Iodization on Occupational Choice . . . . . . . . . . . . . . . . . . . . .

52

2.8

The effect of Iodization on Characteristics of Occupations . . . . . . . . . . . . . . .

57

2.9

Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66

3 Iodine Deficiency Eradication in the USA

68

3.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

68

3.2

Iodine Deficiency Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

70

3.3

Iodine Deficiency and Salt Iodization in the United States . . . . . . . . . . . . . . .

73

3.3.1

Evidence from mortality data . . . . . . . . . . . . . . . . . . . . . . . . . . .

78

3.4

Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81

3.4.1

Defects in Drafted Men . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

3.4.2

World War II Enlistment Data . . . . . . . . . . . . . . . . . . . . . . . . . .

86

3.4.3

Test Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87

3.4.4

The Battle for the Best Enlistees . . . . . . . . . . . . . . . . . . . . . . . . .

89

3.5

Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94

3.6

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95

3.6.1

Linear Probability Model of joining the AAF . . . . . . . . . . . . . . . . . .

96

3.6.2

Logit results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

3.6.3

Falsification exercise with other diseases . . . . . . . . . . . . . . . . . . . . . 102

3.7

Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

3.8

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

3.A Average AGCT scores for the AAF versus the AGF . . . . . . . . . . . . . . . . . . 106
3.B Interpretation of the coefficients:What was the IQ increase following iodization? . . . 106
Bibliography

109

ix

List of Tables
1.1

Summary Statistics: graduation rates . . . . . . . . . . . . . . . . . . . . . . . . . .

10

1.2

Summary Statistics (continued): graduation rates by region . . . . . . . . . . . . . .

11

1.3

Salt sales and education: Coefficient on high-goiter dummy X Born after 1922 dummy 19

1.4

Salt sales and education: Coefficient on Iodized Salt 1 year prior to birth . . . . . .

28

1.5

Coefficient on high-goiter dummy X Born after jump in iodized salt sales . . . . . .

31

1.6

Fuzzy Regression Discontinuity: Coefficient on Iodized Salt 1 year prior to birth . .

34

1A-1 Annual iodized salt sales as a percentage of total salt sales . . . . . . . . . . . . . . .

38

1A-2 Goiter rates and number of observations by canton . . . . . . . . . . . . . . . . . . .

39

2.1

Actively employed individuals in 1970 Swiss Census . . . . . . . . . . . . . . . . . .

46

2.2

Broad occupational categories in 1970 Swiss Census . . . . . . . . . . . . . . . . . .

47

2.3

Job characteristics for occupational categories in 1970 Swiss Census . . . . . . . . .

49

2.4

Salt sales and occupational choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54

2.5

Salt sales and occupational choice if born after jump . . . . . . . . . . . . . . . . . .

56

2.6

High-paying occupation: Fuzzy Regression Discontinuity . . . . . . . . . . . . . . . .

58

2.7

Salt sales and occupational characteristics, whole sample . . . . . . . . . . . . . . . .

60

2.8

Salt sales and occupational characteristics, by gender . . . . . . . . . . . . . . . . . .

61

2.9

Occupational characteristics: Born after iodization, whole sample . . . . . . . . . . .

62

2.10 Occupational characteristics: Born after iodization, by gender . . . . . . . . . . . . .

63

2.11 Occupational characteristics: Fuzzy Regression Discontinuity, whole sample . . . . .

64

2.12 Occupational characteristics: Fuzzy Regression Discontinuity, by gender . . . . . . .

65

3.1

Sample size and probability of joining the AAF by birth year and enlistment semester 88

3.2

Percent of enlistees assigned to the Army Air Forces by year. . . . . . . . . . . . . .

89

3.3

Percent of high school graduates by birth year and branch. . . . . . . . . . . . . . .

92

3.4

Air Force dummy against initial goiter level interacted with birth year dummies . . .

97

3.5

Air Force dummy against high goiter dummy interacted with birth year dummies . .

98

3.6

Logit results of initial goiter level interacted with birth year dummies . . . . . . . . 100

3.7

Logit results of high-goiter group dummy interacted with birth year dummies . . . . 101

3.8

Marginal Probabilities of logit results . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

3.9

Summary of coefficients for falsification exercise with all defects . . . . . . . . . . . . 103
x

3.10 Falsification Exercise with Different Diseases . . . . . . . . . . . . . . . . . . . . . . 104

xi

List of Figures
1.1

Iodized salt penetration in Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . .

8

1.2

Bircher’s Data on goiter in recruits . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13

1.3

Histogram of goiter prevalence in Switzerland . . . . . . . . . . . . . . . . . . . . . .

15

1.4

Deaf-mutism and 1883 Goiter Prevalence before and after iodization . . . . . . . . .

16

1.5

Deaf-mutism and Iodized Salt penetration . . . . . . . . . . . . . . . . . . . . . . . .

17

1.6

Secondary education, difference between high-and low goiter districts . . . . . . . . .

21

1.7

Tertiary education, difference between high-and low goiter districts . . . . . . . . . .

22

1.8

Secondary education, males and females . . . . . . . . . . . . . . . . . . . . . . . . .

23

1.9

Tertiary education, males and females, difference between high-and low goiter districts 24

1.10 Iodized salt penetration in high- and low-goiter districts . . . . . . . . . . . . . . . .

25

1.11 Iodine response function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

1.12 Iodized salt sales by age relative to iodization . . . . . . . . . . . . . . . . . . . . . .

30

1.13 Secondary and Tertiary education before and after jump in iodized salt sales . . . .

32

2.1

Residuals graph of high-paying dummy and iodized salt . . . . . . . . . . . . . . . .

52

2.2

Regression-adjusted probability of being employed in a highly-paid occupation . . .

55

2.3

Change in occupational characteristics after iodization . . . . . . . . . . . . . . . . .

59

3.1

Distribution of Goiter during WWI . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69

3.2

Ads for Iodized Salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75

3.3

Ads for Iodized Salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

76

3.4

A Newspaper Article Discussion of Iodized Salt . . . . . . . . . . . . . . . . . . . . .

77

3.5

US Deaths from exophthalmic goiter, 1920-1960 . . . . . . . . . . . . . . . . . . . . .

79

3.6

US goiter deaths by age group, 1921-1935 . . . . . . . . . . . . . . . . . . . . . . . .

80

3.7

Changes in thyroid mortality vs. pre-existing goiter rates, 1921-1926 . . . . . . . . .

81

3.8

Deaths from exophthalmic goiter, Michigan and other states, 1920-1930 . . . . . . .

82

3.9

Sections from Defects Found in Drafted Men

. . . . . . . . . . . . . . . . . . . . . .

83

3.10 State level Simple Goiter Rates from Defects Found in Drafted Men . . . . . . . . .

84

3.11 Histogram of Section Level Simple Goiter Rates from Defects Found in Drafted Men

84

3.12 Histogram of Population Goiter Rates from Defects Found in Drafted Men . . . . . .

85

3.13 McClendon’s data on the iodine content of water . . . . . . . . . . . . . . . . . . . .

86

xii

3.14 Percentage of Air Corps Recruits with above average AGCT scores . . . . . . . . . .

90

3.15 AGCT Scores for AAF versus all enlistees . . . . . . . . . . . . . . . . . . . . . . . .

92

3.16 High-School Graduation Rates by birth year . . . . . . . . . . . . . . . . . . . . . . .

93

3.17 High-School Graduation Rates by enlistment month . . . . . . . . . . . . . . . . . .

94

3.18 Probability of joining the Airforce . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96

3.19 Graph with Residuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

99

3A-1 Shift in the IQ distribution of high-goiter areas after iodization . . . . . . . . . . . . 107

xiii

Chapter 1

The Impact of Iodine Deficiency
Eradication on Schooling:
Evidence from the Introduction of
Iodized Salt in Switzerland
1.1

Introduction

Health is an important determinant of a population’s prosperity and economic outcomes. Healthier
people make better workers, and cross-country differences in health account for a significant part of
cross-country income inequality (see Weil (2007)). Nutrition is inextricably linked to a population’s
health capital. Malnutrition, especially when it occurs early in life, can have serious detrimental
effects on a person’s lifetime productivity and economic prospects. Micronutrient deficiencies are a
common source of malnutrition, caused by insufficient intake of necessary vitamins and minerals.
Iodine is one such micronutrient.
Iodine deficiency is the leading cause of preventable mental retardation in the world. The WHO
estimates that nearly 50 million people suffer some degree of mental impairment due to iodine
deficiency1 . Two billion people, one third of the world’s population, are at risk, in the sense that their
iodine intake is considered insufficient. According to WHO’s Global Database on Iodine Deficiency,
more than 285 million children receive inadequate amounts of iodine in their diet 2 . Despite efforts
to decrease the prevalence of iodine deficiency in the 1990s, there are still 38 million children born
annually at the risk of developing iodine deficiency disorders. The most vulnerable areas in the
world are South Asia and Central and Eastern Europe (UNICEF 2008).
1 Source:

WHO, http://www.who.int/features/qa/17/en/index.html.

2 Source:

de Benoist, Andersson, Egli, Takkouche and Allen, eds (2004).

1

2
Iodine Deficiency Disorders (IDD) is a generic name given to all defects resulting from a lack of
iodine in the diet. The consequences of iodine deficiency are both visible and invisible; cretinism,
goiter, short stature, and deaf-mutism are among the defects related to iodine deficiency that are
easily detectable. However, iodine deficiency in utero and the first three months of life results in
various degrees of brain damage that might be harder to observe in an affected population. In a
meta-analysis of 21 studies, Bleichrodt and Born estimate that eliminating iodine deficiency increases
average IQ by 13.5 points (Bleichrodt and Born 1994). Such an improvement in cognitive ability
should be reflected in economic outcomes through increased human capital in the form of health,
which will affect one’s educational attainment and labor productivity.
This paper estimates the effects of iodine deficiency eradication on schooling, using data from
Switzerland. Switzerland, due to its geography, had a high prevalence of iodine deficiency, and it
was the first country in the world to introduce iodized salt in 1922. Iodized salt proved a costeffective measure to eradicate visible goiter, deaf-mutism, and cretinism, which were endemic before
its introduction. The invisible effects of iodine deficiency on mental development and cognitive
ability were not fully understood at the time, and public health authorities did not know that they
were fighting against mental retardation as well as endemic goiter. As a result of the countrywide
iodization campaign, there were no more endemic cretins born after 1930, deaf-mutism rates dropped
significantly, and goiter disappeared in children and young recruits (Bürgi, Supersaxo and Selz 1990).
I find that, apart from these effects, salt iodization also had a significant impact on graduation rates
of those born in highly-deficient areas.
I use microdata from the comprehensive 1970 Swiss Census, combined with data on the preexisting variation of iodine deficiency across 185 Swiss districts. I identify the effect of iodization on
the probability of graduating from upper-secondary and tertiary education in three different ways.
First, I look at the education of cohorts born before and after iodized salt first became available
in 1922. I examine high- and low- goiter districts separately, since I expect high-goiter areas to be
more affected by iodization. Second, I corroborate my findings by taking advantage of differences
in the timing of adoption of iodized salt across cantons. Using annual, canton-level data on iodized
salt sales as a percentage of total salt sales, I examine how the penetration of iodized salt affected
schooling outcomes of cohorts exposed to differential degrees of iodine treatment. Finally, I use a
fuzzy regression discontinuity design, where I identify the effect of iodization by looking at sudden
jumps in iodized salt sales, and then compare education outcomes for cohorts born right before and
right after the jump.
The introduction of iodized salt in Switzerland was the first major nutritional intervention to
ever take place, yet its productivity effects have not been estimated. Given that iodine deficiency
is still a problem in much of the developing world, it is useful to know how Swiss productivity was
affected by the introduction of iodized salt.
This paper’s contribution can also be seen from a different perspective. The successful campaign
for the eradication of iodine deficiency provides a rare opportunity to examine the effects of an
“injection of IQ” on schooling outcomes. Iodization had a significant impact on the cognitive ability

3
of people born in the worst-afflicted regions of Switzerland. Heckman and Vytlacil (2001) describe
some problems in the identification of returns to schooling, apart from the usual omitted-ability bias.
The real problem, they note, “is that ability and schooling appear to be inseparable –all interaction
and no main effects– even if ability is perfectly observed”. The advent of iodized salt made whole
cohorts, namely those born after the intervention in previously deficient areas, smarter, through its
effect on brain development in utero. In this light, by using the eradication of iodine deficiency as
a shock to cognitive ability in Switzerland, we can decouple ability and schooling levels, and study
the effects of the former on the latter.
I find that iodization increased the probability of graduating from upper-secondary and tertiary
education for those born in previously highly-deficient areas by around 1 percentage point3 . I also
find that these results are driven by the effect of iodization on females, where iodization explains
between 9 and 14% of the total change in graduation rates from upper education levels over this
period. This could be due to cross-gender differences in the impact of intrauterine iodine deficiency
on brain development. In this case, the same level of iodine treatment will have different effects in
male and female fetuses, which will translate into different effects on schooling later in life. This
is consistent with medical evidence showing that females are more prone to thyroid disorders than
men4 . However, the same findings could be explained by household differences in the sensitivity of
schooling decisions to changes in cognitive ability for male and female offspring. In this case, a given
increase in cognitive ability might raise female schooling by more than male schooling.
Overall, there is strong evidence that eradicating iodine deficiency had a significant impact on
the educational attainment of cohorts born after the introduction of iodized salt in those regions
where one would expect to find such an effect, namely regions which were highly deficient prior to
the intervention.
The rest of the paper is organized as follows: section 1.2 provides a short review of the relevant
literature from Economics. Section 1.3 provides some background on iodine deficiency disorders, and
section 1.4 describes the campaign for salt iodization in Switzerland. Section 1.5 describes the data
from the 1970 Swiss Census, as well as the dataset on the pre-existing prevalence of iodine deficiency
across Swiss municipalities. Section 1.6 introduces additional data, which show the decline in deafmutism following the introduction of iodized salt. Sections 1.7, 1.8, and 1.9 describe the identification
strategy and contain the empirical findings. Section 1.10 provides an interpretation of the results,
and section 1.11 concludes.
3 Primary

and lower-secondary education in Switzerland were obligatory and universal during the period that I
examine, therefore it is not surprising that iodization did not affect graduation rates from these lower education
levels.

4 Thyroid-related

disorders are directly linked to iodine intake, because the thyroid gland is where most of the
body’s stock of iodine is stored and used in the production of hormones which regulate metabolism.

4

1.2

Related literature

I explore the link from better health to improved economic performance by quantifying the effects of
a wide-reaching public health intervention. Other papers exploring the economic effects of smallerscale interventions are, for example, Miguel and Kremer (2001), who study the effects of introducing
deworming drugs in Kenyan schools, taking into account the positive externalities arising from
lower infection rates. In addition, Bleakley (2006) examines the effects of eradicating hookworm
in the American South, and also finds positive effects on school attendance and future earnings.
Finally, Lucas (2006) studies the effects of malaria eradication, and finds evidence of increased
female educational attainment and female literacy as a result of the lower prevalence of the disease.
I concentrate attention on the effects of irreversible mental damage resulting from iodine deficiency in utero and the first three months of life. In that respect, this paper is closely related
to the economic literature on the importance of fetal and early age health inputs on subsequent
economic outcomes. One example of this literature is Douglas Almond’s paper on the long-lasting
effects of the Spanish Influenza of 1918 (Almond 2006). Using U.S. Census data, Almond exploits
the sharp timing and the geographic variation in the severity of the pandemic. He finds that the
cohorts exposed to the virus in utero display lower education attainment rates, lower income and
socioeconomic status, higher rates of physical disability and increased dependence on state transfer
programs. In another paper, Case and Paxson (2008) highlight the importance of in utero and early
childhood health inputs for the adult height of an individual, which in turn is associated with higher
earnings. In particular, taller children perform better on cognitive ability tests, and this higher
performance explains a big part of the variation in earnings later in life.

1.3

Background on Iodine Deficiency Disorders

Iodine is a necessary micronutrient, found in very small quantities in the human body5 . Most of the
body’s iodine is located in the thyroid gland. Iodine is essential in the synthesis of the two thyroid
hormones, thyroxine (T4) and Triiodothyronine (T3). These two hormones regulate metabolism and
“play a determining part in early growth and development of most organs, especially of the brain”
(Delange 2001).
When the thyroid does not receive sufficient amounts of iodine it adapts by enlarging in order
to maximize the use of available iodine. This enlargement is called a goiter. Goiters can occur at
any point in one’s lifetime, whenever the iodine intake is not sufficient6 . Some goiters are reversible,
especially in young individuals. By providing iodine to schoolchildren, numerous studies have shown
that their goiters disappeared (see, for example, Marine and Kimball (1921) for details on their 1917
5 The

recommended daily intake of iodine is 50µg for people aged 0-6 months, 90µg for people aged 6 months-6
years, 120µg for ages 7 to 10, 150µg during adolescence and adulthood, and 200-300µg during pregnancy and
lactation (World Health Organization 1996).

6 For

adults, adequate iodine intake is in the range of 100-199µg/day. There are three degrees of iodine deficiency:
mild (50-99µg/day), moderate (20-49µg/day) and severe (<20µg/day) (ICCIDD, UNICEF and WHO 2001, Table
5, p.36.).

5
experiment in Akron, Ohio). Reversing goiter in adults is harder, especially when they have been
subject to iodine deficiency for many years.
Goiter is a visible effect of iodine deficiency. Apart from goiter, however, iodine deficiency can
have irreversible and harder to observe consequences if it occurs in utero and in the first three months
of life. If production of the thyroid hormones is affected by lack of iodine, normal development of the
fetus or the infant is put at risk. Inadequate iodine intake on the part of the pregnant mother results
in mental retardation for the newborn, and it is associated with a higher incidence of abortions,
stillbirths, low birth weight and increased perinatal and infant mortality1 . Iodine deficiency results
in various degrees of mental retardation and abnormal brain development, which might even go
undetected in a population. In the worst case scenario, severe iodine deficiency causes cretinism.
Cretinism is an acute condition characterized by a combination of mental retardation, stunting and
physical deformation. Cretins are often deaf-mute and have goiters. In endemic areas, cretinism
can affect up to 15% of the population (de Benoist et al., eds 2004). Bleichrodt and Born (1994)
estimate that the average IQ of iodine-deficient groups is 13.5 points lower than the non-deficient
groups7 . If this is true, then iodine deficiency should have sizable economic effects for any afflicted
population.
Endemic goiter and endemic cretinism are primarily due to the geographic location of a population. The main store of iodine is the ocean. As ocean water evaporates, iodine falls on the upper
layers of soil through rainfall. Therefore, geographic areas close to the ocean provide adequate
amounts of iodine to their population, either through the air, the drinking water, or the consumption of iodine-rich foods such as fish, seaweed, meat, or vegetables fed or grown on iodine-rich soil.
On the contrary, some geographic areas are naturally iodine-poor. For example, regions subject to
heavy rain may be iodine-poor due to soil erosion. Furthermore, during the last Ice Age (18,0008,000 years ago), ground erosion brought about by glaciation stripped the soil of its iodine in some
parts of the world. Because it takes thousands of years for rain water to replenish the superficial
layers of soil with iodine, the iodine-content of the soil and water of those regions is extremely low.
For this reason, places such as the Andes, the Alps, the Pyrenees, the Himalayas and, in general,
regions that have been subjected to intense glaciation in the past, are iodine-poor (Koutras, Matovinovic and Vought 1980). Also, there are cases in which even though the soil contains iodine, the
latter never reaches the population. This may happen if, for example, the morphology of the soil is
such that iodine is easily absorbed by the soil, making it harder for iodine to reach the plants, and,
consequently, humans8 (Koutras 1980, p.256).
I use data on goiter prevalence among male recruits as a proxy for the underlying geographic
7 This

estimate is based on a meta-analysis of 21 studies of the effect of iodine deficiency on cognitive ability.

8 Sporadic

goiter might occur as a result of the consumption of certain foods. For example, vegetables in the
Cruciferae family (such as cabbage and turnip) are known to have goitrogenic properties. Geologic factors seem
responsible for the goitrogenic content of both food and water in some areas. Regular and high consumption of such
foods can cause the thyroid gland to enlarge. Goiters may also be caused by an excess of iodine consumption,
although this is rare. Genetic predisposition also seems to play a secondary role in the appearance of goiter.
Goiters caused by these factors are sporadic and do not occur regularly in a population. On the contrary, endemic
goiter and endemic cretinism are the result of iodine deficiency.

6
distribution of iodine deficiency. The Swiss Alps were known for the endemic nature of goiter and
cretinism. The data on goiter prevalence confirm the link between iodine deficiency and goiter
prevalence: the canton of Ticino ranked the lowest among cantons in goiter prevalence. This is to
be expected, since Ticino is in the southernmost part of Switzerland, bordering Italy and enjoying
a milder climate, proximity to the Mediterranean Sea and possibly more iodine-rich foods coming
from Italy than the rest of the country. Another canton with unusually low goiter prevalence is
Vaud. Historically, Vaud had an exclusive salt mine, which was rich in iodine (Bürgi et al. 1990,
p.581). After iodized salt was introduced in Switzerland, cretinism was eradicated and goiters
disappeared from the younger generations, further proving that it was iodine deficiency that caused
these abnormalities in the population.
Iodine has been explicitly used in the treatment of goiter ever since Bernard Courtois isolated
it as an element in 18119. The idea that endemic goiter is due to iodine deficiency was first put
forward in 1846, by Jean-Louis Prévost and A.C. Maffoni (Prévost and Maffoni 1846). Even though
cretinism was associated with endemic goiter, the crucial role of iodine in mental development was
not understood until more than a century later. When large-scale interventions of iodine supplementation took place around the 1920s and after, the objective was goiter eradication. People did
not know that they were also fighting against mental retardation.
After doctors started prescribing iodide to their patients in order to fight goiter, toxic side-effects
resulting from over-dosing triggered opposition to the universal use of iodine. It wasn’t until 1917
that the first larger-scale iodine supplementation program took place in Akron, Ohio. From 1917 to
1922, schoolgirls from 5th grade and above were given sodium iodide regularly, in the form of syrup,
under the direction of David Marine and his assistant, O.P. Kimball. When it began, the intervention
was very controversial, but its undeniable success paved the way for larger-scale programs in the
USA and in Europe.
Iodized salt started circulating in Switzerland in 1922. Almost simultaneously, fortification of
salt with iodine began in the USA, where iodized salt first appeared in 1924. Both interventions
eliminated endemic cretinism and goiter in children, and they decreased goiter prevalence in adults,
even though they were followed by an initial spike in goiter-related surgeries and deaths, which then
subsided10 . More information about the program in Switzerland is provided in the following section.
9 However,

iodine-rich foods and plants, such as seaweed, were used by ancient civilizations, such as the Chinese
and the Greeks, to treat the swelling of the neck before the isolation of iodine as an element.

10 This

adverse consequence of iodine supplementation was due to the existence of nodular goiters in the population.
Nodular goiters were caused by chronic iodine deficiency. Nodular goiters may become toxic following a sudden
increase in iodine intake after a long period of deprivation. This side-effect of iodization is known as iodine-induced
hyperthyroidism).

7

1.4

The campaign for salt iodization in Switzerland

Accounts of goiter and cretinism in Switzerland date back to antiquity11 , although the link between
two conditions was not yet known until much later. Goiter was attributed to something in the water,
rather than something absent from it, and it was often confounded with other diseases visible in the
neck12 .
The late 18th and early 19th centuries were marked by considerable advances in medical knowledge on the thyroid and its operation. These advances were followed by increased interest in goiter
and cretinism, and efforts to assess their geographical prevalence and the extent of the problem. During Napoleonic Wars, the low performance of Swiss recruits for the French Army troubled Napoleon
and the local authorities in today’s canton of Valais. Under Napoleon’s orders, a survey was conducted, according to which there were 4,000 cretins among 70,000 inhabitants (this is an extremely
high prevalence of 5.7%)13 . Since then, other epidemiologic studies followed, usually focusing on
recruits or schoolchildren. However, the standards of goiter and cretinism classification differed from
author to author, so it is impossible to make meaningful comparative studies using different sources.
Nevertheless, the aforementioned studies revealed the extent of the problem. Bircher (1883), in
his monograph, published goiter data taken from recruits during the period 1875-1880 for all towns
and villages in Switzerland, and noticed that even localities adjacent to each other might differ a
lot in their goiter prevalence. I use this rich database to construct measures of iodine deficiency
across localities. In 1889, Theodor Kocher published a goiter study of schoolchildren in Bern, which
revealed a total goiter prevalence ranging from 20 to 100% (Kocher 1889). Because of the great
prevalence of goiter and cretinism, the medical profession and public health authorities focused
attention on the etiology and on ways to provide prophylaxis to the population.
As a result, a Swiss Committee for the study of goiter was established in 1907. At that time,
goiter was still attributed to some agent in the drinking water, even though experiments with iodine
supplementation for the treatment of goiter were already taking place in France and, later, in the
USA. Right before his death in 1917, Kocher suggested goiter treatment with small doses of iodine
(Bürgi et al. 1990).
The first canton to iodize salt was Appenzell-Ausserrhoden, where iodization started in February 1922, with the initiative of a local doctor, H. Eggenberger. In June 1922, the Swiss Goitre
Commission recommended the addition of small amounts of iodine in salt and the additional weekly
consumption of iodine tablets by schoolchildren. Children were encouraged to take more iodine
than the general population, because they were less likely to develop toxic symptoms. At the same
11 Writers

as far back as the 1st century BC wrote about a swelling of the neck, found in people living in the Swiss
Alps. For example, the architect Vitruvius wrote: “[...] the Medulli in the Alps have a kind of water, from drinking
which they get a swelling of the neck” (Langer 1960, p.10-11, translated from Vitruvius, P.M., De Architectura,
lib. VIII, 3, 20).

12 For

instance, some writers attributed goiter to a lack of minerals in the water or thought that it was the result
of rickets. Others came closer to the true cause, linking cretinism to distance from the sea and to air quality
(Langer 1960).

13 Taken

from Bürgi et al. (1990), original reference is: Merke, F. (1971) Geschichte und Ikonographie des endemisches Kropfes und Kretinismus. Berne: H. Huber.

8
time, consumption of the “new salt” would remain voluntary and non-iodized salt would still be
available(Bürgi et al. 1990).
It was in November 1922 that United Swiss Rhine Salt Works (USRSW)14 started adding iodine
to salt (5 mg. KI or 3.75 mg. I per kg salt) and selling the new product at the same price as
non-iodized salt. Even before that date, though, iodine prophylaxis had become popular, especially
since Marine and Kimball’s experiments with schoolchildren in the USA. After the recommendations
of the Swiss Goiter Committee and the success of salt iodization in Appenzell-Ausserrhoden, the
other cantons started allowing the sale of iodized salt in their markets. Figure 1.1 is a graph showing
the population-weighted average iodized salt sales a a percentage of total salt sales in Switzerland.
As is clear from Figure 1.1, iodized salt became popular fast.

.6
.4
.2
0

Iodized salt sales as % of total salt sales

.8

Figure 1.1: Iodized salt penetration in Switzerland

1910

1920

1930
Year

1940

1950

source: Wespi (1962) and 1970 Swiss Census

Not all cantons introduced iodized salt simultaneously, though. For instance, Valais iodized in
1925, Zürich in 1932 and Bern in 193615. On the other hand, Aargau and Basel-Land didn’t iodize
until 1952 and 1950 respectively. In 1925 fewer than one fourth of cantons had iodized salt sales
that exceeded 60% of total salt sales. By 1945 fewer than one fourth of cantons had salt sales that
were below 20% of total salt sales. By 1955, iodized salt sales exceeded 60% of total salt sales in
14 USRSW

was “the exclusive supplier of salt to 24 of the 25 cantons” of Switzerland, the exception being the canton
of Vaud (Bürgi et al. 1990, p.582).

15 This

is the first year that the cantons’ iodized sales exceeded 40% of total salt sales.

