The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Federal Reserve Bank
of Chicago
First Quarter 2008
Economic.
perspectives
2
Obesity, disability, and the labor force
Kristin F. Butcher and Kyung H. Park
17
Avoiding a meltdown: Managing the value
of small change
Frangois R. Velde
29
Corruption and innovation
Marcelo Veracierto
RESLAKCm LibKARY
Federal Reserve Bank
of St. Louis
MAR 1 0 2008
Economic .
perspectives
President
Charles L. Evans
Senior Vice President and Director of Research
Daniel G. Sullivan
Research Department
Financial Studies
Douglas Evanoff, Vice President
Macroeconomic Policy Research
Jonas Fisher, Economic Advisor and Team Leader
Microeconomic Policy
Daniel Aaronson, Economic Advisor and Team Leader
Payment Studies
Richard Porter, Vice President
Regional Programs
William A. Testa, Vice President
Economics Editor
Anna L. Paulson, Senior Financial Economist
Editor
Helen O’D. Koshy
Associate Editors
Kathryn Moran
Han Y. Choi
Graphics
Rita Molloy
Production
Julia Baker
Economic Perspectives is published by the Research
Department of the Federal Reserve Bank of Chicago. The
views expressed are the authors’ and do not necessarily
reflect the views of the Federal Reserve Bank of Chicago
or the Federal Reserve System.
© 2008 Federal Reserve Bank of Chicago
Economic Perspectives articles may be reproduced in
whole or in part, provided the articles are not reproduced
or distributed for commercial gain and provided the source
is appropriately credited. Prior written permission must
be obtained for any other reproduction, distribution,
republication, or creation of derivative works of Economic
Perspectives articles. To request permission, please
contact Helen Koshy, senior editor, at 312-322-5830 or email
Helen.Koshy@chi.ffb.org.
Economic Perspectives and other Bank
publications are available on the World Wide Web
at www.chicagofed.org.
£7
x Chicago fed. org
ISSN 0164-0682
Contents
First Quarter 2008, Volume XXXII, Issue 1
2
Obesity, disability, and the labor force
Kristin F. Butcher and Kyung H. Park
Men of prime working age have increased their non-employment rates over the past 30 years, and
disability rates have also increased. Many have noted that this increase has happened against a
backdrop of generally improving health in the U.S. population. However, obesity has increased
substantially over this period. The authors find that changes in the characteristics of male workers—
including age, race, ethnicity, and obesity levels—can explain a large portion (around 40 percent)
of the increase in non-employment.
17
Avoiding a meltdown: Managing the value of small change
Francois R. Velde
To prevent a shortage of small change, the U.S. Department of the Treasury recently prohibited the
melting and exportation of pennies and other coins. The problem arises because pennies and nickels
are made of inappropriately expensive material, and there is or soon will be a profit to be made from
transferring their content to alternative uses. The author provides a historical context for the problem
of small change and discusses possible remedies.
29
Corruption and innovation
Marcelo Veracierto
In this article, the author illustrates how corruption can affect an industry’s rate of innovation.
An interesting result of the analysis is that, under certain parameter ranges, small increases in the
penalties to corruption or the effectiveness of detection can result in large increases in product
innovation.
Obesity, disability, and the labor force
Kristin F. Butcher and Kyung H. Park
Introduction and summary
In this article, we investigate how the rise in obesity
over the past three decades is related to non-employment. In recent years, unemployment rate figures—
joblessness among those actively seeking work—have
been low by historical standards. At the same time,
however, there has been a rise in the fraction of men
who are not actively seeking work.1 The labor force
participation of men of prime working age is low by
historical standards, and this has coincided with an
expansion in the Social Security Disability Insurance
(SSDI) program.
A number of researchers studying the increase in
men’s non-employment have pointed out that it takes
place against a backdrop of improving health (Juhn,
Murphy, and Topel, 2002; and Autor and Duggan, 2003).
However, these improvements in health are typically
measured by mortality rates, which have been declining over time (Cutler and Richardson, 1997). Obesity
rates, on the other hand, have climbed dramatically
during the past 30 years. To put the increase in perspective, the median male in 2002 would have been heavier
than 75 percent of the male population in 1976, using
a body mass index (BMI) distribution.