9
all cantons, and in many of them, only iodized salt was sold and consumed (Wespi 1962). On a
national scale, the iodine content of salt was raised in subsequent years, to 7.5 mg. in 1962 and to
15 mg. iodide per kg salt in 1980 (Bürgi et al. 1990).
The success of the iodization program was indisputable. According to Bürgi et al. (1990), “no
new endemic cretins born after 1930 have been identified” (p.577). Deaf-mutism rates fell sharply
for cohorts born after 1922 (see section 1.6). In Appenzell-Ausserrhoden, which was the first canton
to provide iodized salt to its inhabitants, the prevalence of goiter in newborns fell from 20% to 6.4%
within the first year after iodization. The prevalence dropped further when, in later years, the iodine
content of salt was raised. The beneficial effects on iodization were also seen in the increased height
of 6-year-olds entering school, as well as young recruits. In the city of Lausanne, 23.7% of young
recruits had large goiters in 1924/1925, but the figure had dropped to 0.2% by 1983-1987 (Bürgi et
al. 1990).

1.5

The 1970 Swiss Census and Bircher’s monograph

The main source of data is the complete 1970 Swiss Census (Federal Statistical Office 1970). Switzerland is made up of 26 cantons (comparable to US States, but much smaller), 184 districts and 2896
municipalities. The 1970 Census contains detailed information on an individual’s year and municipality of birth, as well as other geographic, demographic, work and migration variables. In particular,
I know the municipality and year of birth for each individual. Having detailed information on the
place of birth is important, since the location of the mother during her pregnancy will determine
the extent to which she got adequate amounts of iodine in her diet16 .
I limit the sample to all Swiss-born individuals interviewed in the 1970 Census, who were born
before 1950. The oldest person in my sample was born in 1864. I estimate the effects of iodization
both on all cohorts but also on a smaller age range, which is less likely to be affected by selection
into longevity. The total number of observations is 3,086,287, comprised of 1,503,257 males and
1,583,030 females.

16 Unfortunately,

the Swiss Census asks no income questions. Swiss income data can be found in the Swiss Labor
Survey or the Swiss Household Panel, but these datasets cannot be used here, because the former does not include
place of birth as a question, while the latter is too recent (the first wave of interviews took place in 1999).

Table 1.1: Summary Statistics: graduation rates

Number of observations
% having completed obligatory education
% having completed secondary education
% having completed tertiary education

Number of observations
% having completed obligatory education
% having completed secondary education
% having completed tertiary education

Number of observations
% having completed obligatory education
% having completed secondary education
% having completed tertiary education

Born before
1922

Born after
1922

1,503,277
99.74%
47.00%
9.03%

1,583,010
99.77%
61.86%
17.02%

Born before
1922

Born after
1922

706,047
99.78%
62.85%
11.96%

797,210
99.78%
74.20%
19.67%

Born before
1922

Born after
1922

797,230
99.70%
32.96%
6.43%

785,800
99.76%
49.33%
14.33%

Whole sample
Born before
jump in iodized
salt sales
2,252,036
99.76%
51.14%
11.37%
Males
Born before
jump in iodized
salt sales
1,084,703
99.79%
66.21%
14.53%
Females
Born before
jump in iodized
salt sales
1,167,333
99.73%
37.13%
8.43%

Born after
jump in iodized
salt sales
834,251
99.74%
64.02%
17.88%
Born after
jump in iodized
salt sales
418,554
99.74%
75.77%
19.99%
Born after
jump in iodized
salt sales
415,697
99.74%
52.18%
15.75%

Total

3,086,287
99.75%
54.62%
13.13%
Total

1,503,257
99.78%
68.87%
16.05%
Total

1,583,030
99.73%
41.09%
10.35%

10

Table 1.2: Summary Statistics (continued): graduation rates by region
Whole sample
Born before 1922

Number of observations
% having completed:
obligatory education
secondary education
tertiary education

High-goiter
373,156

Other regions
1,130,121

High-goiter
402,148

99.72%
44.89%
8.64%

99.74%
47.69%
9.16%

99.75%
60.17%
17.14%

Born before 1922

Number of observations
% having completed:
obligatory education
secondary education
tertiary education

Born after 1922

99.77%
62.43%
16.98%
Males
Born after 1922

High-goiter
176,895

Other regions
529,152

High-goiter
202,422

99.77%
60.25%
11.20%

99.78%
63.72%
12.22%

99.75%
72.06%
19.30%

Born before 1922

Other regions
1,180,862

Born before
jump in iodized
salt sales
High-goiter
Other regions
649,013
1,603,023

Other regions
594,788

99.74%
50.02%
11.72%

99.76%
51.59%
11.23%

Born before
jump in iodized
salt sales
High-goiter
Other regions
316,636
768,067

99.78%
99.77%
99.80%
74.93%
64.38%
66.96%
19.80%
14.65%
14.48%
Females
Born after 1922
Born before
jump in iodized
salt sales
High-goiter
Other regions
High-goiter
Other regions
199,726
586,074
332,377
834,956

Born after
jump in iodized
salt sales
High-goiter
Other regions
126,291
707,960
99.72%
67.18%
19.87%

99.75%
63.45%
17.52%

Born after
jump in iodized
salt sales
High-goiter
Other regions
62,681
355,873
99.70%
77.52%
19.93%

99.75%
75.46%
20.01%

Born after
jump in iodized
salt sales
High-goiter
Other regions
63,610
352,087

High-goiter
Other regions
Number of observations
196,261
600,969
% having completed:
obligatory education
99.68%
99.71%
99.75%
99.76%
99.72%
99.73%
99.74%
secondary education
31.05%
33.58%
48.12%
49.74%
36.35%
37.45%
56.99%
tertiary education
6.34%
6.47%
14.95%
14.11%
8.93%
8.23%
19.81%
Notes: High-goiter regions are defined as those districts belonging to the top 25% of the population-weighted goiter distribution

99.74%
51.31%
15.02%

11

12
I use two education outcomes; indicator variables for completing upper-level secondary education,
and completing tertiary education. Obligatory education consists of 6 years of primary education and
3 years of lower-secondary education, and it is universal. Upper-level secondary education17 ranges
between an extra 3 to 5 years after having completed 9 years of obligatory education, depending
on whether someone is preparing for university-level education or just completing a professional
apprenticeship. Tertiary-level education corresponds to a minimum of 13 years of schooling, and
includes studies in colleges, universities, but also upper-level vocational schools.
Tables 1.1 and 1.2 provide some summary statistics for the sample as a whole, as well as separately
by gender. Obligatory education was universal; more than 99% of the population completed at least
9 years of schooling. Secondary education was less pervasive, but still rather common, as roughly
one in two got to this level of schooling. When it comes to tertiary education, however, even if we
look at the younger cohorts, those born after iodization, we see that only one in five men attended
a higher education establishment. The rates for women are consistently lower, even though the
percentage of women advancing to secondary and tertiary levels of education grew by more than the
corresponding percentages for men over my period of examination.

17 I

will henceforth refer to upper-level secondary education simply as “secondary education”, but it is understood
that lower-level secondary education is part of obligatory schooling in Switzerland.

Figure 1.2: Bircher’s Data on goiter in recruits

Distribution of goiter in Switzerland in 1880 - Municipality-level data

Goiter rates
5% or lower
6% - 10%
11% - 20%
21% and up
0

20

®

40
Miles

13

Source: Bircher (1883) and Swiss Federal Statistical Office

10

14
I combine the Census microdata with data on the pre-existing variation of iodine deficiency in
Switzerland. These data come from a monograph written by H. Bircher and published in 1883
(Bircher 1883). Bircher collected data on goiter in Swiss recruits during the period 1875-1880, for
every municipality (village) in every district of every canton in Switzerland. For each locality, he
listed the total number of recruits with goiter that enlisted in the 6-year period, from 1875 to 1880.
Bircher also used 1880 Census data to construct measures of the underlying prevalence of goiter in
the population where the recruits came from.
Because many municipalities have been divided or merged with others since Bircher’s time,
I construct new measures of underlying goiter prevalence in the population at the district level
(one administrative level up from the municipality level), using the same method used by Bircher
(Bircher 1883)18 . Figure 1.2 is a map of Switzerland, showing the geographic variation in goiter, as
depicted by Bircher’s data. It is evident from the map that regions closest to the Swiss Alps were
the ones that were most affected by goiter, whereas regions close to the Mediterranean were not so
deficient in iodine.
Bircher’s data on goiter prevalence are old and prone to measurement error. The data were
collected in 1875-1880, so that the goiter distribution in Switzerland in 1921 might have looked
quite different. Also, to the extent that cretins and youths with extreme cases of goiter would
not enlist in the army in the first place, the data on goiter might be understating the degree of
iodine deficiency for some regions. They might also give a distorted picture of the relative position
of two localities with respect to their pre-existing variation in goiter rates. For instance, a given
locality with a serious iodine deficiency problem might have supplied fewer recruits than another,
less-afflicted locality. In Bircher’s data, the latter locality will be documented as having a more
serious iodine deficiency problem compared to the locality that supplied fewer recruits because of
its higher goiter and cretinism prevalence.
Although measurement error might be an issue, Bircher’s data have been shown to be very good
proxies for the underlying levels of iodine deficiency, in the sense that they correlated well with
estimates of the iodine content of soil and water across localities, whenever such measures have been
available (Bürgi et al. 1990). Bircher’s goiter data also correlate with rates of deaf-mutism on the
eve of iodization (see Figure 1.4 and Section 1.6). They also show that goiter rates displayed great
variation, even within a given district, among villages within a small distance from each other. In
any case, in order to minimize the effects of any measurement error in Bircher’s data, I break the
Census sample down in broad categories according to a district’s goiter prevalence. Also, in some
specifications I only use data in the top and bottom tail of the goiter distribution, so that the exact
distribution of goiter rates doesn’t matter very much.
Figure 1.3 is a population-weighted histogram of the goiter prevalence in 1880, as it appears in
my sample. Each observation from the 1970 Census is matched to an observation from Bircher’s
data, according to one’s district of birth. The red line in Figure 1.3 marks the 75th percentile cutoff
18 Bircher’s

formula for constructing the prevalence of goiter in the population (in percentage terms) is:
The ratio is divided by 6 to get the yearly average number of goitrous
recruits, and it is multiplied by 100 because he estimates that 100 inhabitants supply approximately 1 recruit.
number of goitrous recruits×100×100
.
6×population in locality

15

0

.05

Density
.1

.15

.2

Figure 1.3: Histogram of goiter prevalence in Switzerland

0

10
20
Goiter prevalence in district of birth in 1880

30

source: Bircher(1883) and 1970 Swiss Census

in my sample, which corresponds to a goiter prevalence of 11.7%. For my analysis, I consider any
district with a goiter prevalence equal to 11.7% or higher to be a “high-goiter district”. Outcomes
are similar when I modify this cutoff level to reflect the top 20% or 30% of the population-weighed
goiter distribution.

1.6

Iodized Salt and the decline in deaf-mutism

The success of the iodization campaign in Switzerland was reflected very clearly in the steep drop in
the number of deaf-mutes among children born after 1922. Wespi (1945)19 collected a near-complete
count of deaf-mute schoolchildren who attended specialized institutions during the period 1922-1942,
along with data on their year of birth and the residence of their parents at the time (assumed to
be the children’s canton of birth). Since most admissions occur at the ages 7-10, Wespi limited his
analysis to children born in the period 1915-1932.
Figure 1.4 combines Wespi’s data on deaf-mutism and Bircher’s data on goiter prevalence in
1883 (aggregated at the canton level) in a scatterplot. I divide Wespi’s data in two categories:
cohorts born in 1915-1922 (before iodization), and cohorts born in 1923-1932 (after iodization).
Each observation is a canton-time period combination, so that there are two observations for every
19 I

am grateful to Prof. Dr. Hans Bürgi for kindly sharing this source with me.

16
canton, which correspond to the same rate of goiter but different rates of deaf-mutism. As expected,
rates of deaf-mutism (per 10,000 people) for cohorts born prior to iodization (1915-1922) correlate
positively with rates of goiter in recruits, even though the goiter data was collected many decades
earlier20 . On the contrary, deaf-mutism rates for cohorts born after iodization (1923-1932) do not
correlate at all with goiter rates, as they were recorded prior to the intervention.

Average number of deafmutes per 10000
0
1
2
3
4

Figure 1.4: Deaf-mutism and 1883 Goiter Prevalence before and after iodization

0

5
10
15
1880 Goiter prevalence in canton of birth
deafmutes born 1915−1922
Fitted values, 1915−1922

20

deafmutes born 1923−1932
Fitted values, 1923−1932

source: Bircher (1883), Wespi (1945), and 1970 Swiss Census

Figure 1.5 is a plot of the prevalence of deaf-mutism in each cohort born in 1915-1932, using
Wespi’s data, against the country-wide penetration of iodized salt. It is clear from this figure that
the prevalence of deaf-mutism among schoolchildren decreased rapidly for those born after 1922,
which coincides with the introduction of iodized salt.
20 Wespi

(1945) speculates that this positive relationship might even be understated, because in highly-deficient
areas admission into special institutions may have been incomplete. This may have been the case if in these
highly-affected areas endemic deaf-mutes also had other medical or learning impediments, preventing them from
attending a school.

17

40
10
20
30
Iodized salt as % of total salt sales

10 12 14 16 18
8
6
4

0

0

2

Deaf−mutes per 10000 children

Figure 1.5: Deaf-mutism and Iodized Salt penetration

1915

1920

1925
Year of Birth

1930

Deaf−mutes per 10000 children
Iodized salt as % of total salt sales
source: reproduced from Wespi (1945)

1.7

A first look at the data:
using 1922 as the cutoff date for treatment

Iodized salt first became available in 1922. During that year, iodized salt only circulated in one of
the smallest cantons of the Swiss Confederation, Appenzell-Ausserrhoden, following the initiative
of a local doctor. However, the Swiss Goitre Commission was founded in 1922, and a nationwide
campaign to eradicate goiters from the population was launched. Medical awareness of the potential
benefits of iodine supplementation increased, and iodized salt sales picked up fast, as manifested by
Figure 1.1. Rates of deaf-mutism fell rapidly after 1922 (Figure 1.5), further suggesting that this
year marked the beginning of the intervention. People who were in utero during or after 1922 were
clearly exposed to a different environment than those born before 1922.
In a first look at the data, I compare cohorts born before and after 1922 in a differences-indifferences framework, where I also control for a district-specific time trend. I look at the probability
of graduating from secondary and tertiary education. Treated individuals are those born in highgoiter districts after 1922. A high-goiter district is identified as one being in the top 25% of the
population-weighted goiter distribution21 . I expect high-goiter districts to be the ones benefiting
from the introduction of iodized salt, since underlying iodine deficiency was the most severe in those
21 Results

are similar when I consider the top 20% and the top 30% of the population-weighted goiter distribution.

18
regions. On the contrary, people born in low-goiter districts were less likely to benefit from the
treatment, since their iodine intake in utero was already higher.
Figures 1.6 and 1.7 are regression-adjusted graphs, which show how the probability of graduating
from secondary (Figure 1.6) and tertiary (Figure 1.7) education changed by year of birth, after
controlling for district fixed effects and a district-specific trend. The top panels correspond to the
coefficients for each cohort, separately for those born in high-goiter and non-high-goiter districts,
whereas the bottom panels show the difference between high- and non-high-goiter districts for each
cohort-specific coefficient. It is hard to discern a break in the trends on the top panels, but it is
easier from the bottom panels to see that something changed for those born after 1922 in high-goiter
districts. Whereas there was a converging trend between the two district groups prior to 1922, this
trend seems to have stalled or reversed after 1922.
This departure from trend becomes even more apparent when one looks at males and females
separately. Figures 1.8 and 1.9 show the regression-adjusted difference in coefficients between highand non-high-goiter districts for secondary (Figure 1.8) and tertiary education (Figure 1.9), by
gender. In the case of secondary education, for both males and females, there is a clear departure
from trend starting around 1922, which coincides with the launch of the iodization campaign, and
which continues as iodized salt spread around the country. For tertiary education, males do not
seem affected, but for females there is, again, a change in the trend right around 1922.
Table 1.3 shows results of the regression-equivalent of the above graphs, first for the sample as
a whole, and then separately by gender. I look at all cohorts, but also at a smaller age range, those
born in 1910-1935. More specifically, I run the following regression for an individual i born in district
d in year y:
outcomeidy = α + β· 1 (Born in high-goiter district) X 1 (Born after 1922)
+ District of birth Fixed Effects
+ District of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy
The outcomes which I look at are graduation from secondary- and tertiary-level education. The
coefficient of interest is β. District of birth fixed effects take away the effect of any omitted districtspecific, time-invariant characteristics, which might affect an individual’s education. Cohort fixed
effects control for any unobservable, country-wide characteristics that are common to each cohort in
my sample. Finally, a district-specific linear time trend removes the effect of factors which change
gradually over time and might affect the educational outcomes of cohorts born in a given district.
Such factors include, for example, better access to educational opportunities over time, improving
general health conditions, and ever-lowering transportation and communication costs. The coefficient
β will capture the departure of educational outcomes from this trend for those born in high-goiter
districts after 1922. Standard errors are clustered at the district-cohort level.
It is evident from Table 1.3 that there was a departure from trend for high-goiter districts in
1922. Whether we look at all cohorts, or only at the cohorts born closer to 1922, the probability
of graduating from secondary education increases by between 1 and 1.6 percentage points, whereas

19

Table 1.3: Salt sales and education: Coefficient on high-goiter dummy X Born after 1922 dummy

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

All cohorts
1.59166***
(0.35161)
3086287
0.0923
0.40358**
(0.19633)
3086287
0.0305
All cohorts
0.27304
(0.37112)
1503257
0.0862
0.09915
(0.29536)
1503257
0.0319
All cohorts
2.75442***
(0.47548)
1583030
0.1145
0.64007***
(0.23198)
1583030
0.0345

WHOLE SAMPLE
Born 1910-1935
1.04939***
(0.38659)
1435764
0.073
0.47352*
(0.25965)
1435764
0.0186
MALES ONLY
Born 1910-1935
0.50586
(0.49232)
721813
0.0722
0.16217
(0.38939)
721813
0.0221
FEMALES ONLY
Born 1910-1935
1.29066**
(0.52914)
713951
0.095
0.70804**
(0.30726)
713951
0.0172

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
All regressions include district of birth and cohort fixed effects, as well as a district-specific trend.
Standard errors in parentheses, clustered at the district-year of birth level.

20
the effect is smaller for the probability of graduating from tertiary education, which is a little more
than 0.4 percentage points.
When we look into each gender separately, it becomes clear that these effects are driven by
females. For women born after 1922 in high-goiter districts, the probability of graduating from
secondary-level education increases by 2.7 percentage points. This is a big change; according to Table
1.2, the percentage of women graduating from secondary education increased by 17.07 percentage
points for those born after 1922 compared to earlier cohorts. The coefficient on secondary education
for women indicates that almost one-sixth or 16% of this change was due to iodization.
The change in probability for tertiary education is, again, much smaller, around 0.65 percentage
points. Given that tertiary education graduation rates for women born in high-goiter districts
increased by 8.61 percentage points over this period, the contribution of iodization corresponds
roughly to 7% of the total change in graduation rates.
When we limit the sample to women born in 1910-1935, the effect on secondary education falls
by more than half, which is puzzling. This drop in the coefficient might be related to the actual
penetration of iodized salt in high-goiter districts, which occurred later in time, and therefore affected
younger cohorts, who were born after 1935. This possibility is explored in section 1.8.

1.8

A second look at the data:
The penetration of iodized salt in each canton

1.8.1

A closer look into cantonal iodized salt sales

As manifested by the decrease in deaf-mutism rates described in section 1.6, the campaign for
iodization had strong and immediate effects on the population. The graphs and regression results
in section 1.7 also suggest that high-goiter regions were exposed to some kind of treatment which
affected the education of cohorts born around the time of iodization.
However, a closer look into the data paints a more complex picture. It seems that high-goiter
cantons introduced iodized salt later than the rest of the country. Table 1A-1 in the Appendix
contains annual, canton-level data of iodized salt sales as a percentage of total salt sales for the
period 1922-194922. The data in Table 1A-1 come from a paper published in 1962 by H. J. Wespi23
(Wespi 1962). Table 1A-2 shows population-weighted goiter rates using Bircher’s data, aggregated
at the canton level (Bircher 1883). A quick look at Tables 1A-1 and 1A-2 shows that high-goiter
cantons tended to introduce iodized salt later than the rest of the country.
Table 1A-1 shows how widespread the use of the “new salt” was at each point in time. It
shows that for most cantons the transition to iodized salt was fast, whereas for others it came more
22 Jura

was not yet a canton at that time, so it is not part of Table 1A-1. What later became the canton of Jura
was still part of the canton of Bern at that time.

23 M.D.

and Chief Doctor of Women’s Clinic in Aarau. Incidentally, H.J. Wespi was the son-in-law of H. Eggenberger,
the doctor who first introduced iodized salt in Appenzell-Ausserrhoden in 1922. I am grateful to Prof. Dr. Hans
Bürgi for providing me with H. J. Wespi’s paper, which contained these data.

21

.4

.5

Secondary Education
.6
.7
.8

.9

Figure 1.6: Secondary education, difference between high-and low goiter districts

1910

1920

1930
Year of birth

1950

Secondary − Low goiter

.22

Secondary − High − Low goiter
.225
.23
.235
.24

.245

Secondary − High goiter

1940

1910

1920

1930
Year of birth

1940

source: Bircher (1883) and 1970 Swiss Census

1950

22

.05

.1

Tertiary Education
.15
.2

.25

Figure 1.7: Tertiary education, difference between high-and low goiter districts

1910

1920

1930
Year of birth

1950

Tertiary − Low goiter

.005

Tertiary − High − Low goiter
.01
.015
.02
.025

.03

Tertiary − High goiter

1940

1910

1920

1930
Year of birth

1940

source: Bircher (1883) and 1970 Swiss Census

1950

23

.07

High − Low goiter, males, secondary
.08
.09
.1

.11

Figure 1.8: Secondary education, males and females, difference between high-and low goiter districts

1920

1930
Year of birth

1940

1950

1910

1920

1930
Year of birth

1940

1950

.24

High − Low goiter, females, secondary
.25
.26
.27

.28

1910

source: Bircher (1883) and 1970 Swiss Census

24

−.05

High − Low goiter, males, tertiary
−.04
−.03
−.02

−.01

Figure 1.9: Tertiary education, males and females, difference between high-and low goiter districts

1920

1930
Year of birth

1940

1950

1910

1920

1930
Year of birth

1940

1950

−.05

High − Low goiter, females, tertiary
−.04
−.03
−.02

−.01

1910

source: Bircher (1883) and 1970 Swiss Census

25
gradually. Examples of a fast transition are Schwyz, Luzern and Genève, whereas the transition was
more gradual in Thurgau and Graubünden. The data also indicate that the timing of adoption of
iodized salt differed across cantons; some cantons, such as Nidwalden and Schaffhausen, iodized early,
whereas others, such as Aargau, Basel-Stadt and Basel-Land, were much slower in their adoption.
Figure 1.1 shows that the adoption of the “new salt” for country as a whole was gradual.

.6
.4
.2
0

Iodized salt sales as % of total salt sales

.8

Figure 1.10: Iodized salt penetration in high- and low-goiter districts

1910

1920

1930
Year

Total for Switzerland
Low goiter areas

1940

1950

High goiter areas

source: Bircher (1883), Wespi (1962) and 1970 Swiss Census

Figure 1.10 confirms that high-goiter districts belonged to a large extent to cantons which iodized
relatively late. Figure 1.10 is similar to Figure 1.1 in that it shows population-weighted averages
of iodized salt sales, but apart from showing the countrywide average, Figure 1.10 also breaks the
sample down into high-goiter districts and the rest of the country. High-goiter districts, which seem
to have benefited from iodization immediately after 1922, actually lagged behind in the introduction
of iodized salt; there is one big increase around 1936, and another one after 1942. The sudden
increase in iodized salt availability in high-goiter districts in 1936 is driven by the jump in iodized
salt sales in the canton of Bern, which went from 11% in 1935 to 54% in 1936. Bern, as shown in
Table 1A-2, was among the most populous and worst-afflicted cantons, so a jump in iodized salt
sales there would have a significant impact for the country-wide average. Similarly, the second big
increase shown in Figure 1.10 and taking place after 1940 corresponds to jumps in iodized salt sales
in the cantons of Luzern and Fribourg, which were also populous and had highly-afflicted districts.
It is unclear why high-goiter cantons lagged disproportionately in the introduction of iodized

26
salt in their markets. One possible explanation is that iodized salt and iodine supplementation in
general was regarded with some suspicion in the initial stages of their introduction, because they
were linked to a spike in thyroid-related deaths. Indeed, a sudden, uncontrolled increase in iodine
intake in individuals who have been chronically exposed to iodine deficiency can cause the thyroid
gland to produce excessive amounts of thyroid hormone, leading to thyrotoxicosis, which might be
fatal. Feyrer, Politi and Weil (2008) have documented a significant spike of thyroid-related deaths
in the USA following the introduction of iodized salt in 1924. These deaths, as expected, were
concentrated on older age groups, which were exposed to iodine deficiency over a longer time.
Taking one more look at the graphs in section 1.7, there is some subtle indication of further
departures from trend for high-goiter districts in the late 1930’s and early 1940’s. For example, in
Figures 1.7 and 1.9, the differences in regression-adjusted probabilities of graduating from tertiary
education between high-goiter districts and the rest of the country seem to increase after 1935. At
any rate, iodized salt penetration for each canton deserves closer attention, and incorporating this
extra piece of information into the analysis the next logical step.

1.8.2

An iodine response function

The fact that iodization seems to have had immediate effects in high-goiter areas goes contrary to
the fact that these areas were, in fact, those which iodized in later years. To reconciliate these two
views, it is possible to imagine that the outcomes of treatment are non-linear; in particular, it might
be easier to cure acute conditions and correct sharp outcomes such as cretinism and deaf-mutism
with low levels of iodine supplementation, whereas more subtle outcomes, such as school graduation
rates, might require a more complete and universal degree of treatment. High-goiter cantons were
exposed to a first shock in 1922, which had to do with increased information and awareness of
the prophylactic uses of iodine. Complete, universal treatment came later, when iodized salt was
introduced in these areas.
Such a model is graphically depicted in Figure 1.11. The vertical axis corresponds to the level of
iodine deficiency in a locality, and the horizontal axis to the degree of iodine treatment. Non-goitrous
areas are located on the upper right part of the graph, and increased iodine availability will not
increase average IQ in their population. A high-goiter area with endemic goiter and cretinism, on the
other hand, would be on the lower left part, and would benefit from increased iodine intake. As iodine
supplementation becomes widespread, the high-goiter area will become less and less deficient, and
once treatment is complete, it will reach the upper right part of the graph. The adjustment, though,
doesn’t have to be linear in terms of the impact of iodization on cognitive outcomes. According to
this model, it will be relatively easy to cure cretinism and deaf-mutism with low doses of iodine,
but it will be harder and it will take higher levels of treatment to restore the population’s cognitive
ability to normal levels. Education levels will be mostly affected once treatment brings an area to
the flatter part of the curve.
High-goiter districts may have had a delay in the provision of iodized salt, but they were most
likely subject to some degree of treatment in 1922. Iodized salt is by far the easiest, but it is not the

27

Figure 1.11: Iodine response function

only way to provide iodine to a population. As medical knowledge and awareness of the prophylactic
use of iodine in Switzerland grew, it is possible that partial treatment took place after 1922, even in
areas where iodized salt per se was not available. For example, doctors might have prescribed iodine
supplements to pregnant women who developed goiters. This is consistent with finding a treatment
effect just by looking at post-1922 outcomes, especially for such acute conditions as cretinism and
deaf-mutism. According to the model, though, a higher level of treatment, such as widespread use
of iodized salt, will affect more subtle outcomes, such as school graduation rates. The next step,
therefore, is to incorporate the data on iodized salt sales as part of the analysis24 .

1.8.3

Using iodized salt sales to identify the effect of iodization

In a second specification, I introduce iodized salt sales at the canton level directly as a regressor
in a linear probability model of graduating from secondary and tertiary education. In particular,
I control for the percentage of iodized salt sales in total salt sales one year prior to birth in one’s
canton of birth. I also control for canton and cohort fixed effects, to remove the effect of any omitted
variables that are canton- or cohort-specific. In addition, I introduce a canton time trend, to control
for any other gradual changes that might have affected one’s schooling outcomes. In short, I run
the following regression for an individual i born in canton c in year y:
outcomeidy = α + β· Iodized salt 1 year prior to birth
24 It

is possible that within a geographic location, selection into treatment during the first phase of iodization was
non-random. For example, access to information about the benefits of iodine supplementation for pregnant women
may have been correlated with one’s socioeconomic status. In this case, the break in trend for high-goiter areas
which I’m observing in 1922 comes from a limited part of the population distribution. Non-random selection into
treatment is less of an issue in later stages, when iodized salt was widely used.

28

Table 1.4: Salt sales and education: Coefficient on Iodized Salt 1 year prior to birth

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

ALL AREAS
-0.1561
(0.37471)
3078907
0.0708
0.49531**
(0.23577)
3078907
0.0263
ALL AREAS
-0.04116
(0.36863)
1499495
0.0654
0.39761
(0.33792)
1499495
0.0265
ALL AREAS
-0.53394
(0.46747)
1579412
0.0851
0.56445**
(0.24226)
1579412
0.0307

WHOLE SAMPLE
HIGH-GOITER
LOW-GOITER
1.51967***
-0.34064
(0.52923)
(0.52798)
768353
756364
0.0602
0.0745
1.91756***
0.86382**
(0.48028)
(0.40792)
768353
756364
0.0259
0.0364
MALES ONLY
HIGH-GOITER
LOW-GOITER
0.91858
0.21011
(0.66006)
(0.56328)
375772
367747
0.0583
0.0653
1.811**
1.65157***
(0.72043)
(0.52504)
375772
367747
0.0242
0.0379
FEMALES ONLY
HIGH-GOITER
LOW-GOITER
1.94846***
-1.04845
(0.68709)
(0.68117)
392581
388617
0.0688
0.097
2.04099***
0.05975
(0.46669)
(0.42236)
392581
388617
0.033
0.0385

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include canton of birth and cohort fixed effects, as well as a canton-specific trend.
Standard errors in parentheses, clustered at the canton-year of birth level.

+ Canton of birth Fixed Effects
+ Canton of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where outcomeidy is an indicator variable for having graduated from secondary/tertiary education,
and standard errors are clustered at the canton-cohort level.
Results from this regression are shown on Table 1.4. As shown in this table, iodized salt doesn’t
seem to have had much of an effect for the county as a whole, but breaking the sample into highand low- goiter districts tells a different story. High-goiter districts are those corresponding to
the top 25% of the population-weighted goiter distribution, whereas low-goiter districts are those

29
corresponding to the bottom 25% of the same distribution. The probability of graduating from
secondary and tertiary education if one was born in a high-goiter district after universal use of
iodized salt increases by 1.5 and almost 2 percentage points, respectively. Again, this increase is
largely driven by females. According to Table 1.4, the increase in the probability of graduating
from tertiary-level education is higher than for secondary education. This is consistent with the
notion that affecting secondary-level graduation rates might have been easier in the first stage of
the intervention, which corresponds to increased awareness and the “informational shock” which
occurred in 1922. However, once iodine prophylaxis became universal and complete with the advent
of iodized salt, it affected “harder-to-achieve” outcomes, such as graduation from tertiary education.

1.9

Fuzzy regression discontinuity and jumps in sales of iodized
salt

A closer look into the iodized salt sales data reveals that the transition to iodized salt happened
pretty rapidly for most cantons. For example, Luzern went from 5% to 54% to 100% of total salt
sales being iodized in a period of only two years, from 1942 to 1944. I take advantage of such sudden
jumps in iodized salt sales, as they correspond to jumps in the probability of being treated. Since
these jumps all occurred in a small window of time, this method is “cleaner” than a simple regression
of outcomes on iodized salt sales, as there are fewer possible confounding effects taking place within
same small time frame in a particular canton.
Based on the year of the jump in iodized salt sales, which is particular to each canton, I construct
a new variable, “age relative to iodization”, which is common across all people born in the same
canton in the same year, but will generally be different across cohorts born in different cantons.
Figure 1.12 shows the population-weighted average percentage of iodized salt sales one year prior
to birth, plotted against age relative to iodization. At relative age 0, iodized salt sales jump from
around 22% to around 66%.
I check for some preliminary evidence of how being born in a high-goiter district right after a jump
in iodized salt sales affects one’s educational achievement. Table 1.5 shows results of a regression of
educational outcomes of an indicator variable equal to 1 if someone was born in a high-goiter district
after a jump in iodized salt sales, controlling for district and cohort of birth fixed effects, as well as
a district-specific time trend. I run the following regression for an individual i born in district d in
year y:
outcomeidy = α + β· 1 (Born in high-goiter district) X 1 (Born after jump in sales of iodized salt)
+ District of birth Fixed Effects
+ District of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where outcomes are the probability of graduating from secondary and tertiary-level education, and
where standard errors are clustered at the district-cohort level. First, I run the regression on the
whole sample, and then for males and females separately. Also, I first use the whole sample and

30

0

Average iodized salt sales one year prior to birth
20
40
60
80
100

Figure 1.12: Iodized salt sales by age relative to iodization

−20

−10

0
Age relative to iodization

10

20

source: Wespi (1962), and 1970 Swiss Census

then only the sub-sample of those people born within ten years of the jump in iodized salt sales. In
columns (1) and (3) of Table 1.5, I use cohort fixed effects that are common for high-goiter districts
and the rest of the country, whereas in columns (2) and (4) I allow cohort fixed effects to be different
for high-goiter districts.
According to Table 1.5, the probability of graduating from secondary education grew by between
0.8 and 1.2 percentage points for people born in high-goiter districts after a jump in iodized salt sales,
whereas the same number for tertiary education ranges from 0.3 to 1.3 percentage points. When
I use all cohorts in the sample, some statistical significance is lost when I don’t constrain cohort
fixed effects to be the same across high-goiter districts and the rest of the country. Results seem,
again, to be driven by females, although the difference between male and female coefficients is not
always statistically significant. When I limit the analysis to individuals born within a decade from
iodization, results are no longer driven mainly by females. Also, there’s a much smaller difference
between the two specifications, and, if anything, coefficients are higher for the stricter specification
of column (4) (although not statistically distinguishable from the coefficients in column (3)), where I
allow cohort fixed effects to vary across the two types of districts (high-goiter and non-high-goiter).
According to results in column (4), people born in high-goiter districts after a jump in iodized
salt sales had a 1 percentage point higher probability of graduating from tertiary education. It is
noteworthy that the magnitude of this coefficient is almost double of that suggested by the earlier,

31

Table 1.5: Coefficient on high-goiter dummy X Born after jump in iodized salt sales

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education

Tertiary Education

WHOLE SAMPLE
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
0.94485***
0.27477
0.81078**
1.24831***
(0.29962)
(0.37762)
(0.35746)
(0.37397)
3086287
3086287
1080394
1080394
0.0923
0.0924
0.0746
0.0747
1.28749*** 1.04325*** 0.74916**
1.03269***
(0.21145)
(0.25326)
(0.29383)
(0.313990)
3086287
3086287
1080394
1080394
0.0305
0.0305
0.0193
0.0193
MALES ONLY
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
0.44635
0.05057
0.97789**
0.96029*
(0.33153)
(0.44122)
(0.48179)
(0.50658)
1503257
1503257
544963
544963
0.0862
0.0862
0.0719
0.072
0.46311
0.73809*
0.75222
1.27428***
(0.31468)
(0.38552)
(0.45916)
(0.49252)
1503257
1503257
544963
544963
0.0319
0.0319
0.0233
0.0233
FEMALES ONLY
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
1.11057***
0.41875
0.77607
1.73938***
(0.40643)
(0.52814)
(0.53812)
(0.56875)
1583030
1583030
535431
535431
0.1145
0.1146
0.0958
0.0959
2.00584*** 1.37774***
0.8172**
0.9316**
(0.26661)
(0.33157)
(0.41626)
(0.43615)
1583030
1583030
535431
535431
0.0346
0.0346
0.0226
0.0226

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
Columns (1) and (3) assume common cohort fixed effects for high- and non-high-goiter districts.
Columns (2) and (4) assume different cohort fixed effects for high- and non-high-goiter districts.
All regressions include canton of birth and cohort fixed effects, as well as a canton-specific trend.
Standard errors in parentheses, clustered at the district-year of birth level.

32

Secondary Education, (high−low) goiter
.02
.04
.06
.08

Figure 1.13: Secondary and Tertiary education before and after jump in iodized salt sales, difference
between high-and low goiter districts

−10

−5

0
Age relative to iodization

5

.01

Tertiary Education, (high−low) goiter
.02
.03
.04
.05

secondary education, males, (high−low) goiter
secondary education, females, (high−low) goiter

−10

−5

0
Age relative to iodization

tertiary education, males, (high−low) goiter
tertiary education, females, (high−low) goiter
source: Bircher (1883), Wespi (1962), and 1970 Swiss Census

5

33
preliminary results of Table 1.3.
Figure 1.13 is a graphical representation of the results shown in Table 1.5. It is a regressionadjusted plot showing how the difference in probability of graduating from secondary (top panel)
and tertiary (bottom panel) education between those born in high-goiter districts and the rest of
the country changed according to one’s age relative to the jump in sales of iodized salt, for males
and females separately. The effect is barely there for males, but for females the difference clearly
increased for those born around the jump in iodized salt sales.
The probability of someone being treated with adequate iodine in utero increases discontinuously
if they were born after the jump in sales. This observation provides the framework for a fuzzy
regression discontinuity (FRD) design (for a description of FRD see, for example, Imbens and
Lemieux (2008)). Using jumps in iodized salt sales, I identify the effect of iodization on educational
outcomes. My identification relies on the fact that not all cantons iodized at the same time. Because
of the difference in the timing of intervention, I can control for unobserved, time-invariant canton
characteristics, as well as unobserved, space-invariant cohort characteristics. I regress indicator
variables for having graduated from secondary and tertiary education on the percentage of iodized
salt sales (at the canton-level) one year prior to birth (and additional controls), but I instrument
iodized salt sales with an indicator variable equal to 1 if someone is born after the big jump in sales
which marked each canton’s decisive transition to iodized salt. I run the following regression for an
individual i born in canton c in year y:
outcomeidy = α + β· Iodized salt 1 year prior to birth
+ Trend before jump in iodized salt sales
+ Trend after jump in iodized salt sales
+ Canton of birth Fixed Effects
+ Canton of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where outcomeidy is an indicator variable for having graduated from secondary/tertiary education,
and standard errors are clustered at the canton-cohort level. I include both a linear canton-specific
time trend, but also linear nationwide trends which are allowed to differ before and after the jump
in iodized salt sales.
Estimation results are shown in Table 1.6. First, I estimate the effects for the whole sample and
all areas, and then I use only those observations corresponding to the top and bottom 25% of the
population-weighted goiter distribution, and estimate the effects of iodization separately for each of
these two parts of the distribution. Then I estimate effects for males and females separately, using
the same cuts of the data.
Results using the fuzzy regression discontinuity design suggest even stronger effects for highgoiter areas than previous specifications, and they show the contrast between the effect estimated
for high-and low-goiter districts. For those born in high-goiter districts, there is a 2.1 percentage
point increase in the probability of graduating from secondary education, and a 1.4 percentage point
increase for tertiary education. Females benefited more in the case of secondary education, where

34

Table 1.6: Fuzzy Regression Discontinuity: Coefficient on Iodized Salt 1 year prior to birth

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

Secondary Education
Number of Observations
R-squared
Tertiary Education
Number of Observations
R-squared

ALL AREAS
-1.51933***
(0.42756)
3078907
0.0708
0.01346
(0.27196)
3078907
0.0264
ALL AREAS
-1.12351***
(0.41737)
1499195
0.0655
-0.45101
(0.37484)
1499195
0.0265
ALL AREAS
-1.95353***
(0.56893)
1579412
0.0852
0.46896
(0.29641)
1579412
0.0307

WHOLE SAMPLE
HIGH-GOITER
LOW-GOITER
2.11722***
-1.03942**
(0.5057)
(0.51821)
768353
756364
0.0603
0.0745
1.43328***
0.28196
(0.52231)
(0.41398)
768353
756364
0.026
0.0364
MALES ONLY
HIGH-GOITER
LOW-GOITER
1.26888*
-0.12438
(0.74509)
(0.60595)
375772
367747
0.0584
0.0653
1.66566**
0.86472
(0.7503)
(0.54564)
375772
367747
0.0246
0.038
FEMALES ONLY
HIGH-GOITER
LOW-GOITER
3.48568***
-2.14152***
(0.83893)
(0.68567)
392581
388617
0.0689
0.097
1.55792**
-0.2541
(0.63083)
(0.45283)
392581
388617
0.0331
0.0385

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include canton of birth and cohort fixed effects, as well as a canton-specific trend.
Standard errors in parentheses, clustered at the canton-year of birth level.

35
their probability of graduation increased by 3.5 percentage points.
Interestingly, the sample as a whole experienced a decline in the probability of graduation from
secondary education, which is clearly driven by people born in districts belonging to the middle and
lower parts of the goiter distribution. The magnitude of this decline is smaller than the magnitude
of the increase in graduation rates in high-goiter regions, and it could be explained by the presence
of short-run constraints in the supply of secondary education. Secondary education was much less
widespread in the 1920’s than it is today, and only a little over 50% of the population in my sample
graduated from it, so supply constraints might have been important. If capacity constraints are
behind the negative effect of iodization for the middle and lower parts of the goiter distribution,
however, they don’t seem to matter much in the case of tertiary education. Iodization had a
significant impact on the probability of graduating from tertiary education for high-goiter districts,
but it didn’t affect tertiary education in other regions.

1.10

Interpretation of coefficients

There is strong evidence that iodine provision in previously deficient regions had a significant impact
on the educational attainment of cohorts exposed to it. Looking across the specifications outlined in
previous sections, the probability of graduating from secondary education increased by around 1.1
percentage points, whereas the same number for tertiary education is around 1 percentage point.
Given that secondary school graduation rates increased by about 15 percentage points for those born
after 1922 compared to those born before 1922 in high-goiter districts, iodization alone contributed
about 7.3% of that change. If tertiary education increased by 1 percentage point, then, given that
the change in graduation rates between those born before and after 1922 in high-goiter districts was
8.5 percentage points, iodization is responsible for about 12% of the total change in graduation rates
from tertiary education in high-goiter districts.
Results are very strong for females. Female graduation rates increased by around 1.5 percentage
points for secondary education, and by about 1.2 percentage points for tertiary education, thus
accounting for 9% and 14% of the total change in female graduation rates in high-goiter regions
respectively. Also, the effect of iodization is often statistically higher for females than for males.
This result is consistent with two hypotheses: first, iodine deficiency might affect females more than
males, so correcting it will have larger effects for the female population. Second, schooling outcomes
across the two genders might be affected differently by a given increase in cognitive ability.
The first hypothesis is consistent with some evidence showing women to be more susceptible to
thyroid-related disorders. According to the American Society for Clinical Endocrinologists, “more
than 8 out of 10 patients with thyroid disease are women”, and “women are five to eight times
more likely than men to suffer from hypothyroidism”25. Feyrer et al. (2008) note that there was a
large gender disparity in thyroid-related deaths following salt iodization in the USA. In particular,
thyroid-related death rates were almost seven times higher for women than for men in 1926 (1.1
25 Source:

http://www.aace.com/public/awareness/tam/2005/pdfs/thyroid disease fact sheet.pdf

36
per 100,000 for men, versus 7 per 100,000 for women). My findings are consistent with cross-gender
differences in the effects of iodine deficiency on brain development in utero. A correction of iodine
deficiency in utero would, under such circumstances, affect females more than males, and therefore
have a bigger effect on the schooling outcomes of females once the exposed cohorts reach schooling
age.
However, there is a competing hypothesis which is consistent with my findings. A household’s
response to a given increase in the cognitive ability of its offspring could be different across genders.
In particular, household decisions regarding the schooling levels of its children could be more sensitive
to cognition for girls than for boys. In such a case, even though male and female fetuses are equally
affected by iodine deficiency in utero, a given increase in cognitive ability will produce different
effects on schooling outcomes across genders.
The effect on educational outcomes identified in the previous sections of this paper arises from
increased health capital in utero and early life. Health investments at such an early stage have been
shown to have a significant impact later in one’s lifetime, but I cannot distinguish whether this effect
comes from increased innate cognitive ability alone, better treatment at home and school (before even
reaching upper-secondary and tertiary levels of education), or a combination of the two. It is possible,
for example, that a smarter cohort received different parental and teacher attention in childhood and
early adolescence, in which case the initial effect of iodization was magnified by responses coming
from one’s social environment, which increased their potential for higher educational attainment
even more than the initial improvement in cognitive ability.

1.11

Concluding remarks

The treatment of iodine deficiency in Switzerland provides us with a rare opportunity to examine
how an exogenous increase in the cognitive ability of a population translates into higher schooling.
It also shows that there can be important differences across genders on how large-scale public
health interventions of this kind translate into economically significant outcomes. Although I cannot
distinguish whether this is due to physiological gender differences or to differential environmental
responses for males and females from the part of the household and general social context in which
one grows up, the fact that female education was more affected than male education is something to
keep in mind when thinking about current efforts to eradicate iodine deficiency, and thereby increase
schooling and productivity in much of the developing world.
The potential non-linearity of treatment is also noteworthy. While low levels of treatment might
cure acute conditions, such as cretinism and deaf-mutism, it takes a well-orchestrated public health
campaign to affect more subtle outcomes, such as schooling and graduation levels. This is important
when evaluating public health campaigns; the outcomes one chooses to look at might be important
when deciding on the success of an intervention.
A cost-benefit analysis is beside the point here; iodized salt was clearly cost-effective, even if one
only considers the reduction in medical costs related to goiters and other iodine-related disorders,

37
let alone factoring in the benefits of increased cognition and productivity. Prof. Dr. Hans Bürgi
estimates that, had it not been for iodine prophylaxis, the cost of homes providing for cretins and the
medical bills related to iodine deficiency disorders would have been around 270 million 2005 Swiss
Francs. The cost of adding iodine to salt, on the other hand, amounted to a total of 1.4 million
Swiss Francs in 200526.
The late 1990s saw significant improvements in iodized salt availability in the world, and UNICEF
estimates that 70% of people in the world consumed iodized salt in 2000, whereas less than 20%
did so back in 1990. However, iodine deficiency remains the single most easily preventable cause
of mental retardation today, and 38 million children are born annually at the risk of developing
iodine deficiency disorders (UNICEF 2008). The historical experience of Switzerland can serve as a
paradigm for countries fighting with iodine deficiency.

26 Personal

communication to the author.

38

1.A

Appendix
Table 1A-1: Annual iodized salt sales as a percentage of total salt sales
Canton Name

Zürich
Bern
Luzern
Uri
Schwyz
Nidwalden
Obwalden
Glarus
Zug
Fribourg
Solothurn
Basel-Stadt
Basel-Land
Schaffhausen
Ap. Ausserrhoden
Ap. Innerrhoden
St. Gallen
Graubünden
Aargau
Thurgau
Ticino
Vaud
Valais
Neuchâtel
Genève
Switzerland

22

43

23

24

25

26

Year
27
28

18
1
5

21
1
3

18
4
4

18
4
6

18
4
6

17
4
6

47
7
4
23

1
100
8
83
26

1
5
2
4
55
34
12
3
4
27

2
10
5
3
75
50
24
6
9
36

1
100
8
37
81
2
2
12
5
11
75
50
27
9
11
39

100
100
50
27
97
2
2
12
11
100
67
48
25
9
11
35

100
100
100
37
88
2
3
13
12
100
67
46
26
13
12
34

100
100
100
33
100
3
3
14
9
100
67
53
27
16
12
35

25

100
33
15

100
63
70

100
65
70

8

16

22

26

Canton Name
36
37
Zürich
52
53
Bern
54
64
Luzern
8
9
Uri
79
90
Schwyz
100
100
Nidwalden
100
100
Obwalden
100
100
Glarus
73
76
Zug
100
100
Fribourg
?
3
Solothurn
4
4
Basel-Stadt
14
13
Basel-Land
16
15
Schaffhausen
100
100
Ap. Ausserrhoden
71
71
Ap. Innerrhoden
64
64
St. Gallen
69
68
Graubünden
24
24
Aargau
11
11
Thurgau
38
39
Ticino
100
100
Vaud
100
100
Valais
100
100
Neuchâtel
70
70
Genève
90
90
Switzerland
51
54
Note: The canton of Jura is not in
Source: Wespi (1962)

38
39
40
52
53
55
65
66
63
7
8
8
90
88
90
100
100
100
100
100
100
100
100
100
76
81
82
100
100
100
3
4
7
54
74
69
15
23
25
15
18
17
100
100
100
71
71
68
60
62
57
68
68
67
26
43
75
10
10
8
39
39
36
100
98
97
100
100
100
100
100
100
70
70
70
89
88
90
55
58
59
this table, since it

29

30

31

32

33

34

35

15
5
7
100
100
100
100
41
100
2
3
15
10
100
73
54
47
18
10
36
100
100
80
70
1
30

13
6
7
100
100
100
100
60
100
2
3
14
34
100
74
49
52
17
11
32
100
100
87
70
2
34

14
6
8
97
100
100
100
66
100
1
3
13
15
100
70
51
51
20
13
34
98
100
95
70
3
33

53
6
6
97
100
100
100
67
100
3
3
14
14
100
77
51
58
22
12
37
100
100
96
70
4
39

51
7
8
97
100
100
100
68
100
3
3
14
14
96
67
53
55
21
9
35
100
100
100
70
3
38

53
8
8
93
100
100
100
70
100
?
3
10
12
100
68
39
54
20
10
37
100
100
100
70
27
40

54
11
8
88
100
100
100
72
100
?
3
10
28
100
69
59
64
21
9
35
100
100
100
70
66
40

45
70
73
81
100
100
100
100
94
100
76
67
28
19
100
77
49
89
93
7
46
100
100
100
70
88
73

46
70
69
92
100
100
100
100
93
100
76
64
29
18
100
79
86
91
93
7
67
100
100
100
70
88
75

47
70
74
97
100
100
100
100
94
100
100
63
27
20
97
87
100
91
96
8
76
100
100
100
67
91
77

48
77
73
100
100
100
100
100
95
100
100
60
27
21
96
89
100
93
94
7
84
100
100
100
70
92
78

49
77
75
100
100
100
100
100
97
100
100
58
27
21
100
92
100
93
95
7
88
100
100
100
68
93
79

100
100
75
78
70
70
1
1
29
27
Year
41
42
43
44
55
48
63
67
73
69
71
71
6
5
54
100
87
88
100
100
100
100
100
100
100
100
100
100
100
100
100
100
87
92
94
93
100
100
100
100
39
31
36
50
58
62
65
66
23
25
28
28
17
18
18
18
97
95
100
100
71
70
74
74
59
56
44
48
59
67
78
88
85
86
94
94
7
14
8
7
41
37
36
38
100
100
100
100
100
100
100
100
100
100
100
100
70
70
70
70
89
81
93
91
62
60
64
72
was only created in 1979.

39

Table 1A-2: Goiter rates and number of observations by canton
Canton of Birth
Zürich
Berne
Luzern
Uri
Schwyz
Obwalden
Nidwalden
Glarus
Zug
Fribourg
Solothurn
Basel-Stadt
Basel-Landschaft
Schaffhausen
Ap. Ausserrhoden
Ap. Innerrhoden
St. Gallen
Graubünden
Aargau
Thurgau
Ticino
Vaud
Valais
Neuchâtel
Genève
Jura
Total Switzerland

1970 Population-weighted
average goiter rate
8.69
14.38
13.32
1.97
4.59
4.24
4.87
4.87
9.71
20.16
7.38
7.45
5.36
3.03
5.38
3.37
5.77
3.61
13.91
2.43
1.15
1.96
5.68
5.21
1.50
2.96
8.49

Number of
observations
401656
540017
163472
23668
54857
17358
15925
26672
26395
130704
120707
95673
73803
38522
40949
13807
232667
99746
221113
103106
109378
216993
120191
80368
71884
46656
3086287

Sources: Bircher (1883) and Federal Statistical Office (1970)

Chapter 2

The effects of the generalized use
of iodized salt on occupational
patterns in Switzerland.
2.1

Introduction

Nutrition is inextricably linked to a population’s health capital. Malnutrition, especially when it
occurs early in life, can have serious detrimental effects on a person’s lifetime productivity and
economic prospects. Micronutrient deficiencies are a common source of malnutrition, caused by
insufficient intake of necessary vitamins and minerals. Iodine is one such micronutrient.
Lack of iodine causes many disorders, the most common of which is an enlargement of the
thyroid gland, when there is not enough iodine for the production of hormones which regulate
metabolism. This enlargement is called a goiter. Other symptoms include short stature and deafmutism. However, the most catastrophic consequence of iodine deficiency is brain damage, which
is irreversible and can go unnoticed in a population. Iodine deficiency results in various degrees
of mental underdevelopment when it occurs in utero and the first three months of life. Cretinism,
which is an acute form of mental retardation, often coupled with goiter and deaf-mutism, occurs
when iodine deficiency is utero is most severe.
Iodine deficiency is the leading cause of preventable mental retardation in the world today. The
WHO estimates that nearly 50 million people suffer some degree of mental impairment due to a lack
of iodine in their diets1 . According to WHO’s Global Database on Iodine Deficiency, more than
285 million children receive inadequate amounts of iodine in their diet 2 . Despite efforts to decrease
the prevalence of iodine deficiency in the 1990s, there are still 38 million children born annually at
1 Source:

WHO, http://www.who.int/features/qa/17/en/index.html.

2 Source:

de Benoist et al., eds (2004).

40

41
the risk of developing iodine deficiency disorders. The most vulnerable areas in the world are South
Asia and Central and Eastern Europe (UNICEF 2008).
Although iodine deficiency is eradicated in developed countries today, the picture looked quite
different in early 20th century. Many countries, for reasons related to their geography, had “pockets”
of endemic iodine deficiency within their boundaries. For example, the area around the Great Lakes
in the USA, as well as some Northwestern states had rates of iodine deficiency that were only
paralleled by the prevalence in the Swiss Alps. Indeed, Switzerland was the worst-afflicted country
in the world, because its soil has been stripped of its iodine content in many localities during the
last Ice Age.
This paper estimates the effects of iodine deficiency eradication on occupational patterns using
data from Switzerland. Switzerland was the first country in the world to introduce iodized salt
in 1922. It was the first major nutritional intervention ever recorded. Iodized salt proved a costeffective measure to eradicate endemic goiter, which is an enlargement of the thyroid gland, occurring
when there is not enough iodine in the body. The invisible effects of iodine deficiency on mental
development and cognitive ability were not fully understood at the time, and public health authorities
did not know that they were fighting against mental retardation as well as endemic goiter. As a result
of the countrywide iodization campaign, there were no more endemic cretins born after 1930, deafmutism rates dropped significantly, and goiter disappeared in children and young recruits (Bürgi
et al. 1990). Salt iodization also had a significant impact on graduation rates of those born in
highly-deficient areas (Politi 2009). In this paper I find that occupational patterns in previouslydeficient areas changed too, reflecting a shift towards higher-paying occupations, and also towards
occupations with a lower physical- and a higher cognitive-skills component.
I use microdata from the comprehensive 1970 Swiss Census, combined with data on occupational
characteristics. I identify the effect of iodization on occupational patterns by exploiting variation
in pre-existing prevalence of iodine deficiency, and also differences in the timing of adoption of
iodized salt across localities. I also use a fuzzy regression discontinuity design, where I identify the
effect of iodization by looking at sudden jumps in iodized salt sales, and then compare occupational
outcomes for cohorts born right before and right after the jump. Overall, I find that iodine deficiency
eradication pushed cohorts affected by it towards higher-paying occupations, and occupations with
higher cognitive demands. My estimates imply that between 37% and 45% of the shift towards
higher-paying occupations that took place in high-goiter areas over that period is accounted for by
the eradication of iodine deficiency.
The rest of the paper is organized as follows: Section 2.2 provides some background on iodine
deficiency disorders, and section 2.3 describes the campaign for salt iodization in Switzerland. Section 2.4 describes the data used to define the sources of variation for my identification. Section
2.5 describes the Swiss Census microdata, as well as the occupational variables used to identify the
effect of iodization on occupations. Sections 2.6 outlines my identification strategy, and empirical
results are presented on sections 2.7 and 2.8. Section 2.9 concludes.