There are a number of reasons that increases in
obesity might be linked to decreases in employment.
Increases in obesity might affect the ability to work—
for example, obese people are more likely than others
to have health problems—or the willingness to work,
depending on the availability of alternatives to working. We call these “supply side” factors—those factors
that affect whether or not an individual is willing and
able to take a job. There may also be “demand side”
factors at play. If employers think that obese workers
are likely to be less productive or likely to be more expensive to employ because of health care costs, then
obese workers may have a more difficult time finding a
job than similarly qualified workers who are not obese.
In this article, we examine both self-reported health
and disability outcomes and employment outcomes
to try to distinguish between supply side and demand
side explanations. If, for example, there is no change
in the relationship between obesity and health outcomes, but there is a change in the relationship between obesity and employment outcomes, that would
suggest that demand side factors might play an important role in non-employment among the obese.
We are also interested in whether the changes we
observe over time in health and employment outcomes
are due to changes in the underlying population characteristics, such as a rising incidence of obesity, or due
to an increase in the differences in outcomes between
the obese and the nonobese. For example, if in every
period the obese are more likely to be in poor health
than the nonobese, then an increase in the proportion of
the population that is obese will likely lead to a larger
proportion of the population that does not work. On
the other hand, the propensity to report poor health,
disability, or non-employment among the obese compared with the nonobese may also have changed over
time. This change in propensities may be due to either
supply side or demand side factors that are shaped by
changes in health policies and/or labor market policies.
For example, in 1984 there was a substantial change
in disability insurance (SSDI) criteria that may have
made it more likely that someone with obesity-related
Kristin F. Butcher is an associate professor of economics at
Wellesley College and a former senior economist at the
Federal Reserve Bank of Chicago. Kyung H. Park is a
senior associate economist at the Federal Reserve Bank of
Chicago. The authors thank Dan Sullivan, Anna Paulson,
Bhashkar Mazumder, and seminar participants at the
Federal Reserve Bank of Chicago for helpful comments.
The views expressed here are those of the authors and do
not necessarily represent those of Wellesley College or any
other entity.
1Q/2008, Economic Perspectives
figure 1
Labor force participation, by gender and age group
percent
100
90
80
70
60
disability (defined as requiring the assistance of another person in handling routine tasks, such as personal care, housework,
or shopping) and to apply for disability
insurance has changed over time. Then,
we analyze how much of the change in
non-employment can be explained by
changes in obesity and other demographic
characteristics.
Changes in non-employment, age,
obesity, and disability insurance
First, we look at the changes in labor
force participation by gender and age group
from 1962 through 2006, using the March
40
Current Population Survey (CPS), which
30
is conducted by the U.S. Census Bureau
1962 ’66 ’70 ’74 ’78 ’82 ’86 ’90 ’94 ’98 2002 ’06
for the U.S. Bureau of Labor Statistics
Females, aged 16–65
Males, aged 16–65
(figure 1). Clearly, labor force participaMales, aged 25–55
Females, aged 25–55
tion among women rose dramatically from
the 1960s through the 1990s and leveled
Source: Authors’ calculations based on data from the U.S. Census Bureau,
March Current Population Surveys.
off in the 2000s. The change has been less
dramatic for men, but over the same period, we have seen a continuous decline in
health conditions could qualify for SSDI. This change,
men’s labor force participation. Note that this is the
combined with subsequent changes in the wage struccase even for relatively young men (aged 25–55).
ture that made SSDI benefits more generous relative
If we look at the share of survey respondents who
to low-wage jobs, may have made some obese people
reported that they had not worked the previous week
more likely to opt out of the labor market. Thus, an
(we call this the share “not working last week”)—which
increase in the number of obese people in the populaincludes nonparticipants and the unemployed—we
tion would have a different effect on outcomes, desee a similar pattern (figure 2). While the share not
pending on the period in which the change is evaluated.
working has declined for women, it has risen for men.
We find that, although those who are heavier have
Again, this is true even among relatively young men.
always had worse self-reported health outcomes and
Changes in the age distribution
employment outcomes, there is not much evidence
Some of the changes in the labor supply docuthat the propensity for the obese to have poor outcomes
mented
in the previous section may be related to changes
has changed over time. Non-employment among men
in
the
age
distribution. Figure 3 shows the shift in the
of prime age increased from 10 percent in 1984–85 to
age
distribution
among all 25–54 year olds between
12.5 percent in 2004–05. Increases in obesity alone can
1976–80
and
1999–2002.
As the baby boom generaexplain about 3 percent to 12 percent of that increase.
tion
ages,
there
is
a
change
in the average age among
In addition, population changes in age, race, and ethnici25–54
year
olds.
For
women,
labor supply peaks prity, combined with changes in obesity, can explain beor
to
childbearing
and
again
once
their children are older.
tween 34 percent and 47 percent of the increase in men’s
2
For
men,
Barrow
and
Butcher
(2004)
show that in
non-employment. These results suggest that deterioboth
1978–79
and
1999–2000
periods,
the fraction of
ration in underlying health has played an important role
men
who
did
not
work
at
all
in
the
previous
year inin the decrease in men’s labor force participation and
creased
monotonically
across
age
groups
for
those
that these population changes would have had similar
above
age
40.
Since
morbidity
increases
with
age, it
effects whether evaluated in the mid-1980s or early 2000s.
seems
likely
that
the
aging
of
the
population—even
In the next section, we describe recent trends in
among men aged 25–54—would lead to increases in
non-employment and labor force participation, age,
non-employment.
obesity, and disability insurance receipt. We examine
Barrow and Butcher (2004) point out that there
whether the propensity for the morbidly obese to selfhave
been other demographic changes, for example,
report musculoskeletal conditions and routine needs
50
Federal Reserve Bank of Chicago
figure 2
Share not working last week, by gender and age group
between 1978–79 and 1999–2000 can be
attributed to changes in age, race, and
ethnicity alone.
percent
70
Changes in obesity
Although many of the demographic
60
changes over the past 30 years might lead
us to expect a deterioration of health in
50
the working age population, many health
indicators suggest improvements in health
40
or improvements in individuals’ quality
of life, even when they have a health prob30
lem (Cutler and Richardson, 1997). How20
ever, obesity has become increasingly
common during this period. Obesity is
10
typically defined using the body mass index.3 A BMI lower than 18.5 is considered
0
1962 ’66 ’70 ’74 ’78 ’82 ’86 ’90 ’94 ’98 2002 ’06
underweight; a BMI lower than 25 (but
not lower than 18.5) is considered a
Males, aged 16–65
Females, aged 16–65
healthy or normal weight; a BMI greater
Males, aged 25–55
Females, aged 25–55
than or equal to 25 is deemed overweight;
Source: Authors’ calculations based on data from the U.S. Census Bureau,
a BMI greater than or equal to 30 is deemed
March Current Population Surveys.
obese; and a BMI greater than or equal to
40 is considered morbidly obese.
Figure 4 shows the probability density function for BMI for men and women
figure 3
aged 25–54 years old in the 1976–80 and
Age distribution
1999–2002 National Health and Nutrition
percentage of population
Examination Surveys (NHANES), which
5
are conducted by the U.S. Department of
Health and Human Services, Centers for
1976–80
4
Disease Control and Prevention, National
1999–2002
Center for Health Statistics. These distributions show the rightward shift in the
3
BMI distribution over time.
Although there has been an increase
2
in median BMI, a significant feature underlying the obesity epidemic is that the
variance in BMI has increased. The heavy
1
have gotten much heavier over time.
Panels A and B of figure 5 highlight these
0
changes in the distribution of BMI, using
25 28
31
34 37
40 43
46
49 52
55
age
NHANES data for men and women, respectively. Note the median male in
Note: The distribution is among those aged 25–54.
Source: Authors’ calculations based on data from the U.S. Department of
1999–2002 would have been heavier than
Health and Human Services, Centers for Disease Control and Prevention,
nearly three-quarters of the population in
National Center for Health Statistics, National Health and Nutrition
Examination Survey.
the earlier period 1976–80. A male just on
the cusp of obesity (75th percentile) in
the 1999–2002 BMI distribution would
have been heavier than 90 percent of the earlier perichanges in the racial and ethnic mix of the population,
od’s population. For females, we also see dramatic
that may also be correlated with deteriorating health.
changes in the BMI distribution in the heaviest porTheir analysis, which does not control for obesity,
tions of the distribution.