42

2.2

Iodine Deficiency Disorders

Iodine, together with Iron and Vitamin A, is a necessary micronutrient, found in very small quantities
in the human body. Most of the body’s iodine is located in the thyroid gland. Iodine is essential in
the synthesis of the two thyroid hormones which regulate metabolism and “play a determining part
in early growth and development of most organs, especially of the brain” (Delange 2001).
When the thyroid does not receive sufficient amounts of iodine it adapts by enlarging in order to
maximize the use of available iodine. This enlargement is called a goiter, and it is one of the many
symptoms of iodine deficiency. Goiters can occur at any point in one’s lifetime, whenever the iodine
intake is not sufficient. Some goiters are reversible, especially in young individuals. Reversing goiter
in adults is harder, especially when they have been subject to iodine deficiency for many years.
Goiter is a visible effect of iodine deficiency. Apart from goiter, however, iodine deficiency can
have irreversible and harder to observe consequences if it occurs in utero and in the first three months
of life. Iodine deficiency in utero results in various degrees of mental retardation and abnormal brain
development, which might even go undetected in a population. In the worst case scenario, severe
iodine deficiency causes cretinism. Cretinism is an acute condition characterized by a combination
of mental retardation, stunting and physical deformation. Bleichrodt and Born (1994) estimate that
the average IQ of iodine-deficient groups is 13.5 points lower than the non-deficient groups3 . If this
is true, then iodine deficiency should have sizable economic effects for any afflicted population. It
is this “intelligence-enhancing” aspect of correcting iodine deficiency in a population that is at the
heart of this paper.
Endemic goiter and endemic cretinism are primarily due to the geographic location of a population. The main store of iodine is the ocean. As ocean water evaporates, iodine falls on the upper
layers of soil through rainfall. Therefore, geographic areas close to the ocean are naturally rich in
iodine. On the contrary, regions subject to heavy rain or intense glaciation in the past may be
iodine-poor due to soil erosion. It takes thousands of years for rain water to replenish the superficial
layers of soil with iodine, so the iodine content of the soil and water of those regions is extremely
low. Regions naturally poor in iodine include mountainous areas such as the Andes, the Alps, the
Pyrenees, and the Himalayas (Koutras et al. 1980).
A more detailed description of iodine deficiency disorders is provided in Politi (2009).
Iodine has been explicitly used in the treatment of goiter ever since Bernard Courtois isolated
it as an element in 18114. The idea that endemic goiter is due to iodine deficiency was first put
forward in 1846, by Jean-Louis Prévost and A.C. Maffoni (Prévost and Maffoni 1846). Even though
cretinism was associated with endemic goiter, the crucial role of iodine in mental development was
not understood until more than a century later. When large-scale interventions of iodine supplementation took place around the 1920s and after, the objective was goiter eradication. People did
not know that they were also fighting against mental retardation.
3 This

estimate is based on a meta-analysis of 21 studies of the effect of iodine deficiency on cognitive ability.

4 However,

iodine-rich foods and plants, such as seaweed, were used by ancient civilizations, such as the Chinese
and the Greeks, to treat the swelling of the neck before the isolation of iodine as an element.

43
After doctors started prescribing iodide to their patients in order to fight goiter, toxic side-effects
resulting from over-dosing triggered opposition to the universal use of iodine. The first larger-scale
iodine supplementation program took place from 1917 to 1922 in Akron, Ohio, and it involved
administering sodium iodide regularly to schoolgirls from 5th grade and above. When it began, the
intervention was very controversial, but its undeniable success in decreasing the young girls’ goiters
paved the way for larger-scale programs in the USA and in Europe.
Iodized salt started circulating in Switzerland in 1922. Almost simultaneously, fortification of
salt with iodine began in the USA, where iodized salt first appeared in 1924. Both interventions
eliminated endemic cretinism and goiter in children, and they decreased goiter prevalence in adults,
even though they were followed by an initial spike in goiter-related surgeries and deaths, which then
subsided5 . Information on the program in Switzerland is provided in the following section.

2.3

Swiss Iodization Campaign

Switzerland was known for its high prevalence of goiter and cretinism since ancient times (Roman
writers mentioned it in their works). During Napoleonic Wars, the low performance of Swiss recruits
for the French Army troubled Napoleon and the local authorities in today’s canton of Valais. Under
Napoleon’s orders, a survey was conducted, which showed an extremely high prevalence of cretinism
in the population (Bürgi et al. 1990). Further studies revealed that Switzerland had a much higher
rate of goiter and cretinism than any of its neighboring countries (Italy, France, Germany).
Swiss data on goiter prevalence confirmed the link between iodine deficiency and goiter prevalence: the canton of Ticino ranked the lowest among cantons in goiter prevalence. This is to be
expected, since Ticino is in the southernmost part of Switzerland, bordering Italy and enjoying a
milder climate, proximity to the Mediterranean Sea and possibly more iodine-rich foods coming from
Italy than the rest of the country. Another canton with unusually low goiter prevalence was Vaud.
Historically, Vaud had an exclusive salt mine, which was rich in iodine (Bürgi et al. 1990, p.581).
As a result of the studies showing the extent of the goiter problem, a Swiss Committee for the
study of goiter was established in 1907. At that time, goiter was still attributed to some agent in the
drinking water, even though experiments with iodine supplementation for the treatment of goiter
were already taking place in France and, later, in the USA. Right before his death in 1917, Kocher
suggested goiter treatment with small doses of iodine (Bürgi et al. 1990).
The first canton to iodize salt was Appenzell-Ausserrhoden, where iodization started in February 1922, with the initiative of a local doctor, H. Eggenberger. In June 1922, the Swiss Goitre
Commission recommended the addition of small amounts of iodine in salt and the additional weekly
consumption of iodine tablets by schoolchildren. In November 1922, the Swiss salt monopoly [United
5 This

adverse consequence of iodine supplementation was due to the existence of nodular goiters in the population.
Nodular goiters were caused by chronic iodine deficiency. Nodular goiters may become toxic following a sudden
increase in iodine intake after a long period of deprivation. This side-effect of iodization is known as iodine-induced
hyperthyroidism).

44
Swiss Rhine Salt Works (USRSW)6 ] started adding iodine to salt and selling the new product at
the same price as non-iodized salt. Even before that date, though, iodine prophylaxis had become
popular by means of tablets or other supplements. After the recommendations of the Swiss Goiter
Committee and the success of salt iodization in Appenzell-Ausserrhoden, the other cantons started
allowing the sale of iodized salt in their markets.
Not all cantons introduced iodized salt simultaneously, though. For instance, Valais iodized in
1925, Zürich in 1932 and Bern in 19367. On the other hand, Aargau and Basel-Land didn’t iodize
until 1952 and 1950 respectively. In 1925 fewer than one fourth of cantons had iodized salt sales
that exceeded 60% of total salt sales. By 1945 fewer than one fourth of cantons had salt sales that
were below 20% of total salt sales. By 1955, iodized salt sales exceeded 60% of total salt sales in all
cantons, and in many of them only iodized salt was sold and consumed (Wespi 1962).
The success of the iodization program was indisputable. According to Bürgi et al. (1990), “no
new endemic cretins born after 1930 have been identified” (p.577). Deaf-mutism rates fell sharply for
cohorts born after 1922. In Appenzell-Ausserrhoden, which was the first canton to provide iodized
salt to its inhabitants, the prevalence of goiter in newborns fell from 20% to 6.4% within the first
year after iodization. The prevalence dropped further when, in later years, the iodine content of salt
was raised. The beneficial effects on iodization were also seen in the increased height of 6-year-olds
entering school, as well as young recruits. In the city of Lausanne, 23.7% of young recruits had large
goiters in 1924/1925, but the figure had dropped to 0.2% by 1983-1987 (Bürgi et al. 1990).
More information on the Swiss Iodization Campaign is provided in Politi (2009).

2.4

Bircher’s monograph and Iodized Salt Sales

To identify the effect of iodization on occupational outcomes, I employ two sources of variation: the
first in the naturally-occurring geographical variation in underlying iodine deficiency prior to the
generalized use of iodized salt. The second source of variation arises because of differences in the
timing of adoption of iodized salt across Swiss localities.
In 1883, Swiss physician Heinrich Bircher published a monograph with details on the geographic
variation in goiter rates across Switzerland (Bircher 1883). Over the period 1875-1880, he toured
every town and village in Switzerland and recorded goiter cases in recruits, which served as a proxy
for goiter prevalence in the local population. Bircher’s monograph was eye-opening to public health
authorities at the time, because it showed the extent of the problem across the country, and also the
big differences in goiter prevalence, even in villages within a short distance from each other. The
data correlates well with measurements of iodine content of the water and soil across Swiss localities.
I use the data collected by Bircher to group Swiss districts according to their goiter prevalence.
Goiter prevalence serves as a proxy for underlying iodine deficiency in the population. I classify a
6 USRSW

was “the exclusive supplier of salt to 24 of the 25 cantons” of Switzerland, the exception being the canton
of Vaud (Bürgi et al. 1990, p.582).

7 This

is the first year that the cantons’ iodized sales exceeded 40% of total salt sales.

45
district as being “high-goiter” if it belongs to the top 25% of the population-weighted goiter distribution. This corresponds to districts where goiter prevalence was 11.7% or higher. Correspondingly,
a district is “low-goiter” if its goiter prevalence is within the lowest 25% of the goiter distribution
in the population, corresponding to goiter rates lower than 3.5%. High-goiter districts are where I
expect to observe the treatment effect of iodization. On the contrary, I expect the treatment effect
to be much lower in low-goiter districts.
After iodized salt became available in 1922, it was not adopted at the same time or speed by all
cantons. Some cantons, such as Nidwalden and Schaffhausen, were early adopters, whereas other
cantons, such as Aargau, Basel-Stadt, and Basel-Land were much slower. The sale of iodized salt
had to be approved and allowed by each canton’s constitution. In 1962 H. J. West, M.D. and Chief
Doctor of Women’s Clinic in Aarau, published a paper containing data on yearly iodized salt sales
as a percentage of total salt sales per canton, from 1922 to 1961 (Wespi 1962). These data show
how widespread the use of the “new salt” was at any point in time.
Some cantons, such as Schwyz and Luzern made a very fast transition to iodized salt, whereas for
other cantons, such as Thurgau and Graubünden, the transition was much more gradual. It appears
that low-goiter cantons were the ones that adopted iodized salt earlier than highly-affected cantons.
This reluctance from the part of those most likely to benefit from the intervention is puzzling at
first. However, one needs to take into account that this was the first major nutritional intervention
to ever take place, and it was not without controversy, especially in light of many thyroid-related
deaths resulting from iodine overdoses to people who had been severely deficient throughout their
lives. It is reasonable to expect those cantons where the stakes were higher to be the ones where the
transition took the longer to finally occur, since public debate on the issue would have been more
heated.
I use the data contained in West’s paper as a measure of treatment. Also, based on West’s data
I construct a variable corresponding to one’s age relative to iodization in their canton of birth. In
particular, I identify jumps in the sales of iodized salt, and I look at individuals’ occupational choice
outcomes, comparing those born right before and right after the jump in sales. This is feasible
because once one takes a glance at the data, it becomes obvious that most cantons had a steep
transition to iodized salt, as opposed to a more gradual adoption.
For more information on Bircher’s monograph and on the geographic variation of goiter prevalence, as well as on the iodized salt sales data, and how iodized salt adoption coincides with a drop
in deaf-mutism rates, see Politi (2009).

2.5

The 1970 Swiss Census and Dictionary of Occupational
Titles

I identify the effect of iodization on occupational choice using microdata from the 1970 Swiss Census
(Federal Statistical Office 1970), which includes detailed information on a person’s year and location
of birth. Switzerland is a federation made up of 26 cantons (comparable to US States, but much

46

Table 2.1: Actively employed individuals in 1970 Swiss Census
Males
Females
Total

Active
1,149,692
475,708
1,625,400

Non-active
109,524
841,009
950,533

Total
1,259,216
1,316,717
2,575,933

(Source: 1970 Swiss Census)

smaller), 184 districts and 2896 municipalities. The 1970 Swiss Census records an individual’s
municipality of birth. This low level of aggregation is particularly important, because endemic
iodine deficiency was a very localized phenomenon, and it depended on the iodine content of a
population’s local sources of food and water. Municipalities within a short distance of each other
might have had very different exposure to iodine before iodized salt was introduced, so it is important
to know in as much detail as possible the location of one’s birth.
I use all individuals born in Switzerland from 1900 to 1944 (inclusive) whose occupation is
recorded in the Swiss Census. Unfortunately, this only includes people actively employed (both
part- and full-time) at the time of the Census. This excludes most women in my sample. Table 2.1
shows that out of 2,575,933 individuals born in the period 1900-1944, only 1,625,400 held a job in
1970, and only 36% of females were active in the labor force.
The 1970 Swiss Census contains detailed information on each individual’s occupation, using 4digit codes according to the International Standard Classification of Occupations (ISCO). Using these
data, and grouping individuals into 9 broad occupation categories (according to the first digit of their
ISCO code), I construct an indicator variable for whether an individual was employed in the top three
occupational categories. These categories include individuals employed in executive/managerial
positions, senior officials and legislators, professionals such as physicians, engineers and lawyers, as
well as technicians and associate professionals such as police inspectors, trade brokers, and health
associates. These occupational categories have earned higher wages historically, compared to the
other categories of occupations in the data. For example, in the second trimester of 2007, the
annualized median income of full-time workers in these top categories was over 84,000 Swiss Franks
or more, whereas the corresponding number for all other occupation categories was less than 65,000
Swiss Franks (Communication from the Swiss Federal Statistical Office). Occupational groups not
included in the three top categories include clerks, service and shop sales workers, skilled agricultural
workers, craft and related trades workers, machine operators, and people employed in elementary
occupations such as street vendors, cleaners, and unskilled laborers. Table 2.2 shows how many
individuals were employed in each of the nine broadly-defined occupational categories in the 1970
Swiss Census. A little over 25% of individuals (almost 29% of males) are employed in the three
higher-paying occupational categories.
I combine the occupational data of the 1970 Swiss Census with data on occupational characteristics compiled by Paula England and Barbara Kilbourne from the Dictionary of Occupational

47

Table 2.2: Broad occupational categories in 1970 Swiss Census

Highpaying
occupations

Occupational Category
1: Legislators, senior officials,
managers
2: Professionals
3: Technicians and Associate Professionals
4: Clerks
5: Service, shop and market sales workers
6: Skilled agricultural and fishery workers
7: Craft and related trades workers
8: Plant and machine operators
and assemblers
9: Elementary Occupations
Total

Males

Females

Total

85,317
78,644
168,871
157,625
47,697
124,209
311,402

20,317
18,934
55,286
117,620
104,859
41,832
53,799

105,634
97,578
224,157
275,245
152,556
166,041
365,201

122,068
53,859
1,149,692

19,133
43,928
475,708

141,201
97,787
1,625,400

(Source: 1970 Swiss Census)

Titles for 1980 US Census Detailed Occupations (England and Kilbourne 1988). The dataset is
available in electronic format and freely distributed by the Inter-university Consortium for Political and Social Research (ICPSR). These data contain scores on a variety of characteristics for all
occupational codes used in the 1980 US Census. For example, occupations get scores according to
verbal, numerical, spatial, and other aptitudes needed to perform the job, as well as the physical
demands of the job.
I match occupational codes from the 1980 US Census with ISCO codes, which are used in the
coding of occupations in the Swiss Census. In this way, I am able to get scores of many occupational
characteristics and merge them with the 1970 Swiss Census. As for many four-digit ISCO categories
there were no entries or a direct correspondence with US Occupational codes, I aggregate ISCO codes
to three-digit categories, and compute the average value of each characteristic in each category.
I look into eight occupational characteristics, grouped into two categories: physical requirements
and cognitive demands of occupations [This is the same methodology as the one used in Case
and Paxson (2006).]. The variables corresponding to physical requirements are manual dexterity,
motor coordination, physical demands (such as climbing, kneeling, and reaching), and strength.
The variables related to the cognitive requirements of a given occupation are spatial, verbal, and
numerical aptitude, as well as intelligence. As in Case and Paxson (2006), in some cases I reversecoded some of these variables, in a way that a higher value corresponds to higher requirements for
a given characteristic in an occupation. Values for each characteristics typically range from 1 to
5, except in the case of physical demands, which ranges from 0 to 4, and motor coordination and
intelligence, which range from 1 to 4.
As expected, cognitive requirements such as verbal and numerical aptitude tend to have higher
values for occupations in upper ISCO categories (1, 2, 3), whereas the opposite is true for physical
requirements such as manual dexterity and strength. This is shown in table 2.3, which lists average

48
scores and standard deviations for each job characteristic for each broad ISCO category. For example,
legal professionals (a sub-category of ISCO category 2, which includes lawyers and judges) have
really high scores for numerical and verbal aptitude (4.2 and 4.4 respectively). On the contrary, this
occupational category only has a score of 0.4 for physical demands, and 1.2 for strength. On the
other hand, agricultural laborers (part of ISCO category 9) have scores of 2 and 2.2 for numerical
and verbal aptitude, whereas they have scores of 2.9 and 3.5 for physical demands and strength.

2.6

Identification using geographical and temporal variation

The main idea behind the identification strategy is to use two sources of variation in order to identify
the effect of iodization on occupational outcomes. The first source of variation is geographical; not
all areas in Switzerland were affected by iodine deficiency to the same extent. What defined the
prevalence of iodine deficiency across localities was the iodine content of soil and drinking water,
which in turn was determined by geological developments in that areas, which took place thousands
of years before the period of study. Using Bircher’s rich dataset, I am able to identify districts where
the prevalence of goiter was very high, signifying a high prevalence of underlying iodine deficiency in
the population. I treat all districts where goiter rates are at the top 25% of the population-weighted
goiter distribution as being high-goiter. These are the districts where goiter prevalence is 11% or
higher.
I expect iodization to have a greater impact on the cognitive ability of those populations born
in high-goiter districts compared to any effects measured for people born in low-goiter districts.
Thus, this source of variation identifies a treatment group (people born in high-goiter districts) and
a control group (people born in low-goiter districts).
The second source of variation is temporal; once the salt monopoly started producing iodized
salt, each canton had to authorize its sale by amending its constitution. As a result, the “new
product” did not become available across all cantons at the same time. Therefore, people born after
iodized salt was introduced constitute the treated population, whereas people born before iodization
are a control group.
As discussed in section 2.4, a glance at the data shows that high-goiter cantons allowed the sale of
iodized salt later, compared to low-goiter cantons. This might reflect reluctance from the part of the
authorities in those cantons to introduce such a far-reaching public health measure, in light of the
debate ignited by the spike in thyroid-related deaths following iodine overdose in some chronically
deficient populations. If this is the only reason of the different timing between high- and low-goiter
cantons, then that should not pose a problem for my identification.
The non-random nature of the difference in the timing of adoption is problematic for my identification strategy to the extent that it reflects time-variant unobserved differences between highand low-goiter cantons, which affected the timing of adoption of iodized salt in the 1920s and 1930s,
and also affected occupational patterns many years later, when the treated population entered the
labor market. Although it seems unlikely that a common unobserved factor would affect both these

49

Table 2.3: Job characteristics for occupational categories in 1970 Swiss Census
ISCO
category
1
2
3
4
5
6
7
8
9
Total
ISCO
category
1
2
3
4
5
6
7
8
9
Total

Manual
Dexterity
2.2052
(0.0680)
2.5256
(0.5371)
2.5544
(0.4548)
2.6782
(0.0961)
2.5622
(0.1734)
2.5081
(0.0653)
3.2093
(0.2143)
2.8371
(0.0898)
2.9034
(0.0554)
2.7578
(0.3795)
Spatial
Aptitude
2.3255
(0.1013)
3.1514
(0.6944)
2.6859
(0.5859)
2.2449
(0.129)
2.3215
(0.2453)
3.0425
(0.1222)
3.1094
(0.3124)
2.6209
(0.1900)
2.4313
(0.0934)
2.7014
(0.4879)

Motor
Coordination
2.1933
(0.0557)
2.4086
(0.3076)
2.5896
(0.3985)
3.0950
(0.4348)
2.5112
(0.2752)
2.2774
(0.0633)
2.8661
(0.0818)
2.7467
(0.1407)
2.4706
(0.1305)
2.6830
(0.8017)
Numerical
Aptitude
3.1489
(0.0665)
3.7347
(0.4738)
3.1512
(0.3687)
2.6962
(0.1324)
2.6254
(0.3455)
2.6244
(0.1009)
2.6329
(0.1953)
2.3042
(0.2146)
2.1324
(0.1665)
2.7381
(0.4460)

Physical
Demands
0.7502
(0.2306)
1.0425
(0.6282)
1.3572
(0.7021)
1.7675
(0.2062)
1.4389
(0.3600)
3.0692
(0.1572)
2.6782
(0.5528)
2.0387
(0.1684)
2.3261
(0.2928)
2.0061
(0.7382)
Verbal
Aptitude
3.7353
(0.0779)
4.3509
(0.2069)
3.5000
(0.3124)
3.1454
(0.0909)
2.8821
(0.2330)
2.7457
(0.0624)
2.7408
(0.1531)
2.5312
(0.2015)
2.3319
(0.1096)
3.0214
(0.5258)

Strength
1.7039
(0.1341)
1.7536
(0.3077)
1.8623
(0.3315)
1.5043
(0.4200)
2.2582
(0.2722)
3.4742
(0.0812)
2.8385
(0.3927)
2.6882
(0.2429)
3.1635
(0.2992)
2.3790
(0.7382)
Intelligence
2.8395
(0.0881)
3.4330
(0.2345)
2.6139
(0.2569)
2.1783
(0.0719)
1.9442
(0.2058)
2.2911
(0.1289)
1.9864
(0.0980)
1.7804
(0.2091)
1.5312
(0.1009)
2.2071
(0.4714)

(Source: 1970 Swiss Census and England and Kilbourne (1988))

50
outcomes, I do introduce a linear, canton-specific time trend in all specifications.
Combining the geographical and temporal sources of variation, the treatment group includes
people born in high-goiter districts after iodized salt was introduced. On the contrary,
people born in low-goiter districts, and people born before iodization form the control group.

2.6.1

Iodized salt sales and birth location

I introduce iodized salt sales at the canton level directly as a regressor. The outcome variables
include an indicator for being employed in a high-paying occupation (as defined by belonging to the
top 3 ISCO one-digit occupational categories), as well as each of the 8 occupational characteristics
described in section 2.5.
In particular, I control for the percentage of iodized salt sales in total salt sales one year prior to
birth in one’s canton of birth. I also control for canton and cohort fixed effects, to remove the effect
of any omitted variables that are canton- or cohort-specific. In addition, I introduce a canton-specific
time trend, to control for any other gradual changes that might have affected one’s occupational
outcomes.
I run the following regression for an individual i born in canton c in year y:
outcomeidy = α + β· Iodized salt 1 year prior to birth
+ Canton of birth Fixed Effects
+ Canton of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where standard errors are clustered at the canton-cohort level.
I estimate the above regression for the sample as a whole and then separately by gender. I cluster
standard errors on the canton of birth - cohort level. My sample includes all individuals born in
Switzerland between 1900 and 1944 (inclusive), for whom occupational data were recorded during
the 1970 Census.

2.6.2

Jumps in iodized salt sales; a Fuzzy Regression Discontinuity Approach

I use West’s data on the percentage of iodized salt sales to get information on the timing and speed of
iodization in each canton. Most cantons experienced a big jump in iodized salt sales within a period
of 1 to 2 years, so that the timing of the intervention is easy to identify in the majority of cases.
The probability of someone being treated with adequate iodine in utero changes discontinuously
depending on whether they were born before or after the jump in sales. I use these jumps in
a “fuzzy regression discontinuity” (FRD) framework (for a description of FRD see, for example,
Imbens and Lemieux (2008)).
Based on the year of the jump in iodized salt sales, which is particular to each canton, I construct
a new variable, age relative to iodization, which will be common across all people born in the same
canton in the same year, but will generally be different across cohorts born in different cantons.

51
Because of the difference in the timing of adoption, I can control for unobserved, time-invariant
canton characteristics, as well as unobserved, space-invariant cohort characteristics, and a cantonsspecific trend. Since these jumps all occurred in a small window of time, this method is “cleaner”
than a simple regression of outcomes on iodized salt sales, as there are fewer possible confounding
effects taking place within same small time frame in a particular canton.
I check for some preliminary evidence of how being born in a high-goiter district right after
a jump in iodized salt sales affects one’s occupational outcomes, after controlling for district and
cohort of birth fixed effects, as well as a district-specific time trend.
I run the following regression for an individual i born in district d in year y:
outcomeidy = α + β· 1 (Born in high-goiter district) X 1 (Born after jump in sales of iodized salt)
+ District of birth Fixed Effects
+ District of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where outcomes are an indicator variable for being employed in a high-paying occupation, as well
as the occupational characteristics described in section 2.5. Standard errors are clustered at the
district-cohort level. First, I run the regression on the whole sample, and then for males and females
separately. I first use the whole sample and then only the sub-sample of those people born within
ten years of the jump in iodized salt sales. In some specifications I use cohort fixed effects that are
common for high-goiter districts and the rest of the country, whereas in others I allow cohort fixed
effects to be different for high-goiter districts.
Next, I follow the FRD approach, and regress outcomes on the percentage of iodized salt sales
(at the canton-level) one year prior to birth (and additional controls), but I instrument iodized salt
sales with an indicator variable equal to 1 if someone is born after the big jump in sales which
marked each canton’s decisive transition to iodized salt.
I run the following regression for an individual i born in canton c in year y:
outcomeidy = α + β· Iodized salt 1 year prior to birth
+ Trend before jump in iodized salt sales
+ Trend after jump in iodized salt sales
+ Canton of birth Fixed Effects
+ Canton of birth time trend
+ Cohort of Birth Fixed Effects +ǫidy ,
where outcomeidy is either an indicator variable for being employed in a high-paying occupation, or
one of the occupational characteristics described in section 2.5. Standard errors are clustered at the
canton-cohort level. I include both a linear canton-specific time trend, but also linear nationwide
trends which are allowed to differ before and after the jump in iodized salt sales.
I estimate the treatment effect for the whole sample and all areas, and then I use only those
observations corresponding to high- and low-goiter districts, and estimate the effects of iodization
separately for each of these two populations. I then estimate effects for males and females separately,
using the same cuts of the data.