finds that 14 percent to 33 percent of the increase
in men’s full-year non-employment that occurred
1Q/2008, Economic Perspectives
figure 4
Body mass index density
B. Female
percentage of female population
A. Male
percentage of male population
14
14
1976–80
12
12
10
10
8
8
6
6
1999–2002
4
1976–80
4
2
1999–2002
2
0
10
20
30
40
body mass index
50
0
10
60
20
30
40
50
body mass index
60
Notes: The density is calculated for those aged 25–54. A body mass index lower than 18.5 is considered underweight; 18.5–24.9,
normal weight; 25.0–29.9, overweight; 30.0–39.9, obese; and 40.0 or higher, morbidly obese.
Source: Authors’ calculations based on data from the U.S. Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Health Statistics, National Health and Nutrition Examination Survey.
figure 5
Body mass index distribution
A. Male
body mass index
50
B. Female
body mass index
50
40
1999–2002
40
1999–2002
30
30
1976–80
1976–80
20
20
10
10
0
0
1
5
10
25
50
75
percentile
90
95
99
1
5
10
25
50
75
percentile
90
95
99
Note: The distribution is among those aged 25–54.
Source: Authors’ calculations based on data from the U.S. Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Health Statistics, National Health and Nutrition Examination Survey.
If it is the very heavy who are most likely to suffer ill health from obesity, then the population at risk
of obesity-related health conditions has increased.
Further, if being heavy is more likely to cause one
health problems as one ages, then we would expect
Federal Reserve Bank of Chicago
that as these heavier cohorts age, they will experience
more weight-related health problems than previous,
slimmer cohorts.
Figures 1 through 5 demonstrate that non-employment among men of prime age has increased.
They also document shifts in the population—namely,
the population is older and more likely to be obese—
that are consistent with a health-based reason for this
decline in work among men.
disorders; by 2003, this figure had risen to 26.3 percent.
Mental disorders have also accounted for an increased
share of SSDI awards since 1981.
Figure 7 demonstrates how SSDI awards for various causes have changed on a population basis (per
Changes in disability insurance
10,000 individuals, aged 16–64). Heart disease and
Figure 6 shows that the percentage of the populacancer have held steady as reasons for disability intion receiving disability benefits has risen substantially
surance claims, but musculoskeletal conditions, mental
since the early 1980s and that the increase seems to
illness, and other sources have increased.5 This shift
have begun after 1984.4 Changes to the disability inin the reasons documented for disability receipt is ofsurance eligibility rules in 1984 appear to have inten seen as being due to changes in the criteria used
creased the likelihood that an SSDI applicant would
to judge whether an individual is disabled. Diseases
receive payments. As Autor and Duggan (2003) exthat are easily verifiable by a physician—for example,
plain, the awards criteria now give more weight to an
cancer and heart disease—have declined as a share of
individual’s pain and ability to function in the work
all disability awards. This is not to say, however, that
place; prior to 1984, eligibility was determined by “conthere have not also been changes in underlying health
tinuous disability reviews” by third party physicians.
that would contribute to these shifts in disability inIn addition, rising wage inequality during the 1980s
surance payments.
and 1990s increased the value of SSDI payments relFor example, there are many ways that the increase
ative to wages for many individuals. Many observers
in obesity may be related to the increase in the share
have linked these changes in the SSDI program to inof disability awards for musculoskeletal disorders.
creases in disability insurance receipt and decreases
It may be that the increase in obesity has led to more
in employment.
musculoskeletal disorders, in turn leading to more disCoinciding with these programmatic changes, there
ability claims. In this case, the driver of the increase
have been changes in the primary diagnoses among
is the change in obesity rates leading to more muscurecipients. Table 1 documents the share of disability
loskeletal disorders. On the other hand, changes in
awards attributed to different disorders. In 1981,
disability insurance rules—which now give more emprior to the new disability insurance eligibility criteria,
phasis to an individual’s report of pain—may have
17 percent of all awards were for musculoskeletal
also given those who are obese, and thus have a
better basis for making a claim of musculoskeletal pain, a better chance to qualify
figure 6
for SSDI. Changes in wages relative to
Social Security Disability Insurance award rate
SSDI payments may have given workers
per population
an increased incentive to apply for disaward rate per population, index, 1980 = 100
ability insurance.