52

2.7

The effect of Iodization on Occupational Choice
Figure 2.1: Residuals graph of high-paying dummy and iodized salt

High-goiter areas
6
4
2
0
-2
-4
-6
-8
-10
-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Low-goiter areas
20
15
10
5
0
-5
-10
-15
-0.6

-0.4

-0.2

0

0.2

Residuals of high-paying dummy and iodized salt dummy

0.4

0.6

Linear Fit

Note: Scatterplot of average residuals (at the canton-cohort level) from regressing high-paying occupation dummy
and iodized salt sales one year prior to birth on year and canton of birth fixed effects, as well as a canton-specific time
trend.
(Source: 1970 Swiss Census and England and Kilbourne (1988))

Figure 2.1 is a graphical representation of the regression results outlined later in this section.
Figure 2.1 includes two scatterplots of average residuals (at the canton - year of birth level) from
regressing a high-paying occupation dummy and iodized salt sales one year prior to birth on canton

53
and cohort fixed effects, as well as a canton-specific linear trend. The first scatterplot comes from
including only those individuals born in high-goiter districts, whereas the second one only includes
people born in low-goiter districts. There are fewer observations in the first scatterplot because
iodine deficiency was localized in certain parts of the country, so high-goiter districts are found in
fewer cantons than low-goiter districts. As a result, when residuals are averaged at the canton - year
of birth level, there are fewer observations.
What Figure 2.1 shows is that whereas (regression-adjusted) higher sales of iodized salt around
one’s time of birth are associated with a (regression-adjusted) higher probability of being employed
in a high-paying occupation in the future, such a relationship does not exist for people born in
low-goiter areas, where I expect the treatment effect to have been lower.
Table 2.4 shows results from regressing a high-paying occupation dummy on iodized salt sales (1
year prior to birth), as well as canton and cohort fixed effects. I also include a canton-specific linear
trend. As shown in table 2.4, this makes the coefficients of interest somewhat bigger. The treatment
effect is estimated for the whole sample and then separately by gender, and also for all areas, and
then separately for high- and low-goiter districts only. The coefficients in table 2.4 correspond to
percentage point changes, and standard errors are clustered at the year and canton of birth level.
Looking at the sample as a whole, the treatment effect is significant (when I include the cantonspecific trend) but very small for the country as a whole. However, when I limit the sample to
those observations to individuals born in high-goiter areas, the effect of iodization increases both in
magnitude and significance. Specifically, born in a high-goiter district after iodized salt sales have
gone from 0% to 100% increases one’s probability of employment in the top occupational categories
later in life by 3.2 percentage points. Given that less than a third of the population as a whole is
employed in these categories over the period of examination, this is a significant increase in one’s
chances of receiving higher wages. Low-goiter districts, as expected, were not affected.
When I break the sample down by gender, it is clear that most of the effect of iodization is
driven by males. This is a departure from the conclusion reached in Politi (2009), where the effect of
iodization on increased schooling was mainly driven by the effect on females. However, in this case
females did not participate in the labor force as regularly as males, so there are much fewer female
observations. Indeed, only a third of females in my sample were recorded as being employed. In
addition, female selection into employment was probably not random, therefore it is hard to make
any meaningful conclusions about the effect of iodization on occupational characteristics for females.
Results which are similar in direction but smaller in magnitude are shown on table 2.5, which
estimates a departure from trend in outcomes for those born in high-goiter districts following a
jump in iodized salt sales. Table 2.5, in particular, shows results from a regression of a high-paying
occupation indicator variable on an indicator variable for being in the treatment group (which is the
interaction of two indicator variables: being born in a high-goiter district, and being born after a
jump in iodized salt sales). The regression also includes district fixed effects, a district-specific time
trend (I also show results without including the district-specific trend, and they are very similar),
and year of birth fixed effects, which in columns (1) and (3) are the same for high-goiter districts

54

Table 2.4: Salt sales and occupational choice
Salt sales and occupational choice: Coefficient on Iodized Salt 1 year prior to birth
WHOLE SAMPLE
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
0.69367**
3.20205***
-0.14225
(0.30658)
(0.67162)
(0.40538)
Canton-specific trend
YES
YES
YES
Number of Observations
1620246
402781
397334
R-squared
0.0176
0.013
0.019
High-paying occupation dummy
-0.02581
2.74317***
0.39774
(0.22437)
(0.54258)
(0.35071)
Canton-specific trend
NO
NO
NO
Number of Observations
1620246
402781
397334
R-squared
0.0174
0.013
0.0187
MALES ONLY
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
1.00143***
3.24884***
0.0302
(0.37611)
(0.93727)
(0.50349)
Canton-specific trend
YES
YES
YES
Number of Observations
1145756
288765
280378
R-squared
0.0271
0.0191
0.0288
High-paying occupation dummy
0.26264
3.31585***
0.71948*
(0.27115)
(0.7564)
(0.43602)
Canton-specific trend
NO
NO
NO
Number of Observations
1145756
288765
280378
R-squared
0.0269
0.019
0.0285
FEMALES ONLY
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
0.07639
2.68974**
-0.43441
(0.38722)
(1.19814)
(0.57743)
Canton-specific trend
YES
YES
YES
Number of Observations
474490
114016
116956
R-squared
0.0063
0.0045
0.0089
High-paying occupation dummy
-0.52803*
1.32781
-0.16256
(0.31723)
(0.94442)
(0.51836)
Canton-specific trend
NO
NO
NO
Number of Observations
474490
114016
116956
R-squared
0.0059
0.0044
0.0083
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

55
and the rest of the country, but which are allowed to be different for high-goiter districts in columns
(2) and (4). Standard errors are clustered at the district-year of birth level. Columns (1) and (2)
include the whole sample, whereas columns (3) and (4) only include people born within 10 years of
the jump in iodized salt sales.
As shown on table 2.5, the treatment effect is concentrated on males, regardless of whether we
look at the whole sample or the limited age range. The gain in percentage points for males here is
of the order of 1.2 to 1.5 percentage points, which is somewhat smaller than the effect estimated in
table 2.4. This might because even before the jump in sales, iodized salt was available and consumed
in smaller quantities, so it is likely that part of the control group in this exercise was actually treated
with iodine. According to table 2.5 females did not benefit from iodine supplementation. However,
as discussed above, this is not necessarily accurate since female labor participation was limited over
that period.

High−paying occupation, (high−low) goiter
.01
.02
.03
.04

Figure 2.2: Regression-adjusted probability of being employed in a highly-paid occupation by age
relative to iodization; difference between high- and low-goiter districts

−8

−6

−4

−2
0
2
Age relative to iodization

4

6

8

(Source: 1970 Swiss Census and England and Kilbourne (1988))

Figure 2.2 is a plot of the regression-adjusted difference in probabilities of being employed in a
high-paying occupation people born in high- and low-goiter districts against age relative to iodization.
The line represents the difference in coefficients on relative age indicator variables between those
born in high- and low-goiter areas, after factoring out canton and cohort fixed effects, as well as
a canton-specific trend. It is the regression equivalent of the “Fuzzy Regression Discontinuity”

56

Table 2.5: Salt sales and occupational choice if born after jump
Coefficient on high-goiter dummy X Born after jump in iodized salt sales
WHOLE SAMPLE
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
High-paying occupation dummy 1.10215*** 1.09991***
1.04325**
0.80215
(0.33805)
(0.40945)
(0.45021)
(0.53454)
District-specific trend
YES
YES
YES
YES
Number of Observations
1625400
1625400
689302
689302
R-squared
0.029
0.029
0.0272
0.0272
High-paying occupation dummy 1.20967***
0.88392**
1.1873***
0.62239
(0.3003)
(0.36071)
(0.3248)
(0.46633)
District-specific trend
NO
NO
NO
NO
Number of Observations
1625400
1625400
689302
689302
R-squared
0.0286
0.0287
0.0268
0.0269
MALES ONLY
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
High-paying occupation dummy
1.4672***
1.21396**
1.31045**
1.1849*
(0.42132)
(0.50823)
(0.54966)
(0.68766)
District-specific trend
YES
YES
YES
YES
Number of Observations
1149692
1149692
489047
489047
R-squared
0.0443
0.0443
0.0418
0.0418
High-paying occupation dummy 1.55246***
1.11773**
1.68413***
0.95469*
(0.39325)
(0.458)
(0.41176)
(0.57837)
District-specific trend
NO
NO
NO
NO
Number of Observations
1149692
1149692
489047
489047
R-squared
0.0439
0.044
0.0413
0.0413
FEMALES ONLY
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
High-paying occupation dummy
0.36269
0.84244
0.17463
0.27335
(0.56023)
(0.67139)
(0.82643)
(0.83241)
District-specific trend
YES
YES
YES
YES
Number of Observations
475708
475708
200255
200255
R-squared
0.0104
0.0105
0.0115
0.0116
High-paying occupation dummy
0.48832
0.47301
0.18643
0.20132
(0.42429)
(0.52234)
(0.5024)
(0.72193)
District-specific trend
NO
NO
NO
NO
Number of Observations
475708
475708
200255
200255
R-squared
0.0095
0.0096
0.0106
0.0107
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
Columns (1) and (3) assume common cohort fixed effects for high- and non-high-goiter districts.
Columns (2) and (4) assume different cohort fixed effects for high- and non-high-goiter districts.
Standard errors in parentheses, clustered at the district-year of birth level.

57
approach, the results of which are shown on table 2.6 (A short description of the FRD method can
be found in section 2.6.2). Figure 2.2 shows that right around the time of a jump in iodized salt
sales there was a corresponding jump in a cohort’s probability of being employed in a high-paying
occupation when they entered the labor market.
Table 2.6 tells the same story. The results of the “Fuzzy Regression Discontinuity” approach
are much more similar to table 2.4 than table 2.5 was, although they are still somewhat smaller
in magnitude. According to table 2.6, when looking at the country as a whole, iodization seems
to not have had much of an effect on people’s occupational choice. But the intra-region difference
becomes clear when one looks separately at people born in high- and low-goiter districts. People
born in high-goiter districts after a jump in iodized salt sales had a higher and significant probability
of being employed in a high-paying occupation. The increase in the probability is of the order of
2.3 percentage points, which is economically significant, given the low baseline probability of being
employed in the top occupational categories (less than 30%). As in table 2.5, the effect is driven by
males, where the probability of being employed in a high-paying occupation rose by 3.1 percentage
points.
In high-goiter areas, 23.4% of all people born prior to a jump in iodized salt sales selected into a
high-paying occupation, whereas almost 30% of people born after a jump in sales were employed in
one of the top three occupational categories. Using my estimates, this means that at least 37% of
the change can be directly attributed to iodization. For males, the percentage of people employed
in a high-paying occupation rises from 25.4% to 32.2%. If iodization increases one’s chances of
belonging to that category by 3.1%, then my estimates imply that iodization accounts for almost
45% of the rise in the percentage of males employed in high-paying occupations over that period.
Therefore, iodization accounts for a big part of the shift towards higher-paying occupations observed
in high-goiter areas over the period of examination.

2.8

The effect of Iodization on Characteristics of Occupations

In this section I analyze the effects of the introduction of iodized salt on occupational characteristics.
As outlined in section 2.5, I match eight characteristics of occupations from the book “Occupational
Measures from the Dictionary of Occupational Titles for 1980 Census Detailed Occupations” with
ISCO occupational codes used in the Swiss Census (England and Kilbourne 1988). These characteristics sum up the physical and cognitive skills required by a given occupation. The variables
used are: Manual Dexterity, Motor Coordination, Physical Demands, Strength, Spatial Aptitude,
Numerical Aptitude, Verbal Aptitude, and Intelligence. The first four of these characteristics refer
to physical skills, whereas the final four reflect cognitive skills required by a given occupation.
Figure 2.3 shows how the mix of four of these characteristics changed differentially for high- and
low-goiter areas after iodization. More specifically, figure 2.3 shows the coefficients on an indicator
variable equal to 1 if an individual was born after a jump in sales of iodized salt, in a regression which

58

Table 2.6: High-paying occupation: Fuzzy Regression Discontinuity
Salt sales and occupational choice: Coefficient on Iodized Salt 1 year prior to birth
WHOLE SAMPLE
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
0.15069
2.29497**
-0.42023
(0.36849)
(0.90451)
(0.40623)
Canton-specific trend
YES
YES
YES
Number of Observations
1620246
402781
397334
R-squared
0.0176
0.0131
0.019
High-paying occupation dummy
0.21984
1.71321*
-0.93815**
(0.41011)
(0.9106)
(0.4299)
Canton-specific trend
NO
NO
NO
Number of Observations
1620246
402781
397334
R-squared
0.0174
0.013
0.0188
MALES ONLY
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
0.28545
3.14385**
-0.28615
(0.45591)
(1.30235)
(0.5056)
Canton-specific trend
YES
YES
YES
Number of Observations
1145756
288765
280378
R-squared
0.0271
0.0191
0.0288
High-paying occupation dummy
0.47256
2.88057**
-0.97047*
(0.49684)
(1.2956)
(0.52103)
Canton-specific trend
NO
NO
NO
Number of Observations
1145756
288765
280378
R-squared
0.0269
0.019
0.0286
FEMALES ONLY
ALL AREAS HIGH-GOITER
LOW-GOITER
High-paying occupation dummy
-0.3093
0.44348
-0.69028
(0.45086)
(1.5619)
(0.62258)
Canton-specific trend
YES
YES
YES
Number of Observations
474490
114016
116956
R-squared
0.0063
0.0045
0.0089
High-paying occupation dummy
-0.54469
-0.52011
-0.76598
(0.46783)
(1.41248)
(0.68906)
Canton-specific trend
NO
NO
NO
Number of Observations
474490
114016
116956
R-squared
0.0059
0.0044
0.0083
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Coefficients correspond to changes in percentage points.
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

59

Figure 2.3: Change in occupational characteristics after iodization

Occupational characteristics -- change after iodization

0.02
0.015
0.01
0.005
0
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
physical

strength
High-goiter

numerical

verbal

Low-goiter

(Source: 1970 Swiss Census and England and Kilbourne (1988))

also controls for canton and cohort fixed effects, as well as a canton-specific trend. What is easy to
see from figure 2.3 is that whereas people in low-goiter areas had the same mix of characteristics preand post-iodization, high-goiter areas saw a decrease in the physical and a corresponding increase in
the cognitive component of occupations, signifying a shift towards occupations with higher cognitive
demands than before.
Table 2.11 shows results for the sample as a whole of regressing each occupational characteristic
on the percentage of iodized salt sales one year prior to birth, controlling for canton and cohort fixed
effects, as well as a canton-specific trend. Looking at all areas, there seems to be a slight decrease
in physical skills, and a slight increase in cognitive skills. When we only look at high-goiter regions,
the coefficients increase both in magnitude and significance, with the exception of manual dexterity
and spatial ability. The change is of the order of 0.03 (in absolute value), which, depending on the
coefficient, is between 4 and 8% of a standard deviation in the sample. This is a tiny change, but it
is significant. Moreover, as in the case of high-paying occupations, results are driven by the effect
of iodization on males. This is apparent once we look at Table 2.8, which breaks the sample down
by gender. There’s a big drop in physical demands (kneeling, climbing, reaching, etc) for females,
and a big increase in verbal abilities, but no other characteristic seems affected.

60

Table 2.7: Occupational characteristics: Coefficient on Iodized Salt 1 year prior to birth, whole
sample

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

ALL AREAS
-0.0065267**
(0.0025958)
-0.0019655
(0.0023743)
-0.0093404*
(0.0050766)
-0.0069763
(0.0046804)
0.0015487
(0.0031976)
0.0110898***
(0.0035736)
0.0107766***
(0.0037826)
0.0107693***
(0.0035202)
1542006

WHOLE SAMPLE
HIGH-GOITER
LOW-GOITER
-0.0054148
-0.0055883
(0.007351)
(0.0037123)
-0.0094397
-0.0007528
(0.0070648)
(0.0038487)
-0.0362085**
0.006235
(0.0147821)
(0.007376)
-0.0283815**
0.0039593
(0.0133594)
(0.0061834)
-0.0007003
0.0058368
(0.0097206)
(0.004653)
0.033474***
0.0091073*
(0.0076685)
(0.004735)
0.0270125***
0.0060569
(0.0080725)
(0.004766)
0.0221547***
0.0086688*
(0.0074428)
(0.0045106)
385322
377347

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

61

Table 2.8: Occupational characteristics: Coefficient on Iodized Salt 1 year prior to birth, by gender

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

ALL AREAS
-0.0085004***
(0.0032936)
-0.005372**
(0.0026382)
-0.0108344*
(0.0061104)
-0.0054226
(0.0056623)
0.0028744
(0.0039967)
0.0125323***
(0.0045392)
0.0121351***
(0.0045387)
0.0124379***
(0.004258)
1080860
ALL AREAS
-0.0032167
(0.0031127)
0.0044321
(0.0048784)
-0.0083425
(0.0077268)
-.0111087
(0.007762)
-0.0014131
(0.0049119)
0.0087602**
(0.0044105)
0.0083115
(0.0054498)
0.0079433
(0.0049782)
461146

MALES ONLY
HIGH-GOITER
LOW-GOITER
-0.0083534
-0.0050923
(0.0085289)
(0.0048492)
-0.0125601*
-0.0007246
(0.0066056)
(0.0043759)
-0.0306461*
0.0074013
(0.0159415)
(0.0088791)
-0.026537*
0.0051245
(0.013511)
(0.0076101)
0.0018137
0.0066139
(0.0116746)
(0.0058114)
0.0397279***
0.0075499
(0.0093934)
(0.0061012)
0.0268816***
0.0069954
(0.0095628)
(0.0062836)
0.0230924***
0.0094617
(0.0087863)
(0.0058803)
274227
263677
FEMALES ONLY
HIGH-GOITER
LOW-GOITER
0.0013695
-0.006952
(0.0095191)
(0.0043611)
-0.000404
-0.0020479
(0.0135774)
(0.0073923)
-0.0488137**
0.0044071
(0.0198685)
(0.0092193)
-0.0324223
0.0038235
(0.0201268)
(0.0104108)
-0.0081934
0.0070284
(0.0126684)
(0.0067052)
0.0167783
0.0145821**
(0.0126285)
(0.0059269)
0.0256555*
0.0038328
(0.0145434)
(0.007123)
0.01827
0.0081388
(0.0149586)
(0.0063557)
111095
113670

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

62

Table 2.9: Occupational characteristics: Born after iodization, whole sample
WHOLE SAMPLE
Born +/- 10 years from jump
(2)
(3)
(4)
0.0036855
0.0077488*
0.0071808
(0.0038862)
(0.0045885)
(0.0057727)
-0.0001699
0.0020538
0.0019644
(0.0040769)
(0.004833)
(0.0056653)
-0.0220961***
-0.006845
-0.0130432
(0.0083442)
(0.0097762)
(0.0125692)
-0.0172355**
-0.0102562
-0.0140808
(0.00785)
(0.0093749)
(0.0108008)
-0.0014481
0.0122934**
0.0084329
(0.0050618)
(0.0060556)
(0.0075997)
0.0125876***
0.017797***
0.0174269***
(0.0042332)
(0.0047274)
(0.0057999)
0.0093804*
0.0149158***
0.0143708**
(0.0049353)
(0.005796)
(0.0068862)
0.0042264
0.0116246**
0.0085775
(0.0043923)
(0.0049898)
(0.0058129)
1546952
655388
655388

All cohorts
Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

(1)
0.0040157
(0.0031435)
0.0006691
(0.003453)
-0.013304*
(0.0069865)
-0.0082357
(0.0067016)
-0.0026036
(0.0043484)
0.0086838**
(0.0033974)
0.0065171
(0.0040434)
0.0037387
(0.0036296)
1546952

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Columns (1) and (3) assume common cohort fixed effects for high- and non-high-goiter districts.
Columns (2) and (4) assume different cohort fixed effects for high- and non-high-goiter districts.
Standard errors in parentheses, clustered at the district-year of birth level.

63

Table 2.10: Occupational characteristics: Born after iodization, by gender
MALES ONLY
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
0.0023531
0.0030134
0.0071247
0.0084666
(0.0038217)
(0.0046336)
(0.0054618)
(0.0065581)
-0.0025219
-0.0025243
-0.0003748
-0.0017929
(0.0035052)
(0.0041845)
(0.0049609)
(0.005474)
-0.0125781
-0.0190955*
-0.0114189
-0.0108484
(0.0083503)
(0.0097889)
(0.0112654)
(0.0142136)
-0.0046247
-0.0142717*
-0.0138346
-0.0130591
(0.0075081)
(0.0086543)
(0.0098678)
(0.0113806)
-0.0006729
0.0004197
0.0149853**
0.0167077*
(0.0051193)
(0.0058652)
(0.0072967)
(0.0085963)
0.0127004*** 0.0163939*** 0.0241815*** 0.0263489***
(0.0041941)
(0.0051006)
(0.0054625)
(0.0067043)
0.0091454*
0.0095034
0.0196572***
0.019442**
(0.0048549)
(0.0058286)
(0.0064798)
(0.0077836)
0.0074831*
0.0050152
0.0157109***
0.014659**
(0.0042551)
(0.0050422)
(0.0055496)
(0.0066602)
FEMALES ONLY
All cohorts
Born +/- 10 years from jump
(1)
(2)
(3)
(4)
0.0064285
0.0073436
0.0085276
0.007532
(0.0044443)
(0.0056172)
(0.0066839)
(0.0074156)
0.0060262
0.0048897
0.010965
0.0111891
(0.0059711)
(0.0072029)
(0.0084836)
(0.0099415)
-0.0197688*
-0.0209538
0.0025781
-0.0097583
(0.0115914)
(0.0141631)
(0.0156304)
(0.0185495)
-0.0188164*
-0.0159859
-0.005453
-0.0066519
(0.0110188)
(0.013405)
(0.0156243)
(0.0179133)
-0.0096534
0.0012857
0.0026633
-0.001783
(0.0065412)
(0.008028)
(0.00846)
(0.009888)
-0.0017882
0.0058379
-0.0001548
-0.0005088
(0.0059043)
(0.0071167)
(0.0087851)
(0.0095157)
0.0010073
0.0080343
0.001617
0.0025178
(0.0069625)
(0.0084891)
(0.0108522)
(0.0117831)
-0.0053049
0.0040373
-0.0005824
-0.0023293
(0.0066161)
(0.0080397)
(0.0100097)
(0.0104068)
462336
462336
194259
194259
All cohorts

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
Columns (1) and (3) assume common cohort fixed effects for high- and non-high-goiter districts.
Columns (2) and (4) assume different cohort fixed effects for high- and non-high-goiter districts.
Standard errors in parentheses, clustered at the district-year of birth level.

64

Table 2.11: Occupational characteristics: Fuzzy Regression Discontinuity, whole sample

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

ALL AREAS
-0.0009472
(0.0028367)
0.0000163
(0.0025496)
-0.0076898
(0.0058905)
-0.0040244
(0.0055137)
-0.0011264
(0.0036201)
0.0035306
(0.0042218)
0.0003645
(0.004452)
-0.0009932
(0.0040607)
1542006

WHOLE SAMPLE
HIGH-GOITER
LOW-GOITER
0.0041189
-0.0024479
(0.0118249)
(0.0039328)
-0.0024329
-0.0007358
(0.008796)
(0.0038433)
-0.0532868**
0.0066527
(0.0211891)
(0.0079619)
-0.0363888**
0.0057858
(0.0174469)
(0.0067376)
-0.0008482
0.004219
(0.013549)
(0.0049332)
0.0294672***
0.0045859
(0.0101903)
(0.0045452)
0.0258381**
0.001003
(0.0107883)
(0.0048576)
0.0129775
0.0033626
(0.0091739)
(0.0043635)
385322
377347

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

65

Table 2.12: Occupational characteristics: Fuzzy Regression Discontinuity, by gender

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

Manual Dexterity
Motor Control
Physical Demands
Strength
Spatial Ability
Numerical Ability
Verbal Ability
Intelligence
Number of observations

ALL AREAS
-0.0030607
(0.0035813)
-0.003096
(0.0029474)
-0.0092432
(0.0070682)
-0.000927
(0.0064875)
-0.0006293
(0.0048196)
0.0049506
(0.0053631)
0.0002278
(0.0052992)
-0.0008748
(0.004987)
1080860
ALL AREAS
0.0040697
(0.0035782)
0.0088829
(0.0055917)
-0.0030604
(0.0090979)
-0.011987
(0.0095886)
-0.0031325
(0.00581)
-0.0008476
(0.004911)
-0.0005356
(0.0061051)
-0.0021015
(0.0057062)
461146

MALES ONLY
HIGH-GOITER
LOW-GOITER
0.0053442
-0.0026801
(0.0140892)
(0.0050988)
-0.0086938
-0.0013496
(0.0098406)
(0.0045082)
-0.0439352*
0.0055413
(0.0246903)
(0.0097755)
-0.0306259*
0.006546
(0.0174439)
(0.00822)
0.0129399
0.0042163
(0.0183486)
(0.0062829)
0.041329***
0.0030917
(0.0128449)
(0.0058297)
0.0331184***
0.0018294
(0.0118991)
(0.0064191)
0.0221407**
0.0035545
(0.0106555)
(0.0058077)
274227
263677
FEMALES ONLY
HIGH-GOITER
LOW-GOITER
0.0060009
-0.002506
(0.0144809)
(0.0046641)
0.0132889
0.0007584
(0.0190334)
(0.0077163)
-0.0587841**
0.0073969
(0.0236899)
(0.0100802)
-0.0324252
0.0024774
(0.0290438)
(0.011036)
-0.0160187
0.0055695
(0.0155782)
(0.0071093)
0.0087092
0.0096535
(0.0168866)
(0.0060933)
0.0088448
-0.0008637
(0.0193026)
(0.0074627)
-0.0023815
0.0034093
(0.0206379)
(0.0067172)
111095
113670

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%
High-goiter districts are those belonging to the top 25% of the population-weighted goiter distribution.
Low-goiter districts are those belonging to the bottom 25% of the population-weighted goiter distribution.
All regressions include people born in Switzerland in 1900-1944 and interviewed for the 1970 Census.
Standard errors in parentheses, clustered at the canton-year of birth level.

66
Tables 2.9 and 2.10 show the results of regressing occupational characteristics on an indicator
variable for belonging to the treatment group (born in a high-goiter district after a jump in iodized
salt sales) and other covariates. There are cohort fixed effects which are allowed to be both common
across all districts and different for high-goiter districts and the rest of the country. As in the case
of table 2.5, which was the equivalent of table 2.9 when I used a high-paying occupation dummy
as the outcome of interest, the results here are weaker than in tables 2.7 and 2.8, and for most
characteristics the average value of a significant coefficient is between 0.011 and 0.022. However, the
general results are the same: positive coefficients are related to cognitive skills, whereas negative
coefficients correspond to physical skills. These results are strong for males (table 2.10), especially
when I limit the sample to those born within ten years of the intervention. As in the previous table,
females don’t seem affected, although there is some evidence of a decrease in the physical demands
and strength requirement of the occupation that they select into.
Results from the fuzzy regression discontinuity exercise appear on tables 2.11 and 2.12. The
values of coefficients for the high-goiter group are somewhat larger in absolute value compared
to those in tables 2.7 and 2.8, especially for physical skills, and for the results on males born in
high-goiter districts (Table 2.12). Coefficients on occupational characteristics change by 0.025-0.053
points, or 7.2% of a standard deviation in the case of physical demands. There is a big decrease
in the physical demands component for females (table 2.12), which goes down by 0.06 points, more
than 8% of a standard deviation.
To summarize, the mix of occupational characteristics shifted slightly towards a bigger cognitive
component for those cohorts born after iodization in previously highly-deficient areas. The shift was
very small -accounting at most for about 8% of a standard deviation- but it was significant. The
effect is stronger for males than for females, but as discussed above, female labor market participation
was very limited and possibly non-random over that period, so the results on females are harder to
interpret.

2.9

Concluding remarks

The Swiss Iodization Campaign was the first major nutritional intervention, and many others, such
as milk fortification with vitamin D, followed. Yet its effects on economic outcomes have never
been studied before. The effects of correcting iodine deficiency on cognition had real and significant
effects for the populations treated. The cost-effectiveness of providing universal prophylaxis with
iodized salt is indisputable, even if one only takes into account the lower hospitalization costs of
people with thyroid disorders, without factoring in the big impact on the economy that increased
cognition might have in a population.
Apart from an policy impact evaluation, though, the Swiss Iodization Campaign can also be
viewed as an experiment of what happens to a given country and a given population within that
country when there is an arguably exogenous change in its cognitive ability. Indeed, one can view

67
iodine deficiency eradication as an “injection of IQ” in the population8 .
Once one thinks about iodization in that light, new and interesting empirical questions pop up.
For example, it is usually very hard to de-couple innate ability from schooling when measuring the
returns to education in a population, because there is usually not enough variation in the data, as
presumably people of low ability do not achieve high education levels, whereas most high-ability
people do (for a discussion of this issue, see Heckman and Vytlacil (2001)). In the case of iodization
interventions, however, we are faced with a change in overall ability in the population for all levels
of schooling, so we are already one step further in solving that problem. It is still a complex issue,
however, since one would have to make assumptions about the initial distribution of ability in the
population prior to the intervention, and how it changed after the correction. Average observed
ability could increase either by a shift of the ability distribution to the right, or by a decrease of its
left tail, however the implications for the interpretation of the findings would be different in each
case, so would be our capacity to de-couple ability from schooling.
Another question arises from treating iodization in a given population as resulting in a shift of
its comparative advantage, as workers select into more cognitive-demanding occupations. This can
have implications for the local economy (which we can treat as open), as well as the economy of
the country as a whole. For example, do we observe worker migration into other areas, and if yes,
what is the profile of the workers who migrate and of the receiving regions? How are local and
countrywide relative wages affected? These are intriguing questions which could be answered by
looking at the various iodization interventions that have taken place around the world, starting with
the Swiss and North American experience of the 1920s.
Tremendous steps have been taken for the global eradication of iodine deficiency in the past 20
years, but there is still a long way to go before nobody is affected by iodine deficiency disorders.
Given the obvious cost-effectiveness of universal salt iodization, and the beneficial impacts for the
treated populations as they were estimated in this paper, it should be clear that eradicating iodine
deficiency should be at the top list of priorities for governments and health-promoting organizations.

8I

thank Prof. Yona Rubinstein who came up with this graphic expression.