300
In the next section, we examine whether the propensity of the obese to claim
250
various health ailments, to self-report
routine needs disability, or to apply for
Female
200
SSDI has changed over time. The 1984
change in the SSDI rules does not fall in
Overall
150
the span of our data on self-reported health,
so this exercise does not shed light on
how that policy change may have affected
100
Male
behavior. Instead it allows us to answer
50
the following question: During the period
after 1984 when awards for musculoskel0
etal disorders continue to rise, do we see
1980
’83
’86
’89
’92
’95
’98
2001
’04
a rise in the propensity of the obese to report these ailments?
Note: The sample population is made up of those aged 30–54.
Sources: Authors’ calculations based on data from the U.S. Social Security
In the rest of this article, we focus
Administration, Annual Statistical Supplement to the Social Security Bulletin,
only on men aged 25–54 years old, since
2005; and Haver Analytics.
it is this group that has shown a rising
1Q/2008, Economic Perspectives
Table 1
Share of total Social Security Disability Insurance awards, by diagnosis
Total
Males
2003
1981
Females
Diagnosis
1981
( - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - percentage - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - )
Heart disease
Cancer
Mental disorders
Musculoskeletal disorders
Nervous system
Respiratory system
Endocrine system
All other
Total
24.9
16.3
10.5
17.0
8.3
6.2
4.3
12.5
100.0
11.4
9.4
25.4
26.3
8.5
4.2
3.1
11.6
99.9
27.9
15.1
10.2
15.6
7.9
6.5
3.8
12.9
100.0
2003
1981
14.4
9.1
23.0
24.7
8.2
4.1
3.1
13.5
100.1
2003
18.0
19.1
11.2
20.1
9.1
5.5
5.5
11.5
100.0
7.8
9.7
28.2
28.2
8.9
4.4
3.1
9.9
100.2
Note: Columns may not total because of rounding.
Sources: Authors’ calculations based on data from the U.S. Social Security Administration, Annual Statistical Supplement to the Social Security
Bulletin, 1981, and Annual Statistical Report on the Social Security Disability Insurance Program, 2003.
trend in non-employment over this period. The underlying health conditions have changed in similar ways
for women, making their large increase in labor force
participation even more striking.
Self-reported health conditions, disability,
SSDI receipt, and obesity
Since one can report a health condition without claiming to be disabled by it
and since one can claim to have a disability without applying for disability insurance, we examine the relationship between
obesity and each of these outcomes separately. We show how the relationship has
changed over time. We are particularly
interested in whether the propensity for
those who are heavy to report poor health
outcomes has increased over time, which
would be consistent with changes in the
incentives of the obese to apply for SSDI
and leave the labor force.
Figure 8 shows the unadjusted prevalence of musculoskeletal disorders for
men who are underweight, normal
weight, overweight, obese, and morbidly
obese. From 1984 through 1996, those
who are heavier are more likely to report
a musculoskeletal problem. There is an
increase in reports of musculoskeletal
problems among the morbidly obese
from 1984 through 1988, but there is a
decline in later years. In general, there is
little evidence of an increase in the propensity for the obese and morbidly obese
Federal Reserve Bank of Chicago
to report a musculoskeletal problem. This finding may
be somewhat misleading, however, because it does
not control for other demographic differences that may
be correlated with obesity and with reports of musculoskeletal problems. To address this, we use regression
figure 7
Social Security Disability Insurance awards
per population, by diagnosis
awards per 10,000 persons
14
12
10
8
6
4
2
0
1981
’84
’87
’90
’93
’96
’99
2002
Mental disorders
Cancer
Musculoskeletal conditions
Other
’05
Heart disease
Notes: The sample population is made up of those aged 16–64. Changes
to the disability insurance eligibility rules were made in 1984. See the text
for further details.
Source: Authors’ calculations based on data from the U.S. Social Security
Administration, Annual Statistical Report on the Social Security Disability
Insurance Program, 2005.
analysis, which allows us to hold constant other demographic differences and
examine whether the likelihood of reporting a given health issue has changed over
time by weight category.