Chapter 3

The Economic Effects of
Micronutrient Deficiency:
Evidence from Salt Iodization in
the United States1
3.1

Introduction

Public health interventions in developing countries have lately been the center of media attention and
scholarly work. This interest is justified by the potentially significant impact of large-scale health
interventions on people’s productivity, longevity, and quality of life. Improvements in productivity
and life expectancy could have strong positive effects on developing countries’ economic growth, while
differences in access to health might help explain a big part of income gaps between countries.2 It is
interesting to note that many of the health problems faced by developing countries were only recently
solved in the developed world. Many authors have therefore turned to the historical experience of
currently developed countries in order to estimate the potential benefits of health interventions.3
Micronutrient deficiencies are widespread in developing countries, and in the last decades a global
effort has been launched to ameliorate them. However, little evidence exists about the aggregate,
1 This chapter represents co-authored work with David N. Weil (Brown University) and James Feyrer (Dartmouth
College).
2 For

instance, Sachs (2003) shows that efforts to eliminate malaria have substantial effects on income, through their
effect on health, reduced absenteeism, etc. Acemoglu and Johnson (2007) look at the effect of health interventions
and find little effect. Weil (2007) charts a middle course.

3

For example, Bleakley (2007) examines the effects of hookworm eradication in the American South in early
20th century and finds significant effects on education and future incomes of those cohorts that benefited from
the intervention. Also, Watson (2006) finds that improvements in sanitation of Indian reservations in the 1960’s
explain a big part of the convergence in infant mortality rates between Whites and Native Americans.

68

69
long-run effects of such interventions on either population health or economic measures such as labor
productivity.4 In this paper we address these issues by looking to the history of the United States,
which like many other developed countries, undertook early in the 20th century interventions similar
in scope and intent to those being applied in developing countries today. In particular, we examine
the iodization of salt.
Severe iodine deficiency in utero is known to cause significant development damage, including
decreases in cognitive ability. Iodine deficiency is linked directly to geography through the food and
water supply. In adults, the most noticeable symptom of iodine deficiency is goiter, the enlargement
of the thyroid gland. Prior to salt iodization, endemic goiter and other iodine deficiency disorders
were present in specific regions of the US and absent from others, depending on the iodine content
of the soil and water. Figure 3.1 illustrates the geographic distribution of goiter in the US as
measured among World War I recruits (we discuss the data further below). In 1924 iodized salt
was introduced in the United States explicitly to reduce the goiter rate. This intervention rapidly
reduced the incidence of iodine deficiency.
Figure 3.1: Distribution of Goiter during WWI

source: Defects found in Drafted Men (Love 1920)
4 The

only study that we know of examining the effect of iodization is Field, Robles, and Torero (2007). They
find that in Tanzania, treatment of mothers with iodated oil resulted in a rise in schooling of 0.33 years among
children, with a larger effect for girls than boys.

70
Iodization of salt in the US provides a particularly good natural experiment due to the geographic
distribution of the disease combined with a rapid, complete treatment. Since there are large in utero
effects of iodine deficiency, we should see a significant difference between those born before and after
the introduction of iodized salt in locations with low levels of environmental iodine. Those living in
high iodine regions provide a control group.
We exploit two unique data sources to look at the effects. After World War I, statistics from
draft physicals were compiled by geographic location. From this source we know the incidence of
goiter for 151 geographic regions before the introduction of iodized salt. This provides us with a
pretreatment measure of iodine deficiency.
Our outcome measure is provided by an extensive dataset of men who enlisted in the Army
during World War II. The timing of the war generates a large sample of men born in the years
1920-1927, neatly covering the introduction of iodized salt. Upon enlistment, each recruit took the
Army General Classification Test (AGCT), a forerunner to the AFQT. The Air Corps was assigned
draftees with significantly higher average test scores than the ground forces. We exploit this fact to
get a crude estimate of each recruit’s AGCT score based on their assignment.
The probability of assignment to the Air Corps rises significantly in low iodine (i.e. high goiter)
counties in the years after the introduction of iodized salt. In the lowest iodine regions, our estimates
suggest a 10-20 percent increase in the probability of a man being assigned to the Air Corps after
iodization.
Using information about average scores of Air and Ground Force recruits we can infer a onequarter to one half standard deviation increase in average test scores in these regions. The average
level of iodine deficiency in the US was significantly lower than in the highest regions, so the overall
effect in the US was much more modest, though iodization was undoubtedly extremely cost effective.
The increase in cognitive ability due to salt iodization may have contributed a small amount to trend
rise in measured IQ that took place over the course of the twentieth century, the so-called Flynn
Effect.
The paper proceeds as follows: section 3.2 provides some background on iodine deficiency disorders. Section 3.3 outlines the history of salt iodization in the US. Section 3.4 describes our data
and provides some background on their collection. Section 3.5 explains our identification strategy
and section 3.6 presents our results. Section 3.7 interprets our results and puts them in the context
of related research on the effects of iodization. Section 3.8 concludes.

3.2

Iodine Deficiency Disorders

Recent work has shown that the quality of maternal health and nutrition during pregnancy has
persistent effects on adult health outcomes. For instance, Almond (2006) shows that cohorts exposed
to the Spanish Influenza of 1918, either in utero or during the first months of life, had worse health
and socio-economic outcomes in their lifetime. Behrman and Rosenzweig (2004) find that differences
in birth weight among identical twins are reflected in differences in school attainment and adult

71
earnings. Childhood height, which depends, among other factors, on the environment in utero and
maternal health during pregnancy, is a significant predictor of both the performance in cognitive
tests at a young age, and also of adult earnings (Case and Paxson 2008).
Iodine is one of the “big three” micronutrients, deficiencies in which are a major source of ill
health in developing countries (the other two are vitamin A and iron). Iodine deficiency, in particular,
is the leading cause of preventable mental retardation in the world. WHO estimates that nearly 50
million people suffer some degree of mental impairment due to iodine deficiency5 . Two billion people
-one third of the world’s population- are at risk, in the sense that their iodine intake is considered
insufficient. According to WHO’s Global Database on Iodine Deficiency, more than 285 million
children receive inadequate amounts of iodine in their diet (de Benoist et al., eds 2004).
Iodine Deficiency Disorders (IDD) is a term used to describe a range of anomalies, ranging from
goiter to cretinism, due to inadequate provision of iodine. Iodine is a micronutrient essential for
the synthesis of the two thyroid hormones, thyroxine and triiodothyronine. These hormones are
necessary for metabolism, the neuromuscular system and the reproductive function.
Around 70-80% of the total iodine content in the human body is found in the thyroid gland
(Fleischer, Forbes, Harriss, Krook and Kubota 1974). The thyroid gland uses iodine to produce
thyroxin, a hormone that regulates the metabolism. When there is too little iodine in the diet,
the thyroid enlarges, forming a goiter. This enlargement (governed by thyroid stimulating hormone
produced by the pituitary gland in response to low thyroxin) allows the thyroid to produce more
thyroxin for a given availability of iodine, and can fully or partially compensate for the shortage
of iodine. When dietary iodine is only slightly inadequate, the enlarged thyroid will be able to
produce sufficient thyroxin for normal body functioning. This is known as euthyroid goiter. At
lower levels of dietary iodine, the enlarged thyroid will produce inadequate thyroxin, a condition
known as hypothyroid goiter, characterized by slow metabolism, lethargy, and weight gain.
In most individuals with goiters due to iodine deficiency, an increase in dietary iodine will result
in the thyroid gland returning to its normal size and thyroxin production remaining at or returning
to its proper level. However, in some people with this condition, increased iodine consumption results
in the thyroid glad producing excessive quantities of thyroxin, resulting in hyperthyroidism. This
is called iodine-induced thyrotoxicosis. Hyperthyroidism is characterized by a too-fast metabolism,
with symptoms including rapid heartbeat, weight loss, temperature elevation, nervousness, and
irritability67 . The problem of iodine induced thyrotoxicosis is most likely to occur in individuals
who have experienced long periods of iodine deficiency and those with “nodular goiter”. Nodular
goiter represents a later stage of the disease; it is preceded by diffuse enlargement or simple goiter8 .
5

source: WHO, http://www.who.int/features/qa/17/en/index.html.

6 Reference:

”Goiter” in Health A to Z, www.healthatoz.com

7 Goiter

and hyperthyroidism can also result from Graves Disease, also called Basedow disease, an immune condition
in which the thyroid is stimulated to produce excess thyroxin.

8 Iodine-induced

thyrotoxicosis is also called Jod-Basedow disease. “Jod” is German for iodine. The name indicates
that iodine consumption is resulting in the symptoms of Basedow disease. Examining the rise in thyroid disease
that followed the introduction of iodized bread in Tasmania, Connolly (1971) found that most patients with
iodine-induced thyrotoxicosis had pre-existing nodular goiter, and few had Graves disease.

72
Beyond goiter and the associated effects of iodine on the metabolism, a second effect of iodine
deficiency is in utero. Iodine deficiency early in pregnancy causes serious brain damage to the
foetus. Unlike the effects of goiter, this damage is permanent. Severe iodine deficiency can result
in cretinism, which is characterized by “profound mental deficiency, dwarfism, spastic dysplasia and
limited hearing” (Scrimshaw 1998, p.364)9 . However, as Scrimshaw points out, “even in areas where
cases of cretinism due to iodine deficiency in the mother are few, the linear growth of the infant, its
intellectual capacity, and certain other of its neurological functions are permanently compromised
to varying degrees” (Scrimshaw 1998, p.351). In other words, even if iodine deficiency does not
result in cretinism, an iodine-deficient region will be marked by the lower cognitive performance
of its population. Typically, non-deficient populations differ from iodine-deficient populations by
approximately 10 IQ points, whereas the whole normal IQ distribution of a population shifts to
the left as a result of iodine deficiency10 . In endemic areas, cretinism can affect up to 15% of the
population (de Benoist et al., eds 2004). Bleichrodt and Born (1994) estimate that the average IQ
of iodine-deficient groups is 13.5 points lower than the non-deficient groups. If this is true, then
iodine deficiency should have sizable economic effects for any affected population.
In economies where the diet is composed primarily of locally produced food, the main determinant
of whether a population will be iodine-deficient or not is a region’s geography. Ocean water is rich
in iodine, which is why endemic goiter is not observed in coastal areas. From the ocean, iodine is
transferred to the soil by rain. This process, however, only reaches the upper layers of soil, and it
can take thousands of years to complete (Koutras et al. 1980). Heavy rainfall can cause soil erosion,
in which case the iodine-rich upper layers of soil are washed away. The last glacial period had the
same effect; iodine-rich soil was substituted by iodine-poor soil from crystalline rocks (Koutras et
al. 1980). This explains the prevalence of endemic goiter in regions that were marked by intense
glaciation, such as Switzerland and the Great Lakes region. Iodine is taken up by plants when it
is present in the soil, and can reach humans either through plants or animals which have eaten
them. Iodine is also present in subsurface water in some locations. Finally, deposits of mineral salt
(the remains of evaporated seawater) contain iodine, but this is lost when the salt is refined. The
human body does not naturally store a great deal of iodine, so that seasonal variations in iodine
consumption may result in seasonal manifestations of IDD.
Even before the discovery of iodine, ancient civilizations treated goiter with burnt sponge or
seaweed (Curtis and Fertman 1951, Langer 1960). After iodine was discovered in 1811 by Courtois,
continuous clinical research for over a century proved its essential role as a measure of prophylaxis
against IDD. Doctors and public health officials have used different ways to ensure that adequate
quantities of iodine are provided for a given population. Salt iodization has proved to be the
cheapest and most wide-reaching way to protect a population from iodine deficiency. Alternatives
9

According to one interpretation, the word cretin comes from the French term for Christian. Cretinism was
endemic in the French Alps, where the term was apparently invented for those who were too dumb to commit a
sin and who, therefore, were good Christians.

10 Scrimshaw

(1998) provides a list of studies and experiments that have been conducted, and which have shown
the hindering effects on mental development of iodine deficiency in utero.

73
have included the iodization of water supplies and bread, as well as the provision of iodine-enriched
chocolates or milk to babies and schoolchildren and injections of slow-releasing iodated oil11 .

3.3

Iodine Deficiency and Salt Iodization in the United States

It was only in the last century that micronutrient deficiency diseases were eliminated in the United
States and other developed countries. Pellagra, a deficiency in niacin that results from diets dependent on maize, was endemic in the US South at the beginning of the twentieth century. Rickets, a
bone-deforming disease caused by deficiency of vitamin D, was common in industrial cities of the
North. Both diseases were controlled by a combination of dietary improvements and fortification.
Vitamin D was added to milk in the 1930s, and B vitamins (including niacin) were added to fortified
baked goods starting in the 1940s (Bishai and Nalubola 2002).
Salt iodization was the first experiment in the systematic fortification of food to combat micronutrient deficiency. This public health intervention was made possible by the nearly simultaneous
discovery of a widespread health problem and of its underlying cause.
In the First World War draft, a little more than 2.5 million draftees were examined for various
physical and mental shortcomings. From these examinations, a lengthy collection of countrywide
data was compiled, showing the geographic distribution of many diseases and defects across the
United States (Love and Davenport 1920). Goiter was among the defects that were measured,
because an unexpectedly high number of soldiers had trouble wearing a uniform because of their
enlarged thyroids.
According to the draft examinations, almost 12,000 men had goiter and a third of these were
judged unfit for service, because the size of their neck was too big for the military tunic to be buttoned
(Kelly and Snedden 1960, p.34). Most of them came from states in the Northwest (Washington,
Oregon, Idaho, Montana) and the area around the Great Lakes. In Northern Michigan, for instance,
more draftees were judged unfit for service “for large and toxic goiters than for any other medical
disorder” (Markel 1987, p.221). On the other hand, goiter was rare in people coming from coastal
areas.
The realization of the problem led to multiple surveys of goiter, which confirmed the geographical
variation in the prevalence of goiter. By that time, there was medical and veterinary evidence
showing that goiters could be reduced by adding iodine to the diet. Experiments with school children
confirmed that the size of goiters decreased after receiving iodine12 . These observations prompted
a public debate on the possible ways to provide iodine prophylaxis to the American population.
Some objections were raised as to the potential side-effects of such a global measure. It had been
documented that large amounts of iodine could cause hyperthyroidism to develop in some adults and
11 Iodization

of water supplies proved wasteful since only a small proportion of water is used for drinking and cooking
purposes. Bread iodization was used in the Netherlands as a wartime measure (Matovinovic and Ramalingaswami
1960).

12 The

first such experiment took place in Akron, Ohio in 1917, under the direction of David Marine and O.P.
Kimball. For details see Marine and Kimball (1921), and Carpenter (2005).

74
thyrotoxicosis in others13 . Despite these concerns, the medical consensus was that small amounts
of iodine in the diet were beneficial for the vast majority of an iodine-deprived population, and this
was confirmed by the experimental results in schoolchildren.
Public health authorities in Michigan, one of the worst-afflicted states, held a symposium on
thyroid disease in 1922. The idea of salt iodization (which had been proposed by researchers in
Switzerland, and was first implemented in that country in 1922) was introduced by David Murray
Cowie, M.D. as a cheap and effective means of providing iodine to all population groups, regardless
of social status. As a result, the Iodized Salt Committee was set up, with the mission of investigating
the matter further. The Committee, chaired by Cowie, produced reports on the low iodine content
of drinking water in Michigan and the possibility of effective prevention of goiter though iodized
salt. Its members agreed upon the launch of a statewide educational campaign on goiter and its
prevention though iodized salt, sponsored by the Michigan State Medical Society. The campaign,
launched in 1922, included lectures to physicians and the general public delivered across the state.
The Committee also contacted the state’s salt manufacturers. The salt producers, although
convinced about the public-service character of the project, had initial qualms about its economic
feasibility and profitability; it would be financially impossible to separate the salt intended for the
Michigan market and then add iodine to it. Instead, the Salt Producers Association decided to launch
iodized salt nationwide; they saw the new product as an improved commodity for which there would
be a much larger market -and corresponding profits- than that of Michigan. As a result, in May 1924,
Michigan was the first state to introduce iodized salt14 . The actions of the Michigan salt producers
were important, because Michigan was the largest producer of salt for human consumption in the
country15 .
The salt companies contributed to the educational campaign through aggressive advertising of
the “new salt” throughout the country, in order to create a market for the new product (Markel 1987,
p. 224). Figures 3.2 and 3.3 shows two ads from this period. Figure 3.4 is a copy of a newspaper
clipping from the era.
After Michigan introduced iodized salt, the penetration of the new product statewide was quite
rapid. As mentioned above, salt producers made the new product available nationwide, since this
13 Thyrotoxicosis

might occur as a result of iodization in those individuals that have suffered from long-term iodine
deficiency and whose goiters have become nodular. In such cases, iodine supplementation causes the output of
hormone to jump to toxic levels.

14 Before

that, in April 1923, public authorities in Rochester, New York, introduced iodine in the water supplies
of one reservoir, in what is known as “the Rochester experiment” (Kohn 1975). Subsequent goiter surveys show
an important decrease in incidence. However, it seems unlikely that this decrease was due to the iodization of
the water supply, for the following reasons: first, only one of the reservoirs was iodized. Second, by that time,
iodized salt was available and widely used in Rochester. Third, because of bigger awareness and improved medical
monitoring, doctors were more likely to prescribe iodine supplements to anybody with a palpable goiter.

15 Salt

production takes three forms: evaporated, rock salt, and the production of liquid brine. In 1924 the quantities
produced by these three methods were 2.22, 2.06, and 2.51 million short tons, respectively. Brine was used
exclusively as a feedstock by the chemical industry. According to the Salt Institute (http://www.saltinstitute.org),
as of today, virtually all food grade salt sold or used in the United States is produced by evaporation. This was
the case in 1924 as well (personal communication from Richard Hanneman, president Salt Institute, March 6,
2008). In 1924, Michigan was the largest producer of evaporated salt in the country, accounting for 36% of the
total. The next largest producers were New York (18%) and Ohio (14%) (Katz 1927).

75

Figure 3.2: Ads for Iodized Salt

76

Figure 3.3: Ads for Iodized Salt

77

Figure 3.4: A Newspaper Article Discussion of Iodized Salt

78
was the only way that the project would be financially feasible. The Morton Salt Company, the
largest producer in the country at the time, began selling iodized salt on a nationwide basis in the
fall of 1924

16

. At the same time, public awareness of the problem, especially in those areas that

were afflicted the worst, was gaining momentum. Articles in newspapers and magazines around the
country advocated the use of the new salt for all cooking and eating purposes, making references to
successful goiter prophylaxis in Switzerland17 . Most state health authorities urged the public to use
iodized salt. In advertisements of iodized salt, the new commodity bore the endorsement of state or
national medical associations and educational booklets were provided by the salt companies upon
demand. We have not been able to find precise numbers on the penetration of iodized salt (or other
iodine dietary supplements) for the rest of the country, but all the evidence suggests that the new
product became very popular very quickly, especially in the goiter-belt region, where it mattered the
most. We have many newspaper sources that show the generalized availability and use of iodized
salt from 1924 onwards.
The focus of the salt iodization campaigns was goiter eradication. This is sensible given goiter’s
obvious physical symptoms. It is not clear that the link between iodine deficiency and mental
function was suspected at the time. The decrease in mental retardation that we suspect resulted
from iodization campaigns was a very positive unintended consequence. The cognitive benefits are
the main motivation for modern iodization campaigns (though goiter remains the most obvious sign
of an iodine deficient population).
The pre-existing geographical variation in the prevalence of iodine deficiency, as measured by
goiter rates in recruits, along with the timing of iodization, provide us with a nice natural experiment
with which to look at the long-term impact of iodine deficiency eradication.

3.3.1

Evidence from mortality data

A second source for evidence on the spread of iodized salt comes from data on mortality from thyroid
disease. Mortality data provides a good indicator of iodization because, as discussed above, the
treatment designed to help goiter sufferers sometimes ended up killing them. The doctors and public
health officials who worked for the introduction of iodized salt were aware of the potential danger
of iodine-induced hyperthyroidism, but viewed the danger as minimal. In Europe, the potential
negative side-effects of iodine treatment had been discussed as early as the nineteenth century [see
McClure (1934a) and Carpenter (2005)].
Figure 3.5 shows the annual rate of deaths in the US over the period 1910-1960 due to exophthalmic goiter, which accounted for the overwhelming majority of deaths due to thyroid disease over
16 (Markel

1987). Collusion in the evaporated salt industry was widespread, and Morton acted as the price setter.
Many companies literally made copies of Morton’s price schedule, simply replacing their company letterheads
for that of Morton (Fost 1970). Morton’s decision to iodize salt in 1924 would thus likely have affected a large
percentage of households, both directly and through Morton’s influence on smaller companies.

17 For

example: Lima News (Lima, Ohio), on August 29, 1924, reports that iodized salt is now marketed “thru the
regular grocery trade”. In Appleton Post Crescent (Appleton, Wisconsin), on January 28, 1926, it is mentioned
that “iodized salt is now sold by grocers everywhere, and families can use it instead of ordinary salt”. More
sources are available from the authors upon request.

79

Figure 3.5: US Deaths from exophthalmic goiter, 1920-1960

4.5

4

3.5

3

2.5

2

1.5

1

0.5

0
1900

1910

1920

1930

1940

1950

1960

1970

US Total Goiter Deaths

this period. Exophthalmic goiter is an enlargement of the thyroid accompanied by bulging of the
eyes, which is sign of hyperthyroidism

18

. There is an extremely large rise in the death rate at the

time of iodization: from 2.1 per 100,000 in 1923 to 2.7 in 1924, 3.4 in 1925, and 4.0 in 1926. Deaths
remained elevated for at least a decade. There was also a large gender disparity. In 1926, the death
rate was 1.1 per 100,000 for men 7.0 per 100,000 for women. The population of the United States
in 1926 was 117 million, and so the rise of approximately two deaths per 100,000 people represented
an extra 2,340 deaths in that year. Over the period 1925-1942 there appear to be at least 10,000
excess deaths that resulted from the introduction of iodized salt. We have found little discussion in
the literature of what appears to be a high short-term price the country paid for long-run benefits
resulting from this public health intervention (McClure 1934b).
As would be predicted by the medical evidence that iodine-induced hyperthyroidism is most
common among those with long-standing iodine deficiency, the rise in the death rate was highest,
and persisted the longest, among older age groups. As figure 3.6 shows, deaths in the 25-34 age
category less than doubled from 1921 to 1926, and had fallen below their 1921 level by 1935. In
the 65-74 age category, deaths more than tripled between 1921 and 1926, and were still three times
18 While

some medical dictionaries use the definition stated here, others define exophthalmic goiter as being synonymous with Graves disease, which is the dominant cause of the condition today. It seems clear that the definition
used in text was being applied in the vital statistics data.

80

Figure 3.6: US goiter deaths by age group, 1921-1935

16
14

death rate per 100,000

12
10
8
6
4
2
0
<1

1--4

5--14

15--24 25--34

35-44

45--54 55--64 65--74 75--84

>=85

age group

1921

1924

1926

1929

1932

1935

their 1921 level in 1935. The link between iodine deficiency and the rise in deaths at the time of
iodization is also apparent looking across states. Figure 3.7 shows a scatter plot of rate of simple
goiter among World War I recruits (from data discussed in the next section) and the change in the
mortality rate from thyroid disease over the period 1921-26. Large increases in mortality all took
place in states that had high levels of goiter due to iodine deficiency.
Given the strong evidence that it was salt iodization that caused the rise in thyroid disease
deaths, we can use data on deaths to learn about the timing of iodization. We do not know the
exact time lag between increased iodine consumption and death from thyroid disease in a susceptible
person. However, we do know the timing of iodization in Michigan, so we can at least ask whether
timing of deaths in that state looks similar to other high-goiter areas. Figure 3.8 shows annual
data on thyroid disease death rates for Michigan along with the other 10 states in the upper right
quadrant of Figure 3.7. For visual clarity, we divide the states into two groups, and in each picture
we show data for Michigan for comparison. Compared to the other states in the upper Midwest,
there is some evidence that Michigan may have experienced the surge in deaths earlier, but only by
one year. Compared to the states in the mountains and Pacific west, there is no evidence of any lag
compared to Michigan. Time series for these latter states are very jagged, however, because of their
low populations.

81

Figure 3.7: Changes in thyroid mortality vs. pre-existing goiter rates, 1921-1926

6.00

IN

y = 0.1866x + 1.1416
R2 = 0.5163

exo. goiter mortality rate, 1921–1926

MI

OR
WA

5.00
OH
IL

4.00

WI

MN

NE

3.00

MT

UT

MO
CO

MD
KS

2.00

1.00

PA

FLRI NY TN
CA
NJ SC
NC
VA
DE KY
NH
MA CT
VT
MS
LA

0.00
0.00

5.00

10.00

15.00

20.00

25.00

ME

-1.00

WWI simple goiter rate

3.4

Data Sources

We wish to examine the impact of the iodization of salt in 1924. In order to do so, we need two
pieces of information. First, who was likely to benefit from salt iodization? Second, how did the
outcomes of these individuals change after the iodization of salt in 1924?
As previously described, the presence or absence of endemic iodine deficiency depends on the
iodine content of the soil and ground water. Individuals born in low iodine areas are much more
likely to be iodine deficient than those born in high iodine areas. Iodizing salt will have a large
effect on the former and no effect on the latter. The abrupt switch to iodized salt completes the
natural experiment. Individuals born in high iodine areas should see no effect from salt iodization.
Individuals born in low iodine areas, by contrast, should see effects whose size is related to the
severity of iodine deficiency before the treatment. The latter group provides our treatment group
and the former group our control.
To implement this strategy we need data on the prevalence of iodine deficiency before 1924, as
well as data on outcomes of individuals born in these areas both before and after 1924. Our primary
data sources take advantage of two previously unused surveys of prime age American males in the
early part of the twentieth century; we use data collected during the military drafts of World War I
and World War II.

82

Figure 3.8: Deaths from exophthalmic goiter, Michigan and other states, 1920-1930

exo. goiter death rate, per 1000

10
9
8
MI
WI
OH
IL
MN
IN

7
6
5
4
3
2
1
0
1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930
Year

exo. goiter death rate, per 1000

10
9
8

MI
OR
WA
MT
UT
ID

7
6
5
4
3
2
1
0
1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930

Year

83

Figure 3.9: Sections from Defects Found in Drafted Men

3.4.1

Defects in Drafted Men

For data on the prevalence of iodine deficiency before 1925 we use a volume entitled Defects Found
in Drafted Men, published by the War Department in 1920 (Love and Davenport 1920)19 . Defects
summarizes the results of all the physical exams performed on draftees during World War I for
both accepted and rejected men. Data on prevalence rates per 1000 are recorded for 269 different
medical conditions. The data are regional, organized by units called draft sections. All but the
lowest population states are broken down into multiple sections. Illinois and New York, for example
are broken down into 8 sections. Each section is defined as a collection of counties20 . In total,
Defects has data on 151 separate regions of the country. Figure 3.9 is a map of the US showing the
locations of sections from Defects.
The outcome of interest for our study is the rate of simple goiter, which is a direct result of
iodine deficiency. Simple goiter is relatively common in the data, with a population weighted average
prevalence of about 5 cases per 1000 and a median prevalence of 2.5 per 1000. The prevalence rates
range as high as 29.85 cases per 1000. Though there are no sections with a zero rate of simple goiter,
about one third of the sections have rates of less than 1 per 1000. Figure 3.10 shows the simple
goiter rate at the state level. Rates range from a high of almost 27 per 1000 in Idaho to a low of
0.25 per 1000 in Florida.
The fact that the data are at a finer level of aggregation than the state level is important because
19 Many
20

thanks to Hoyt Bleakley for making us aware of this marvelous book.

Since county borders in the US are relatively static, it is straightforward to map the Defects sections to present
day US counties.

84

Figure 3.10: State level Simple Goiter Rates from Defects Found in Drafted Men

Figure 3.11: Histogram of Section Level Simple Goiter Rates from Defects Found in Drafted Men

0

10

Percent
20

30

40

Distribution of simple goiter across state sections

0

10

20
simple goiter

Source: Love (1920)

30

85

Figure 3.12: Histogram of Population Goiter Rates from Defects Found in Drafted Men

0

10

Percent
20

30

40

Distribution of simple goiter in WWII US Army data

0

10

20

30

simple goiter
Source: Love (1920)

there is significant regional variation within the high goiter states. For example, in the five sections
in Michigan the rates reported inDefects range from over 25 in the Upper Peninsula to less than 10
in Detroit and the surrounding areas.
The incidences of iodine deficiency identified in Defects are almost certainly geographic in origin.
In a paper published in the Journal of the American Medical Association in 1924, J.F. McClendon21
and Joseph C. Hathaway provided measures of the iodine content of drinking water from 69 localities
across the US22 . These measures came from lakes, springs, rivers and wells. Their paper includes
US maps with the low-iodine areas being shaded, and other maps where the high-goiter areas are
shaded (their data on goiter come from the Defects Book). The two shaded areas in the two maps
largely overlap (McClendon 1924). The data on the iodine content of water are not plentiful enough
to be used as alternative measures of iodine deficiency, in addition to the goiter data that we have.
They are useful, however, because the negative correlation between the iodine content measures and
the goiter in their section of origin suggests that goiter variation was due to geographical factors.
Figure 3.13 is a scatterplot relating the log of iodine to the level of simple goiter in the section where
the origin of the water sample belongs (typically a town). The corresponding regression line is:
21 Professor
22 Parts

of Physiologic Chemistry at the University of Minnesota Medical School.

of Iodine per hundred billion parts of drinking water.

86
simple goiter

=

10.854−1.422 ×log iodine
(1.445) (0.355)

(Standard errors in parentheses, N=67)

Figure 3.13: McClendon’s data on the iodine content of water

−10

Goiter in WWI recruits
0
10
20

30

Goiter and iodine content of water in select US localities

0

2

4
6
log of iodine content of water
Simple Goiter

8

10

Fitted values

source: McClendon (1924)

We expect that the effect of iodizing salt will be larger in areas where the prevalence of goiter in
Defects is highest. Since the medical literature suggests that the largest impact of iodine deficiency
is in utero, we would like to identify individuals born in high and low goiter areas both before and
after 1924. Luckily for our purposes, the United States government performed a survey of this
population during the draft for World War II.

3.4.2

World War II Enlistment Data

The World War II enlistment data are from the National Archives and Records Administration
(NARA). The data has its origin in punch cards produced during the enlistment process for the
United States Army, including the Women’s Army Auxiliary Corps. It includes both volunteers and
draftees. After the war these punch cards were converted to microfilm. In 1994 NARA hired the
census department to scan the microfilm into a collection of over 9 million enlistment records from
1938-1946. Though not complete, these records represent the majority of the enlistment into the

87
Army during this period. Of the 1,586 rolls of microfilm, 1,374 (87%) were successfully scanned,
leaving approximately 1.5 million punch cards unrecorded. The missing rolls are not sequential, and
there are no indications that the records are missing in any systematic fashion. In addition to the
missing rolls, several hundred thousand individual records were unreadable.
Though the format of the punch cards changed somewhat over the course of the war, the coding
for basic demographic information was consistent. The demographic fields are name, serial number,
state and county of residence, place and date of enlistment, place and year of birth, race, education,
and marital status. In addition, the particular branch of the Army that the enlistee entered is coded.
One can also infer through the serial number whether the person was drafted. There are no records
for individuals who entered the Navy or Marines.
The sample we have available to us is obviously very large and the timing of the draft is nearly
perfect for our purposes. Limiting our sample to white men, we have data on over 300,000 from
each birth year between 1921 and 1927, giving us extremely complete coverage on both sides of the
1924 salt iodization date. Unfortunately, the data do not include the county of birth, only the state.
We therefore limit our sample to individuals whose birth state is identical to their state of residence
and we assume that the county of residence upon enlistment is the county of birth. This reduces
the sample by less than half, leaving us with almost 2 million records of individuals born between
1921 and 1927.

3.4.3

Test Scores

All enlistees were given the Army General Classification Test (AGCT), a predecessor to the AFQT
that is currently given to enlistees. This test score would be an ideal outcome for our study, since
the primary effect of iodine deficiency in utero is reduced cognitive ability. Unfortunately, the score
is not recorded in the data that we have available(with one notable exception, discussed below). We
can, however, make some crude inferences about the test scores by examining which army branch
the enlistees were assigned to.
Each test taker was assigned a grade of I, II, III, IV, or V, with class I being the highest score
on the test and V the lowest. Jobs within the Army were deemed to require soldiers from different
groups. For example, skilled positions like mechanics tended to get class I or II enlistees, while
lower skill jobs, like cooks tended to get class IV or V enlistees. We do not know the particular job
assignment of each recruit. However, we can identify enlistees who were assigned to the Army Air
Forces (AAF) versus those who were assigned to the Army Ground Forces (AGF). Roughly 14%
of all enlistees were assigned to the AAF over the course of the war, though this proportion varied
from year to year. The year to year variation is described in Table 3.223 .

23 The

year to year variation is largely driven by changes in US participation in the war. In 1941, the US was largely
engaged in the Air War and had less need for ground forces. By 1943 the need for ground forces rose considerably.

Table 3.1: Sample size and probability of joining the AAF by birth year and enlistment semester
semester of
enlistment
40.1

20
43
0.19

21
65
0.12

22
32
0.09

23
10
0.00

40.2

24,849
0.15

28,448
0.13

20,704
0.14

66
0.08

41.1

19,864
0.19

17,851
0.16

15,617
0.20

3,086
0.36

41.2

34,252
0.38

6,732
0.68

6,243
0.61

42.1

74,192
0.16

33,054
0.33

42.2

107,383
0.07

43.1

24

Year of Birth
25
26

27

28

Total
150
0.13

Sample size
% going to AAF

74,067
0.14

Sample size
% going to AAF

15
0.00

56,433
0.19

Sample size
% going to AAF

6,604
0.57

6
0.50

53,837
0.46

Sample size
% going to AAF

12,145
0.45

10,276
0.40

2,897
0.30

85
0.21

132,649
0.25

Sample size
% going to AAF

185,140
0.07

151,019
0.16

44,984
0.59

29,751
0.60

104
0.22

518,381
0.17

Sample size
% going to AAF

20,338
0.06

23,146
0.09

60,682
0.04

155,601
0.02

144,109
0.01

29,310
0.02

96
0.09

433,282
0.02

Sample size
% going to AAF

43.2

10,761
0.07

10,681
0.08

10,449
0.08

15,281
0.06

28,303
0.04

71,648
0.09

66
0.32

147,189
0.08

Sample size
% going to AAF

44.1

14,817
0.04

11,180
0.06

9,397
0.06

9,711
0.05

11,099
0.04

41,627
0.18

30,354
0.26

14
0.14

128,199
0.14

Sample size
% going to AAF

44.2

11,559
0.01

9,163
0.02

7,019
0.02

8,824
0.01

11,071
0.00

16,005
0.01

76,846
0.13

27
0.00

140,514
0.08

Sample size
% going to AAF

45.1

4,803
0.01

4,313
0.01

4,081
0.01

5,022
0.00

6,791
0.00

8,318
0.00

56,995
0.20

30,342
0.07

55
0.09

120,720
0.11

Sample size
% going to AAF

45.2

8,855
0.28

9,751
0.26

9,467
0.22

10,947
0.19

13,280
0.15

16,014
0.11

41,103
0.11

82,226
0.12

9,738
0.37

201,381
0.15

Sample size
% going to AAF

46.1

5,405
0.37

6,694
0.33

6,811
0.30

7,762
0.25

9,426
0.19

11,572
0.14

19,540
0.12

80,613
0.11

66,660
0.27

214,483
0.19

Sample size
% going to AAF

46.2

1,946
0.41

2,143
0.40

2,438
0.39

2,609
0.37

3,292
0.31

3,809
0.22

5,252
0.17

20,767
0.10

50,859
0.19

93,115
0.19

Sample size
% going to AAF

Total

339,067
0.14

348,361
0.12

316,104
0.16

280,783
0.16

260,040
0.10

198,492
0.10

230,252
0.16

213,989
0.11

127,312
0.25

2314400
0.14

Sample size
% going to AAF

88

89
Table 3.1 is a summary of our enlistment data. It gives the total sample size and the percentage of
recruits going to the Airforce for each cohort and each enlistment semester. The number of enlisted
men started increasing in 1942, and it peaked in the second semester of that year, as well as the first
semester of 1943. People born after iodization enlisted in large numbers starting the first semester
of 1943. The proportion of recruits going to the Airforce was particularly low in 1943.
These AAF enlistees were systematically different than other enlistees during the war. There is
ample evidence that the AAF enjoyed preferential assignment of inductees compared to other Army
branches24 .
Table 3.2: Percent of enlistees assigned to the Army Air Forces by year.
Year
1940
1941
1942
1943
1944
1945
1946
Total

3.4.4

Percentage in Air Corps
13.7
32.6
18.8
03.8
10.8
13.9
19.1
14.0

The Battle for the Best Enlistees

The Army Air Forces (AAF), which was still part of the Army during World War II, had a large proportion of jobs that required skilled recruits relative to the Army Ground Forces (AGF). Throughout
the war, the AAF pushed to have a large proportion of the more highly skilled recruits assigned
to the Air Forces. In February 1942, the AAF successfully got the 75 percent rule put into place.
Under this rule, 75 percent of the men assigned to the AAF were to have scored above 100 (the
median score) on the AGCT25 . From this, we can infer that individuals assigned to the AAF during
this period have, on average, higher test scores than those assigned to the AGF. Because the AGCT
is a normalized exam, we can even infer how much higher this average was.
Unsurprisingly, the AGF was not pleased and this rule was not in place for the entire war.
Though lower skilled recruits could easily be used in the infantry, the AGF was concerned about
having a supply of recruits who could become high-quality combat leaders. The AGF successfully
lobbied the War Department to change the rule on August 1, 1942 so that the proportion of above
average men received by the AAF was reduced to 55 percent26 .
The AAF fought back against this change by using a second test, the mechanical aptitude (MA)
24 See,
25 US

for example, Palmer, Wiley and Keast (1948, p.21)

Air Force Historical Study #76, Classification and Assignment of Enlisted Men in the Air Arm 1917-1945, p.

44.
26 ibid,

p.46.

90

Figure 3.14: Percentage of Air Corps Recruits with above average AGCT scores

source: US Air Force Historical Study #76, Classification and Assignment of Enlisted Men in the Air Arm
1917-1945.

test, as a screen for AAF recruits. At first, they simply requested that a higher proportion of men
assigned scored above average on the MA test. This was later formalized. From December 1942
until June 1943, the AAF was supposed to be assigned 55 percent of their new recruits from the
group with scores greater than the mean on both the AGCT and the MA tests. Combining the
two tests was obviously more restrictive than just using one test. In fact only 37.5 percent of all
recruits were above average on both tests27 . This rule was allowed to expire, but it is clear that the
AAF continued to get higher quality of recruits even after the rule formally expired. For example,
for those inducted in 1943, 41.3% of soldiers assigned to AAF were class I or II. This percentage
is higher than the one corresponding to Ground Combat Arms (29.7%) and Services (36.5%) [data
come from Palmer et al. (1948)].
Figure 3.14 is a graph from the War Department showing the percentage of recruits assigned to
the AAF with above average AGCT scores during the period of time that these rule changes were
occurring. During the early part of the graph, the 75 percent rule was clearly in operation. At the
end of July, the abolition of the 75 percent rule can be seen, with the AAF only getting 55 percent
of recruits from the above average group. By September, the AAF had managed to return to the
old proportions via the 55 percent mandate on both the AGCT and MA tests.
Additional evidence on the positive selection of men with high cognitive function into the AAF
comes from an anomaly in the WWII enlistment data. As mentioned above, all new enlistees in
27 ibid,

p.56.

91
the army took the AGCT test, but this test result was not generally recorded on the punch cards
that are the primary source for the enlistment data. However, for the the three months March-May,
1943, AGCT scores were recorded in the fields marked weight for almost all recruits.

28

The fact

that AGCT was coded in this field for some subset of the war is suggested in the documentation.
Observing the actual distribution of values in the weight field confirms this is true for a subset of
observations. Examining observations from enlistments through 1942, the weight field has a mean
of 150 and a standard deviation of 22. For the period of March through May of 1943, the mean is
roughly 100 and the standard deviation it about 20, consistent with normalized AGCT scores.
Figure 3.15 shows a histogram of AGCT scores for recruits entering the Air Corps alongside a
histogram for all other recruits.29 For the non Air Corps group there are 422,516 reported scores with
a mean of 99.4 and a standard deviation of 20.7. This distribution matches the normed distribution
of the AGCT (which were supposed to be mean of 100 and standard deviation of 20). The second
panel shows the histogram of recruits entering the AAF in this period. The mean (126.5) suggests
that the AAF was, at this point, receiving substantially better recruits than the ground forces.
However, as seen in Table 3.1, this was also a period in which very few recruits were entering the
AAF – this histogram is based on only 384 recruits.
The preferential treatment of the AAF lasted until the end of 1943, when the Infantry crisis
broke out. The need for high-quality ground forces grew more acute in 1944 and lasted until the
end of the War, while, at same time, air operations were not as important as in previous years. This
meant that priorities between the AAF and the AGF reversed in favor of the latter, and the Army
classification system was revised to allow for better-quality soldiers to join the Ground Forces30 .
With this classification procedure, as well as with tranfers within the Army commands, the AGF
had an influx of high-quality men, especially in the end of 1944 and afterwards, as opposed to the
Air Force. At the same time, the distribution of recruits among the Army commands changed, and
most of the new inductees were assigned to the Ground Forces.
Through our enlistment records we know whether or not an individual was assigned to the AAF
or the AGF. We know that individuals assigned to the AAF have higher test scores than the average
enlistee for the initial phase of the war. We can also see the selection into the AAF over the course
of the war by looking at the proportion of high school grads assigned to each branch. Table 3.3 and
Figure 3.16 give the proportion of high school graduates in each branch by birth year. Figure 3.17
shows the proportion of high-school graduates by enlistment month, separately for the Airforce and
the rest of the Army.
28 We

are grateful to Joseph Ferrie for making us aware of this data.

29 Because

the recording of AGCT scores in this field was for such a short period of time there is some question as
to whether all enlistment places coded this field the same way. For this reason the histograms are drawn using
only data from enlistment places with over 500 recruits and where the mean of the weight field is between 80 and
120 within the enlistment place. This eliminates less than 11,000 observations and does not substantially change
the distributions.

30 The

New classification procedure was based on “The Physical Profile System”, which became operative in 1944,
and classified recruits into three profiles, according to their ability to withstand strenuous combat conditions.
80% of men assigned to the AGF had to belong to the top profile, whereas only 10% of the AAF recruits came
from the top group.

92

Figure 3.15: AGCT Scores for AAF versus all enlistees

Assigned to AAF

.02
0

.01

Density

.03

.04

Not Assigned to AAF

50

100

150

50

100

150

AGCT Score
Graphs by aircorps

source: WWII enlistment records.

Table 3.3: Percent of high school graduates by birth year and branch.
Birth Year
1920
1921
1922
1923
1924
1925
1926
1927
1928
Total

HS Graduation Rate
Air Corps Ground Forces Total
76.0
42.8
47.5
75.9
41.3
45.6
70.1
39.7
44.5
60.5
43.6
46.3
52.1
39.4
40.7
56.7
35.9
37.9
50.9
36.7
39.0
33.3
38.4
37.9
26.3
40.5
37.0
59.1
40.1
42.8

93

Figure 3.16: High-School Graduation Rates by birth year

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
20

21

22

23

24

25

26

27

28

Year of Birth

aircorps=1

aircorps=0

Total

As expected, Table 3.3 and Figure 3.16show that the younger cohorts were less likely to have a
high-school degree than the older cohorts. It is also clear that there was positive selection into the
Airforce, which became less and less pronounced over the course of the war. By the time the 1927
and 1928 cohorts enlisted, the Ground Forces were getting equally, if not better-qualified recruits
than the Air Forces. Figure 3.17 shows the downward trend in the proportion of AAF recruits who
had high-school diplomas, which becomes steeper towards the end of 1943, when the Infantry crisis
broke out and priority was given to the Ground Forces. We can also see from Figure 3.17 the spike
in the proportion of high-school graduates in the Air Force around February 1942, when the 75%
rule was put into effect. Also, Figure 3.17 shows the temporary decrease in quality of AAF recruits
in the second half of 1942, when the 75% rule was withdrawn. In late 1942 and early 1943, when
the Mechanical Aptitude Test was put into use, the Air Force returned to the preferential-treatment
status it enjoyed before.
It is clear that the Air Corps selected higher quality enlistees for the cohorts born from 1920
until 1926. Luckily, this gives us good coverage of pre and post treatment enlistees. As will be seen
below, the fact that the for the 1927 and 1928 birth cohorts there was not positive selection into the
Air Corps is also consistent with our empirical results.
We will take advantage of the selection of enlistees into the Air Corps in two ways. First, if
iodine deficiency affects cognitive ability we should expect a jump in the relative rate of assignment

94

Figure 3.17: High-School Graduation Rates by enlistment month

1.2

1

0.8

0.6

0.4

0.2

46
10

46
07

46
01
46
04

45
10

45
07

45
04

45
01

44
10

44
07

44
04

44
01

43
10

43
07

43
04

43
01

42
10

42
07

42
04

42
01

41
10

41
07

41
04

41
01

40
07
40
10

40
04

40
01

0

enlistment year-month

aircorps=1

aircorps=0

Total

to the AAF after 1924 in those counties where goiter rates were high in the Defects data. Second,
by exploiting the normal distribution of the AGCT test, and the joint distribution of the AGCT
and MA tests, we can infer the average test scores of individual assigned to the AAF versus those
assigned to the AGF under each of the assignment regimes. We can then assign to each of the
individuals in our data set a test score based on their assignment and date of enlistment31 .

3.5

Identification

Our identification strategy relies on two important sources of variation. First, the impact of the
intervention is a function of the size of the problem in the treated populations. In regions where
goiter was low among men drafted in WWI, we expect little change after iodizing salt. In regions
where goiter was high we expect to see a significant increase in the proportion of men being assigned
to the Army Air Forces after iodizing salt. Second, the rapid introduction of the iodized salt provides
for a clean distinction between the treated and the untreated. We expect to see a sharp difference
between the years before 1924 and the years after. High goiter areas therefore provide the treatment
group and low goiter areas the control group. Years before 1924 are pre-treatment and years after
31 See

Appendix 3.A for the details of these calculations.

95
1924 are post-treatment.
The basic regression specification is:
yi = α +

X

βt [goiter × I(t = birthyear)] + Controls + ǫ

(3.5.0.1)

t6=1924

where yi is one of two outcome variables measured at the individual level. First, yi is a dummy
variable coded one if the individual entered the AAF and zero if they did not. Second, yi is the
implied AGCT score assigned to each individual based on the value of the AAF dummy and their
induction date.
The goiter rate interacted with a set of birth year dummies provides the main coefficients of
interest. The year iodized salt was introduced, 1924, is the excluded category. The pattern of
coefficients on this set of dummies will show how the relationship between the geologically determined
level of iodine deficiency (as measured by goiter rates in WWI enlistees) and cognitive ability (as
measured by the likelihood of entering the AAF) changes over time. The iodization of salt in 1924
should make these coefficients significantly larger (less negative) in the later years.
In all the regressions we include a number of additional controls. We include a full set of section
dummies. We also sometimes include a goiter x year term to control for nutritional trends that
may have been reducing the impact of low iodine soil over time. Since the proportion of enlistees
being assigned to the AAF varies dramatically over the course of the war, we include a full set of
birth year dummies x enlistment year dummies in one specification. This will also control for any
systematic relationship between birth year, enlistment year and the propensity to enter the AAF
which is shared across all members of the cohort. This implicitly provides a control for enlistment
age. In another specification, instead of including birth year dummies interacted with enlistment
year dummies, we include birth year dummies and enlistment month dummies separately32. Results
are consistent across specifications.
Since we have many observations utilizing a single section level data point for the goiter level we
cluster our standard errors at the Defects section level.

3.6

Results

Figure 3.18 is a basic, “first-cut” graphical preview of our results. We plot the probability of joining
the Airforce for each cohort of recruits, by high-goiter and low-goiter group, according to the goiter
level in their section of birth. The high goiter group contains Defects sections which are at the top
25% of the distribution with a cutoff of 5.4 goiter cases per 1000. From Figure 3.18 we see the jump
in the probability of joining the Airforce for the 1925 cohort coming from a high-goiter area, relative
to the same cohort coming from a low-goiter area. The jump is even more pronounced for the 1926
cohort. This is reasonable, if we assume that iodized salt caught up in the market with some lag.
32 For

this specification, we only use those recruits who enlisted from July 1940 to December 1946, because there
were very few men who enlisted in the period January-June 1940.

96

.1

Probability of joining the AAF
.15
.2
.25

.3

Figure 3.18: Probability of joining the Airforce

20

21

22

23

24
25
Year of Birth

High goiter group

26

27

28

Low goiter group

Note that the 1927 and 1928 cohorts only enlisted after the AAF stopped receiving preferential
treatment among Army commands.
A better graphical representation of our results is given by Figure 3.19. Figure 3.19 plots the
average residuals for each cohort - goiter group combination, after running an OLS regression of an
Airforce dummy on enlistment month dummies (for each birth year separately, so we ran 9 regressions
for this graph). Figure 3.19 tells the same story as Figure 3.18, but it’s more convincing, because
it accounts for the effect of enlisting in any particular month over the course of the war. Figure
3.19 shows clearly that for the cohorts born after iodization in a previously high-goiter area, and
enlisting when there was positive selection into the Airforce, there was a jump in their probability of
joining the Airforce, as opposed to those coming from low-goiter areas. One can go as far as saying
that Figure 3.19 also shows the negative selection into the Airforce for the 1928 cohort, who enlisted
during the final phase of the war.

3.6.1

Linear Probability Model of joining the AAF

Table 3.4 displays results estimated as a linear probability model using OLS. The dependent variable
is a dummy indicating whether the individual entered the Army Air Forces (AAF). The main
regressors are the level of goiter in the region where the recruit was born interacted with birth
year dummies. The excluded year is 1924, the year that iodized salt was introduced. In the first

97

Table 3.4: Air Force dummy against initial goiter level interacted with birth year dummies
(1)
(2)
(3)
(4)
Air Force
Air Force
Air Force
Air Force
dummy
dummy
dummy
dummy
goiter X birthyear20
0.00132
0.00081
(0.00069)
(0.00121)
goiter X birthyear21
0.00101
0.00146
0.002
0.00207
(0.00040)* (0.00045)** (0.00067)**
(0.00106)
goiter X birthyear22
0.00058
0.0007
0.00124
0.00111
(0.00067)
(0.00067)
(0.00079)
(0.00102)
goiter X birthyear23
-0.00057
-0.00085
-0.00024
-0.00065
(0.00033)
(0.00048)
(0.00026)
(0.00034)
goiter X birthyear25
0.00378
0.00323
0.00345
0.00303
(0.00110)** (0.00088)** (0.00110)** (0.00084)**
goiter X birthyear26
0.00593
0.00533
0.00527
0.00493
(0.00229)* (0.00168)**
(0.00227)* (0.00187)**
goiter X birthyear27
0.0012
0.00041
0.00021
-0.0002
(0.00100)
(0.00122)
(0.00084)
(0.00068)
goiter X birthyear28
-0.00299
-0.00363
-0.00432
-0.00444
(0.00190)
(0.00266) (0.00162)**
(0.00183)*
goiter X year of birth
-0.00033
-0.0002
(0.00017)
(0.00030)
Constant
0.08622
0.64526
0.05618
0.62685
(0.04280)* (0.03220)**
(0.03680) (0.02576 )**
Birth year X Enlistment
YES
NO
YES
NO
Year Dummies
Birth year dummies
NO
YES
NO
YES
Enlistment month
NO
YES
NO
YES
dummies
Section dummies
YES
YES
YES
YES
Observations
2,275,622
2,274,698
2,275,622
2,274,698
R-squared
0.13
0.13
0.13
0.13
* significant at 5%; ** significant at 1%
All regressions clustered at the state-section level
Regressions (1) and (3) - white men enlisted in the period 1940-1946
Regressions (2) and (4) - white men enlisted in the period July 1940- December 1946

98

Table 3.5: Air Force dummy against high goiter dummy interacted with birth year dummies
(1)
(2)
(3)
(4)
Air Force
Air Force
Air Force
Air Force
dummy
dummy
dummy
dummy
highgoiter X birthyear20
0
-0.0853
0.0067
-0.00908
[0.00000]
[0.05016]
[0.01187]
[0.02315]
highgoiter X birthyear21
0.01991
-0.03943
0.02493
0.01772
[0.00761]**
[0.03582]
[0.01076]*
[0.01773]
highgoiter X birthyear22
0.01346
-0.0286
0.01681
0.00951
[0.01227]
[0.02288]
[0.01260]
[0.01479]
highgoiter X birthyear23
-0.00557
-0.02347
-0.0039
-0.00441
[0.00415] [0.00782]**
[0.00291]
[0.00414]
highgoiter X birthyear25
0.04065
0.04637
0.03897
0.02731
[0.01381]** [0.01389]**
[0.01500]* [0.00934]**
highgoiter X birthyear26
0.08746
0.11092
0.08411
0.07281
[0.02332]** [0.03195]** [0.02618]** [0.01964]**
highgoiter X birthyear27
0.01854
0.05782
0.01351
0.00066
[0.01140] [0.02018]**
[0.00992]
[0.00861]
highgoiter X birthyear28
-0.06017
-0.06687
-0.07621
[0.03161]
[0.02490]**
[0.03106]*
highgoiter X year of birth
-0.00168
-0.01905
[0.00297]
[0.00776]*
Constant
0.0635
0.73807
0.05373
0.62699
[0.04598] [0.05636]**
[0.03712] [0.02578]**
birth year X enlistyear
YES
NO
YES
NO
dummies
birth year dummies
NO
YES
NO
YES
enlist month dummies
NO
YES
NO
YES
Observations
2275622
2274698
2275622
2274698
R-squared
0.13
0.13
0.13
0.13
* significant at 5%; ** significant at 1%
All regressions include state-section level dummies
All regressions clustered at the state-section level
Regressions (1) and (3) - white men enlisted in the period 1940-1946
Regressions (2) and (4) - white men enlisted in the period July 1940- December 1946

99

Figure 3.19: Graph with Residuals

two columns of the table we include goiter trend (defined as the section goiter rate interacted with
year of birth) to control for any secular trends in the incidence of goiter.
There is a clear break before and after the excluded year 1924. The coefficients for 1922 and 1923
are small and insignificant. The coefficients on 1925 and 1926 are positive and highly significant.
The fact that the coefficient rises from 1925 to 1926 is consistent with there being some lag in the
diffusion of iodized salt. The later years, 1927 and 1928 are not significantly different from zero
when the trend is accounted for, although 1928 is negative and significant when the trend is not
included. However, as Table 3.3 shows, in 1927 and 1928 the selection appears to be pushing lower
quality recruits toward the AAF, not higher, so we are not surprised that these years do not match
the 1925 and 1926 coefficients.
The effects are large in high goiter areas. The coefficients on 1925 and 1926 are between 0.003 and
0.005. The highest goiter areas have levels of roughly 30 per 1000 cases. Multiplying the coefficients
times 30 indicates that the highest goiter areas saw a 9-15 percentage point increase in the likelihood
of joining the AAF if they were born after 1924.
In Table 3.5 we separate the sample into high and low goiter sections. The high goiter areas
are those in the top 25% of the distribution with a cutoff of 5.4 goiter cases per 1000. The main
regressors are the high goiter dummies interacted with birth year dummies.
The results echo the earlier results. Individuals from high goiter areas see a 3-11% increase in
the probability that they enter the Air Corps after the iodization of salt in 1924. Once again, the

100

Table 3.6: Logit results of initial goiter level interacted with birth year dummies
(1)
(2)
(3)
(4)
Air Force
Air Force
Air Force
Air Force
dummy
dummy
dummy
dummy
goiter X birthyear20
0.0132
0.00735
0.01556
0.00964
[0.00662]*
[0.01100]
[0.00750]*
[0.01174]
goiter X birthyear21
0.02062
0.01873
0.0224
0.02044
[0.00722]**
[0.00988]
[0.00797]**
[0.01046]
goiter X birthyear22
0.01211
0.00858
0.01329
0.00972
[0.00615]*
[0.00928]
[0.00647]*
[0.00958]
goiter X birthyear23
-0.00147
-0.00554
-0.00088
-0.00497
[0.00288]
[0.00292]
[0.00294]
[0.00296]
goiter X birthyear25
0.03754
0.03297
0.03694
0.0324
[0.01276]** [0.00842]** [0.01248]** [0.00828]**
goiter X birthyear26
0.03905
0.03527
0.03787
0.03412
[0.01746]* [0.01237]**
[0.01690]* [0.01219]**
goiter X birthyear27
0.00709
0.00037
0.00532
-0.00134
[0.01138]
[0.00776]
[0.01062]
[0.00740]
goiter X birthyear28
-0.0211
-0.02544
-0.02347
-0.02773
[0.01000]*
[0.01356]
[0.00982]*
[0.01314]*
goiter X year of birth
-0.00059
-0.00057
[0.00031]
[0.00023]*
Birth year X Enlistment
YES
NO
YES
NO
Year Dummies
Birth year dummies
NO
YES
NO
YES
Enlistment month
NO
YES
NO
YES
dummies
Section dummies
YES
YES
YES
YES
Constant
-2.20675
-0.59892
-2.18858
-3.42273
[0.59722]**
[0.18681]**
[0.60949]**
[0.20194]**
Observations
2275622
2274698
2275622
2274698
* significant at 5%; ** significant at 1%
Standard errors in brackets, clustered at the state-section level
Regressions (1) and (3) - white men enlisted in the period 1940-1946
Regressions (2) and (4) - white men enlisted in the period July 1940- December 1946

effect is bigger for 1926 than for 1925. Given that the average rate of assignment to the Air Corps
was roughly 14% for the entire sample, this represents a large effect.