The National Health Interview Survey (NHIS)—conducted by the U.S. Department of Health and Human Services,
Centers for Disease Control and Prevention, National Center for Health Statistics—asks a series of health questions
that allow us to examine components
of musculoskeletal disorders. Figure 9
presents differences in reporting of
lower back pain between morbidly obese
men and those of normal weight in the
1997–2005 National Health Interview
Surveys. We calculated these differences
by running a linear probability model on
whether the individual reports lower back
pain, controlling for indicator variables
for underweight, overweight, obese, and
morbidly obese. Normal weight is the
omitted category. Only the morbidly obese
were statistically significantly more likely
to report these ailments. We ran separate
regressions without any controls, as well
as controlling for age alone and then controlling for age, race, and Hispanic ethnicity.6 We ran a separate regression for
each year, thus allowing the effect of the
regressors to differ each year. (Figures 9,
10, and 11 also include the 95 percent
confidence intervals for the difference in
reporting between the morbidly obese
and those of normal weight.)
We see that over this period, those
who are morbidly obese are more likely
to report lower back pain, although for
some years this difference is not statistically significantly different from zero.
Although the point estimate for the
difference in reporting lower back pain
is higher later in the period, the difference in the effects between the two periods is not statistically significant. Thus,
there is little evidence of an increase in
the difference in reports of lower back
pain between the morbidly obese and
those who are of normal weight during
this period. Also, note that our estimates
do not vary substantially as we add control variables. The results are similar for
figure 8
Prevalence rates of chronic musculoskeletal conditions
among males
percent, musculoskeletal condition per male population
35
30
25
20
15
10
5
0
1984
1985–
86
1987–
88
1989–
90
1991–
92
1993–
94
Underweight
Obese
Normal weight
Morbidly obese
1995–
96
Overweight
Note: The sample population is made up of males aged 25–54.
Source: Authors’ calculations based on data from the U.S. Department of
Health and Human Services, Centers for Disease Control and Prevention,
National Center for Health Statistics, National Health Interview Survey.
figure 9
Incidence of lower back pain: Morbidly obese vs.
normal weight males
estimate of morbidly obese indicator
.25
.20
.15
.10
.05
.00
–.05
–.10
1997
’98
’99
No controls
Age
2000
’01
’02
’03
’04
’05
Age, race, ethnicity
95 percent confidence interval
Note: The sample population is made up of males aged 25–54.
Source: Authors’ calculations based on data from the U.S. Department of
Health and Human Services, Centers for Disease Control and Prevention,
National Center for Health Statistics, National Health Interview Survey.
1Q/2008, Economic Perspectives
other components of musculoskeletal disorders, such
as reported arthritis or other joint pain.
Figure 10 examines whether the morbidly obese
have become relatively more likely over time to report routine needs disabilities. The data are from the
NHIS from 1984 through 2005. There were significant
changes in sequence and wording of the disability
questions between 1996 and 1997, and thus, we show
a break in the series.7 The figures are based on linear
probability models that are analogous to those described
for figure 9.
Again, we see that the morbidly obese are more
likely to report a routine needs disability; and controlling for age, race, and ethnicity makes little difference
in the size of that effect. However, there is no statistically significant difference in the size of the effect of
morbid obesity across time periods.
Finally, figure 11 shows the difference in the probability of ever having applied for disability insurance
between the morbidly obese and those categorized as
having normal weight, controlling for age, race, and
ethnicity. Information on applications for disability
insurance are only available after 1996, and all respondents are asked if they have “ever applied for” disability
insurance. While the morbidly obese have always been
statistically significantly more likely to have applied
for disability insurance than those of normal weight,
this difference is stable over the period observed.
Table 1 (p. 7) and figures 6 (p. 6) and 7 (p. 7) in
the previous section showed that disability awards have
been increasing since the mid-1980s, particularly for
musculoskeletal ailments. In this section, we examined
the relationship between obesity, health, disability,
and application for SSDI. The evidence shows that
obesity has increased, with morbid obesity having increased in particular. In addition, since the mid-1980s
the morbidly obese, in particular, have reported worse
health outcomes than their nonobese counterparts. However, over this period we have not seen an increase in
the propensity to report worse health outcomes by the
morbidly obese, nor an increase in the likelihood of
their applying for SSDI. What we have seen is that
there are now more of the category of people—very
obese people—who have always reported worse health
outcomes, but not much evidence of an increase in
the likelihood of reporting worse health outcomes
among the very obese.