3.6.2

Logit results

We also ran logit models of the probability of joining the AAF, using the same right-hand-side
variables as in the linear probability model of the previous section. Logit results are displayed in
Tables 3.6 and 3.7. Table 3.6 lists results of the interaction of goiter in the Defects section of origin
and birth year dummies, whereas Table 3.7 lists results of the interaction of birth year dummies and
a dummy variable for belonging to a high-goiter group. Regardless of the exact specification, we

101

Table 3.7: Logit results of high-goiter group dummy interacted with birth year dummies
(1)
(2)
(3)
(4)
Air Force
Air Force
Air Force
Air Force
dummy
dummy
dummy
dummy
highgoiter X birthyear20
-0.25107
-0.42923
0.16764
-0.01719
[0.10312]*
[0.19089]*
[0.09757]
[0.19335]
highgoiter X birthyear21
0.02616
-0.09918
0.34019
0.20985
[0.10400]
[0.15833]
[0.10696]**
[0.16080]
highgoiter X birthyear22
0.01727
-0.10447
0.22662
0.10155
[0.09485]
[0.13124]
[0.09843]*
[0.13312]
highgoiter X birthyear23
-0.09949
-0.14021
0.00519
-0.0372
[0.03256]**
[0.04119]**
[0.03295]
[0.04106]
highgoiter X birthyear25
0.60413
0.4365
0.49945
0.33349
[0.16388]** [0.10208]** [0.16002]** [0.10131]**
highgoiter X birthyear26
0.8798
0.74234
0.67044
0.53632
[0.18410]** [0.12199]** [0.17565]** [0.11999]**
highgoiter X birthyear27
0.54155
0.36096
0.22752
0.05193
[0.12685]**
[0.08787]**
[0.11617]
[0.08373]
highgoiter X birthyear28
0.13087
-0.01788
-0.28785
-0.42992
[0.13545]
[0.20632]
[0.14515]*
[0.20574]*
highgoiter X year of birth
-0.10468
-0.10301
[0.00495]**
[0.00334]**
Birth year X Enlistment
YES
NO
YES
NO
Year Dummies
Birth year dummies
NO
YES
NO
YES
Enlistment month
NO
YES
NO
YES
dummies
Section dummies
YES
YES
YES
YES
Constant
-2.40148
0.41482
-3.36703
-2.05741
[0.59429]**
[0.12229]**
[0.72408]**
[0.13248]**
Observations
2275622
2274698
2275622
2274698
* significant at 5%; ** significant at 1%
All regressions include state-section level dummies
All regressions clustered at the state-section level
Regressions (1) and (3) - white men enlisted in the period 1940-1946
Regressions (2) and (4) - white men enlisted in the period July 1940- December 1946

102

Table 3.8: Marginal Probabilities of logit results of high-goiter group dummy interacted with birth
year dummies
[1]
[2]
[3]
[4]
variable
dy/dx
dy/dx
dy/dx
dy/dx
highgoiter X birthyear20*
-0.02077
-0.03077
0.016192
-0.00144
[-2.68]
[-2.65]
[1.62]
[-0.09]
highgoiter X birthyear21* 0.002398
-0.00805
0.03502
0.019134
[0.25]
[-0.65]
[2.82]
[1.21]
highgoiter X birthyear22* 0.001578
-0.00846
0.022386
0.00889
[0.18]
[-0.83]
[2.11]
[0.73]
highgoiter X birthyear23*
-0.0087
-0.01119
0.000472
-0.00309
[-3.17]
[-3.56]
[0.16]
[-0.92]
highgoiter X birthyear25* 0.068926 0.043587 0.054817 0.031998
[3.01]
[3.63]
[2.63]
[2.9]
highgoiter X birthyear26* 0.110453 0.083019 0.078164 0.055527
[3.66]
[4.72]
[3.08]
[3.68]
highgoiter X birthyear27* 0.060258
0.034958
0.022516
0.004463
[3.54]
[3.59]
[1.8]
[0.61]
highgoiter X birthyear28* 0.012494
-0.0015
-0.02339
-0.03061
[0.92]
[-0.09]
[-2.21]
[-2.47]
(*) dy/dx is for discrete change of dummy variable from 0 to 1
Standard errors of z statistic in brackets

always see a jump in the coefficients for those cohorts born in 1925 and 1926, after salt iodization.
Table 3.8 lists the marginal probabilities from the logit specifications where the sample is separated in high- and low- goiter areas. The marginal probabilities are very similar to those derived
with the linear probability model. Recruits born in 1925 or 1926 in previously high-goiter areas are
3-11 percentage points more likely to join the Airforce.

3.6.3

Falsification exercise with other diseases

As a falsification exercise we ran regressions using data on 58 defects other than goiter, which were
also reported in Defects. For each defect, we computed the same linear probability model of joining
the airforce as the one in column (3) of Table 3.4. Table 3.9 is a summary of the values for the
coefficients corresponding to the interaction of birth year dummies with the defect prevalence at the
section level. Finding significant coefficients for cohorts born in 1925-1926, similar to those of Table
3.4, would cast doubt on our results from the previous regressions.
As can be seen from Table 3.9, the time pattern for high goiter regions is replicated exactly
(positive and significant coefficient for those born in 1925 and 1925) in only six of 58 defects:
exophthalmic goiter, hemophilia, multiple sclerosis, acromegaly, tabes dorsalis, and curvature of the
spine. Exophthalmic goiter is a more acute version of simple goiter, which we have been using in our
previous specifications. The fact that its coefficients are similar to those of simple goiter is a good
validity check of our results. Hemophilia, multiple sclerosis, acromegaly, and tabes dorsalis are very

103

Table 3.9: Summary of coefficients for falsification exercise with all defects

Disease X born in 1922

Positive and significant
coefficient
Diseases
Total
leukemia, general dis- 5
eases (other),
tobes
dorsalis, paralysis (unknown
location
and
cause), neuro-circulatory
asthenia
simple goiter, leukemia, 4
tobes dorsalis, neurocirculatory asthenia
leukemia
1

Disease X born in 1923

obesity, monoplegia

Disease
1925

X

born

in

Disease
1926

X

born

in

simple goiter, exoph- 8
thalmic
goiter,
curvature of the spine,
acromegaly,
ductless
glands,
hemophilia,
tabes dorsalis, multiple
sclerosis
simple goiter, exoph- 8
thalmic goiter, curvature
of the spine, acromegaly,
hemophilia, alcoholism,
tobes dorsalis, multiple
sclerosis

Disease X born in 1920

Disease X born in 1921

2

Disease X born in 1927

syphilis,
chorea

huntington’s 2

Disease X born in 1928

dysentery,
pelagra, 9
syphilis,
gonococcus,
cancer, tumors benign,
arthritis,
huntington’s
chorea, hysteria

Negative and Significant
coefficient
Diseases
Total
diabetes mellitus, para- 4
plegia, neurasthenia, tic

pelagra,
idcarriers,
syphilis, addison’s disease, paraplegia, tic
mycosis,
pelagra,
syphilis,
rickets,
tumors benign, addison’s
disease, anemia, tic
dysentery, pelagra, tb
suspected, rickets, tumors benign, arthritis,
diabetes mellitus, anemia, muscular rheumatism, huntington’s chorea
pelagra, syphilis, gonococcus, arthritis, paralysis (unknown location
and cause), huntington’s
chorea

6

8

10

6

dysentery,
pelagra, 10
syphilis,
gonococcus,
arthritis,
gigantism,
drug addiction, paralysis
(unknown location and
cause),
huntington’s
chorea, hysteria
obesity, purpura, drug 8
addiction,
herniplea
and apoplexy, paralysis
(unknown location and
cause),
neurasthenia,
hysteria, migraine
simple goiter, exoph- 10
thalmic goiter, idother,
curvature of the spine,
obesity, alcoholism, drug
addiction, tobes dorsalis,
neurasthenia, neurosis

104

Table 3.10: Falsification Exercise with Different Diseases
(1)
(2)
(3)
(4)
Simple Exophthalmic Dysentery
Epilepsy
Goiter
Goiter
Air Force
Air Force Air Force Air Force
dummy
dummy
dummy
dummy
disease X birthyear20
0.00132
-0.00009
0.07478
-0.00252
[0.0007]
[0.002]
[0.0826]
[0.0030]
disease X birthyear21
0.002
0.00339
0.01929
-0.00188
[0.0007]**
[0.002]
[0.0748]
[0.004]
disease X birthyear22
0.00124
0.00004
0.02241
-0.00456
[0.0008]
[0.002]
[0.0910]
[0.003]
disease X birthyear23
-0.00024
-0.00016
-0.05137
0.00055
[0.0003]
[0.0006]
[0.0214]*
[0.0007]
disease X birthyear25
0.00345
0.01293
-0.11918
-0.00299
[0.0011]**
[0.0032]**
[0.0745]
[0.0027]
disease X birthyear26
0.00527
0.01852
-0.26338
0.00131
[0.0023]*
[0.0051]** [0.0950]**
[0.0042]
disease X birthyear27
0.00021
0.00108
-0.00325
-0.00136
[0.0008]
[0.0025]
[0.0465]
[0.0016]
disease X birthyear28
-0.00432
-0.01221
0.275
0.00175
[0.0016]**
[0.0045]**
[0.1253]*
[0.0055]
Constant
0.05618
0.05447
0.05702
0.06414
[0.0368]
[0.0377]
[0.0365]
[0.0376]
Observations
2275622
2275622
2275622
2275622
R-squared
0.13
0.13
0.13
0.13
* significant at 5%; ** significant at 1%
All regressions clustered at the state-section level
All regressions include state-section level dummies
All regressions include section fixed effects, and birth year x enlistment year fixed effects.

rare and many sections report no cases of these defects. Curvature of spine correlates highly with
goiter and may be related to iodine deficiency. The coefficient for ductless glands jumps for the 1925
cohort, but remains insignificant afterwards. Ductless glands are directly related to the function of
the endocrine system, which is affected by the lack of iodine. Surprisingly enough, the coefficient for
alcoholism jumps significantly for the 1926 cohort, whereas it is negative and significant for the 1928
cohort, similarly to the coefficient for goiter. Alcoholism had a much lower prevalence than goiter,
however (its maximum value was 1.82, compared to 29.85 for goiter), and it was absent (or, at least,
not recorded) in one third of the sections. Interestingly, cretinism, which is a consequence of severe
iodine deficiency, has the same pattern of coefficients as goiter, though the jump is not significant.
In Table 3.10 we report the results for epilepsy and dysentery alongside our baseline simple
goiter results. We also include the regression results for exophthalmic goiter, as a second measure
of goiter. We choose to report epilepsy because it is a hereditary defect and does not depend on
geographical factors in the way that iodine deficiency does. We also included dysentery in this

105
falsification exercise, because it is a defect affected by general sanitation and health conditions.
Exophthalmic goiter coefficients closely follow those of simple goiter. For epilepsy, no coefficient
comes in significantly. In case of dysentery, we get a big negative jump for people born in 1926,
whereas we get a big positive coefficient for 1928. We are still unclear why we are getting such
big and changing coefficients, but we note that almost 60% of sections had no reported cases of
dysentery.

3.7

Interpretation

How large are these effects? The highest goiter places have goiter rates of 30. The coefficients (from
Table 3.4) suggest that about 15% more recruits from the highest goiter regions go into the AAF
after salt iodization. From Appendix 3.A we know that the Air Force recruits have, on average 9
point higher AGCT scores (almost a half a standard deviation). The average increase for the recruits
from that section is therefore 0.15 time 9 points, or 1.35. So the average cognitive ability in the
section goes up by greater than one twentieth of a standard deviation.
Examining the high-low regressions yields similar figures. In the high goiter group, we have a
5-10 percent higher assignment rate to the AAF after salt iodization. Ten percent times 8 AGCT
points results in 0.8 points higher on average, or about a twentieth of a standard deviation. This
implies a twentieth of a standard deviation increase in cognitive ability for 25 percent of the US
population.

3.8

Conclusion

The results suggest relatively modest effects at the population level in the US. However, for the one
half a percent of US residents with a serious goiter the impact was likely quite large. For a larger
population, more modest, but still positive impacts were seen. Our results show that there were
measurable cognitive benefits from iodizing salt that went beyond the obvious effect of reducing
goiter in the US.
These results must be seen in the context of many, many other health interventions that were
happening at the same time. Ten interventions of the magnitude of iodizing salt will generate a very
large and noticeable effect on overall cognitive ability.
Many countries have goiter rates that are much larger than those seen in the US. The effect of
successful iodization programs in these regions would likely be larger than we see in the US.

106

3.A

Average AGCT scores for the AAF versus the AGF

From February 1942 until August 1, 1942 the AAF was assigned 75 percent of their men from those
scoring higher than average on the AGCT. The test was normed to have a mean of 100 and a
standard deviation of 20. Assuming the test is distributed N(100,20), conditional on scoring above
the median, the mean score is 116. Conditional on scoring below the median, the mean score is 84.
This implies that the AAF has an average test score that is a 75/25 percent weighted average of
these two scores for an average score of 108.
If the AAF score average were above 100, this implies that the average AGF recruit was below
100. During the war, the total proportion of enlistees from each birth cohort assigned to the AAF
varied from 10 to 16 percent. If, for example, the AAF took 10% of all enlistees with an average
test score of 108, this implies that the AGF got the other 90% and that this group has an average
AGCT of roughly 99.
On August 1, 1942 the 75 percent rule was rescinded and the proportions went to 55/45. Similar
to the calculations above this implies an average AGCT for the AAF of 102 during this period.
From December 1942 until June 1943 the AAF received 55 percent of their recruits from the
pool with above average scores on both the AGCT and the ME test. We are also told that only
37.5 percent had above average scores on both. If we assume that the MA and AGCT tests are
jointly normally distributed with a correlation of 0.70 we obtain a joint distribution where 37.5 of
the observations are above the average on both tests. Among the pool of observations where both
test scores are above average, the average AGCT score should be 0.91 standard deviations above
the mean, corresponding to a test score of roughly 118. The pool of observations with either test
below average has an average AGCT score 0.55 below the mean or roughly 89. Therefore the AAF
recruits on average had an AGCT score 0.55*118+0.45*89 ˜105 during this period.

3.B

Interpretation of the coefficients:
What was the IQ increase following iodization?

We are assuming that the IQ of recruits from non-goitrous areas was distributed normally, with
mean µ = 100 and standard deviation σ = 15. The IQ distribution of recruits from goitrous areas
was also distributed normally, with a standard deviation also equal to σ = 15. We are looking for
the mean of the IQ distribution of recruits from goitrous areas. Call that x. Iodization shifted the
IQ distribution of recruits from goitrous areas to the right, increasing their mean IQ to 100 (Figure
3A-1).
Let α be the % of enlistees going to AAF.
Let p be the % of AAF enlistees who scored over the median, which was 100, in the AGCT.
This is based on Army rules, e.g. p=75% of enlistees in the AAF had to have scores over 100 in the
AGCT (later p=55%, then p went back up). Suppose that the same percentage of AAF enlistees

107

Figure 3A-1: Shift in the IQ distribution of high-goiter areas after iodization

had IQ scores over the mean. The mean IQ of the population prior to iodization, assuming that
25% of recruits came from goitrous areas, is: 0.75 · 100 + 0.25 · x.
Let e be the increase in the probability of joining the AAF for recruits from goitrous areas
following iodization. This is the effect that we estimate in our regressions.
Since e is the effect of iodization, then prior to iodization, a percentage z = α − e of recruits
from goitrous areas went to the AAF.
Let r be the % of enlisted men scoring above the mean IQ (= 0.75 · 100 + 0.25 · x), who joined
the AAF.
Then,

1−p
p

· r is the % of enlisted men scoring below the mean IQ, who joined the AAF.

By definition, the following holds:
1
2

·r+

1
2

·

1−p
p

· r = α ⇒ r = 2 · α · p and

1−p
p

· r = 2 · α · (1 − p)

[For example, if 15% of enlistees go to the AAF and the 75%-25% rule holds, then 22.5% of the
enlistees whose IQ was higher than average, and 7.5% of enlistees whose IQ was below average went
to the AAF.]
Let y be the percentage of people from iodine-deficient regions, who had IQ higher than average.
A percentage r of those people joined the AAF (where r = 2 · α · p). Similarly, (1 − y) % of people
from iodine-deficient regions had IQ’s lower than average, and
went to the AAF.
It has to be the case that:

1−p
p

· r = 2 · α · (1 − p) of these people

108
y · r + (1 − y) ·

1−p
p

·r =z ⇒y =

1
2

−

e
(2p−1)·2·α .

[For example, if 15% of recruits went to the AAF, the 75% rule was in effect, and iodization
increased the probability of joining the AAF by 5 percentage points, then approximately 17% of
recruits from goitrous areas had IQ levels that were higher than average, and 22.5% of them went
to the AAF.]
So, we need to find the mean x of the normal distribution with σ = 15, for which y % of
observations are above 0.75 · 100 + 0.25 · x. In essence, we are looking for a fixed point.
Let w be the critical value in the standard normal distribution. In other words, y % of values
in the standard normal distribution are over w. (In stata, w = invnorm(1 − y), in excel w =
norm sin v(1 − y)).
The mean of the IQ distribution for goitrous areas prior to iodization will be:
x = 0.75 · 100 + 0.25 · x − 15 · w ⇒ x = 100 − 20 · w.
[Using the above example, it turns out that the mean of the IQ distribution in high-goiter areas
was x = 80.65. In other words, iodization increased average IQ in iodine-deficient areas by almost
20 points, which corresponds to an increase in average IQ by 1.3 standard deviations.]

Bibliography
Acemoglu, Daron and Simon Johnson, “Disease and Development: The Effect of Life Expectancy on Economic Growth,” Journal of Political Economy, 2007. forthcoming.
Almond, Douglas, “Is the 1918 Influenza Pandemic Over? Long-Term Effects of In Utero Influenza
Exposure in the Post-1940 U.S. Population,” Journal of Political Economy., August 2006, 114,
672–712.
Bircher, H., Der endemische Kropf und seine Beziehung zur Taubstummheit und Cretinismus,
Basel, Switzerland: B. Schwabe, 1883.
Bishai, David and Ritu Nalubola, “The History of Food Fortification in the United States: Its
Relevance for Current Fortification Efforts in Developing Countries,” Economic Development
and Cultural Change, October 2002, (51), 3753.
Bleakley, Hoyt, “Disease and Development: Evidence from Hookworm Eradication in the American South,” Quarterly Journal of Economics. Forthcoming, 2006.
Bleakley, Hoyt, “Disease and Development: Evidence from Hookworm Eradication in the American South,” Quarterly Journal of Economics, February 2007, 122 (1), 73–117.
Bleichrodt, N and M P Born, “A meta-analysis of researh on iodine and its relationship to
cognitive development,” in John B. Stanbury, ed., The damaged brain of iodine deficiency, New
York: Cognizant Communication, 1994, pp. 195–200.
Bürgi, Hans, Zeno Supersaxo, and Beat Selz, “Iodine deficiency diseases in Switzerland one
hundred years after Theodor Kocher’s survey: a historical review with some new goitre prevalence data,” Acta Endocrinologica (Copenh), 1990, 123, 577–590.
Carpenter, Kenneth J., “David Marine and the Problem of Goiter,” Journal of Nutrition, April
2005, 135, 675–680.
Case, Anne and Christina Paxson, “Stature and status: Height, ability, and labor market
outcomes,” NBER Working Paper 12466, 2006.
and

, “Stature and status: Height, ability, and labor market outcomes,” Journal of

Political Economy, 2008, 116 (3), 499–532.
109

110
Connolly, R. J., “An increase in thyrotoxicosis in southern Tasmania after an increase in dietary
iodine,” The Medical Journal of Australia, 1971, 1 (24), 1268–1271.
Curtis, George M. and M. Been Fertman, “Iodine in Nutrition,” in “Handbook of Nutrition,”
second ed., American Medical Association, 1951, chapter 6, pp. 111–135.
de Benoist, Bruno, Maria Andersson, Ines Egli, Bahi Takkouche, and Henrietta Allen,
eds, Iodine status worldwide: WHO Global Database on Iodine Deficiency, Geneva, Switzerland:
World Health Organization, 2004.
Delange, F., “Iodine Deficiency as a cause of brain damage,” Postgraduate Medical Journal, 2001,
77, 217–220.
England, Paula and Barbara Kilbourne, “Occupational Measures from the Dictionary of Occupational Titles for 1980 Census Detailed Occupations [Computer file]. Richardson, TX,”
Inter-university Consortium for Political and Social Research [distributor]. Ann Arbor, MI,
1988.
Federal Statistical Office, Swiss Population Census 1970.
Feyrer, James D., Dimitra Politi, and David N. Weil, “The Economic Effects of Micronutrient
Deficiency: Evidence from Salt Iodization in the United States,” manuscript, Brown University,
2008.
Fleischer, Michael, Richard M. Forbes, Robert C. Harriss, Lennart Krook, and Joe
Kubota, “Iodine,” in “The relation of selected trace elements to health and disease,” Vol. 1 of
Geochemistry and the Environment, Washington, D.C.: National Academy of Sciences, 1974,
chapter 3, pp. 26–28.
Fost, Carolyn Ann, “The Salt Industry: A Case Study in the Evaluation of Public Policy,”
unpublished Ph.D. dissertation, Southern Illinois University, 1970, pp. 49–50.
Heckman, James and Edward Vytlacil, “Identifying the role of cognitive ability in explaining
the level of and change in the return to schooling,” The Review of Economics and Statistics,
2001, 83 (1), 1–12.
ICCIDD, UNICEF and WHO, Assessment of Iodine Deficiency Disorders and Monitoring their
Elimination WHO/NHD/01.1, Geneva, Switzerland: World Health Organization, 2001.
Imbens, Guido W. and Thomas Lemieux, “Regression discontinuity designs: A guide to practice,” Journal of Econometrics, February 2008, 142 (2), 615–635.
Katz, Frank J., Mineral Resources of the United States 1924, Department of Commerce, Washington: US Government Printing Office, 1927.

111
Kelly, F.C. and W.W. Snedden, “Prevalence and Geographical Distribution of Endemic Goitre,”
in “Endemic Goitre” number 44. In ‘WHO Monograph Series.’, Geneva, Switzerland: World
Health Organization, 1960, p. 27.
Kocher, Theodor, Vorkommen und Vertheilung des Kropfes im Kanton Bern, Berne, Switzerland:
KJ Wyss, 1889.
Kohn, Lawrence A., “Goiter, Iodine and George W. Goler: the Rochester Experiment,” Bulletin
of the History of Medicine, 1975, 49, 389–399.
Koutras, Demetrios A., “Trace Elements, Genetic and Other Factors,” in John B. Stanbury
M.D. and Basil S. Hetzel, eds., Endemic Goiter and Endemic Cretinism, John Wiley and Sons,
1980, chapter 13, pp. 255–268.
, Josip Matovinovic, and Robert Vought, “The Ecology of Iodine,” in John B. Stanbury
M.D. and Basil S. Hetzel, eds., Endemic Goiter and Endemic Cretinism, John Wiley and Sons,
1980, chapter 9, pp. 185–195.
Langer, P., “History of Goitre,” in “Endemic Goitre” number 44. In ‘WHO Monograph Series.’,
Geneva, Switzerland: World Health Organization, 1960, pp. 9–25.
Love, Albert G. and Charles B. Davenport, Defects Found in Drafted Men. Statistical Information Compiled from the Draft Records Showing the Physical Condition of the Men Registered
and Examined in Pursuance of the Requirements of the Selective-Service Act., Washington,
D.C.: Government Printing Office, 1920.
Lucas, Adrienne, “Economic Effects of Malaria Eradication: Evidence from the Malarial Periphery,” manuscript, Brown University, 2006.
Marine, David and O.P. Kimball, “The prevention of simple goiter in man,” Journal of the
American Medical Association, October 1, 1921, 77 (14), 1068–1070.
Markel, Howard, ““When it rains, it pours”:Endemic Goiter, Iodized Salt and David Murray
Cowie, MD,” American Journal of Public Health, 1987, 77 (2), 219–229.
Matovinovic, J. and V. Ramalingaswami, “Therapy and Prophylaxis of Endemic Goitre,”
in “Endemic Goitre” number 44. In ‘WHO Monograph Series.’, Geneva, Switzerland: World
Health Organization, 1960, pp. 385–410.
McClendon, J. F., “Inverse relation between Iodin in Food and Drink and Goiter, Simple and
Exophthalmic,” Journal of the American Medical Association, May 1924, 82 (21), 1668–1672.
McClure, Roy D., “Thyroid surgery as affected by the generalized use of iodized salt Salt in an
Endemic Goitre Region - Preventive Surgery,” Annals of Surgery, November 1934, 100 (5),
924–932.

112
, “Thyroid Surgery In Southern Michigan As Affected By The Generalized Use Of Iodized Salt,”
Journal of the Michigan State Medical Society, February 1934, 33, 58–62.
Miguel, Edward and Michael Kremer, “Worms: Education and Health Externalities in Kenya,”
NBER Working Paper 8481, 2001.
Palmer, Robert R., Bell I. Wiley, and William R. Keast, The Procurement and Training
of Ground Combat Troops United States Army in World War II: The Army Ground Forces,
Washington, D.C.: Center of Military History, United States Army, 1948.
Politi, Dimitra, “The Impact of Iodine Deficiency Eradication on Schooling: Evidence from the
Introduction of Iodized Salt in Switzerland,” manuscript, Brown University, 2009.
Prévost, Jean-Louis and A.C. Maffoni, Atti Acad. Sci. med. Torino, 1846, 2, 453.
Sachs, Jeffrey, “Institutions don’t rule: direct effects of geography on per capita income,” NBER
Working Paper 9490, February 2003.
Scrimshaw, Nevin S., “Malnutrition, Brain Development, Learning and Behavior,” Nutrition
Research, 1998, 18 (2), 351–379.
UNICEF, Sustainable Elimination of Iodine Deficiency, UNICEF, May 2008.
Watson, Tara, “Public Health Investments and the Infant Mortality Gap: Evidence from Federal
Sanitation Interventions on U.S. Indian Reservations,” Journal of Public Economics, September
2006, 90, 1537–1560.
Weil, David N., “Accounting for The Effect of Health on Economic Growth,” Quarterly Journal
of Economics, February 2007, 122 (3), 1265–1306.
Wespi, H. J., “Abnahme der Taubstummheit in der Schweiz als Folge der Kropfprophylaxe mit
jodiertem Kochsalz,” Schweizerischen Medizinischen Wochenschrift, 1945, 75 (28), 625.
, “40 Jahre Kropfprophylaxe mit jodiertem Salz,” Bulletin des Eidg. Gesundheitsamtes, 1962,
2 (Suppl. B), 1–4.
World Health Organization, “Iodine,” in “Trace elements in human nutrition and health,”
Geneva, Switzerland: World Health Organization, 1996, pp. 49–71.