Non-employment and obesity
In this section, we examine the relationship between obesity and employment. This relationship may
be different from the relationship between obesity
figure 10
Routine needs disability: Morbidly obese vs. normal weight males
estimate of morbidly obese indicator
0.12
0.10
0.08
0.06
0.04
0.02
0.00
–0.02
1984
1985–
86
1987–
88
No controls
1989–
90
Age
1991–
92
1993–
94
1995–
96
1997–
98
Age, race, ethnicity
1999–
2000
2001–
02
2003–
04
2005
95 percent confidence interval
Notes: The sample population is made up of males aged 25–54. Routine needs disability is defined as requiring the assistance
of another person in handling routine tasks, such as personal care, housework, or shopping. There were significant changes in
sequence and wording of the disability questions between 1996 and 1997, and thus, we show a break in the series. See note 7
for further details.
Source: Authors’ calculations based on data from the U.S. Department of Health and Human Services, Centers for Disease Control
and Prevention, National Center for Health Statistics, National Health Interview Survey.
Federal Reserve Bank of Chicago
on the ability to perform a job may be
smaller in the current technological era
Ever applied for Social Security Disability Insurance:
than it would have been when heavy
Morbidly obese vs. normal weight males
physical exertion was a frequent requireestimate of morbidly obese indicator
ment at work.
0.14
In addition, there is some evidence of
discrimination
against obese people (see
0.12
Carpenter, 2006, and Cawley and Danziger,
2004). Suppose there are two equally pro0.10
ductive individuals—one obese and one
0.08
not obese—and employers are less willing
to hire the obese individual. If that prefer0.06
ence for the nonobese was constant over
time, the increase in the obese population
0.04
could lead to an increase in the fraction
0.02
of individuals who are not working. In
addition, however, employers’ “prefer0.00
ence” for hiring nonobese people could
1997–
1999–
2001–
2003–
2005
change over time. On the one hand, tech98
2000
02
04
nological changes that reduce the physical
No controls
Age, race, ethnicity
requirements of jobs would seem to narAge
95 percent confidence interval
row any perceived productivity gap beNote: The sample population is made up of males aged 25–54.
tween obese and nonobese workers. On the
Source: Authors’ calculations based on data from the U.S. Department of
Health and Human Services, Centers for Disease Control and Prevention,
other hand, even if productivity is not a
National Center for Health Statistics, National Health Interview Survey.
concern, the rising costs of employer-provided health insurance may make employers less inclined to hire those they perceive
as being costly employees over time.
and self-reported health measures. For example, across
Finally, of course, those who are obese may be
a number of periods, an increase in obesity may affect
less likely to work than individuals of normal weight
health in a similar way. However, the employment refor other, more personal reasons. They may be in
sponse to that change in health may differ, depending
poorer health, making work more difficult, or they
on both demand side (from the employers’ perspecmay find work less enjoyable than their counterparts
tive) and supply side (from the workers’ perspective)
of normal weight. Changes in working conditions
changes in the employment–obesity relationship.
may also have an impact on obese workers. These conFirst, there is some debate about the ways in which
ditions could include demand side factors, discussed
changes in employment itself may have contributed
previously, or supply side factors. If, for example, wages
to the rise in obesity (Philipson and Posner, 1999).
for the obese fall or SSDI becomes either easier to get
For many of us, technological changes have tended to
or more generous relative to the wages they could
reduce the calories we expend at work by letting us
likely command, then the obese might change their
spend more time at our desks. This is true even in empropensity to work in a given period.
ployment sectors that typically required more physical
In the analysis that follows, we want to disentangle
activity, as more and more processes in industrial and
the increase in non-employment that has arisen because
manufacturing environments have become automated.
there are more obese people, and particularly more
This trend may have contributed to the long-term inmorbidly obese people, today than there were 20 years
crease in BMI, although much of the recent rise in obeago from any increase that has occurred because the
sity seems to have begun in the 1980s, when one might
effect of obesity on non-employment has changed.8
argue that the transition from hard physical labor to
We focus on measures of non-employment that are
sedentary work had already happened. Nonetheless,
available in the data sets that also track obesity over
that transition may have important implications for the
time. The two main data sets are the National Health
effect of obesity on one’s ability to work—if most
Interview Survey and the National Health and Nutripeople at work are engaged in sedentary tasks that retion Examination Survey. Note that the information
quire little physical exertion, then the effect of obesity
available in the data set usually used to track labor
figure 11
10
1Q/2008, Economic Perspectives
market statistics (CPS) and the information available
in the data sets usually used to track health statistics
(NHIS and NHANES) are not the same. In particular,
the data sets that contain information on BMI and obesity have less detailed information on whether one is
working. In the CPS, one can examine the fraction of
the year spent not working, for example, or the fraction of the population that is not employed for the entire year (see Barrow and Butcher, 2004). In the health
data sets, the available data restrict us to classifying
people as non-employed if they report not working in
the previous one to two weeks. Table 2 compares the
health and labor force data available in the Current
Population Surveys, National Health Interview Surveys,
and National Health and Nutrition Examination Surveys.
Table 3 shows the differences in the reported
share of non-employed by year using the different
data sets. We see that the NHIS closely tracks the
non-employment figures calculated from the CPS. In
contrast, the NHANES overstates the growth in nonemployment among men in the prime age category by
more than twofold. For this reason, we focus on the
NHIS in the analysis that follows.
In order to examine how much of the change in
non-employment can be explained by changes in
obesity, we use an Oaxaca–Blinder multivariate decomposition (see Oaxaca, 1973, and Blinder, 1973).
Here, we run linear probability regressions with not
working in the past one to two weeks as the outcome
variable. We control for underweight, overweight,
obese, and morbidly obese as the weight categories,
with the normal weight category omitted. In some regressions, we also control for age, race, and ethnicity,
as well as for pairwise interactions between weight
categories and age and race. We run these regressions
in both the early (1984–85) and later (2004–05) years
of our data series:
1)
0
1
Y84 −85 = β84
− 85 + β84 − 85 X 84 − 85 + ε84 − 85 ;
2) Y04 − 05 = β004 − 05 + β104 − 05 X 04 − 05 + ε04 − 05 .
Typically, these equations are then rearranged to
examine how much of the difference in outcomes between the two years is due to differences in the explanatory (X) variables, and how much is due to differences
in the effects of these variables on the outcomes, the
β values. Differences attributable to changes in the
Table 2
Comparison of labor force and health data, by data source
Labor force data
CPS
March
NHIS
1984–96
NHIS
1997–2005
NHANES
1976–80
Worked last 1–2 weeks
X
X
X
X
Reason not working last week
X
X
X
Class of worker
X
X
X
X
Hours worked last week
X
X
Full/part time
X
X
X
Weeks worked
X
Months worked
X
X
Wage data
X
Industry
X
X
X
X
Occupation
X
X
X
X
Health data
Body mass index or weight/height
X
X
X
Disability/physical limitations
X
X
X
X
Conditions causing disability
X
X
X
Ever applied for Social Security
Disability Insurance
X
X
NHANES
1999–2002
X
X
X
X
X
X
X
X
X
X
X
Notes: CPS means Current Population Survey. NHIS means National Health Interview Survey. NHANES means National Health and Nutrition
Examination Survey. In the NHIS 1984–96 and NHANES 1976–80, the employment status question asks whether or not the respondent has
worked in the past two weeks, while the NHIS 1997–2005 and NHANES 1999–2002 ask about employment status in the past one week.
The March CPS employment status variables (esr and mlr) also ask about employment status in the past one week. The March CPS also
asks questions related to disability status. One variable notes whether or not “health or disability limits kind or amount of work.” Another
records whether someone left a job for health reasons. Finally, the data include a variable indicating whether or not the household receives
disability income.
Sources: U.S. Census Bureau, March Current Population Surveys; and U.S. Department of Health and Human Services, Centers for Disease
Control and Prevention, National Center for Health Statistics, National Health Interview Survey and National Health and Nutrition Examination
Survey.
Federal Reserve Bank of Chicago
11
Table 3
Comparison of share of non-employed males,
by data source
CPS
NHIS
( - - - - - - - - percentage - - - - - - - )
2004–05
1984–85
Change
13.4
11.5
2.0
12.5
10.3
2.2
CPS
NHANES
1999–2002
1976–80
Change
11.9
9.8
2.1
12.0
7.5
4.6
Notes: The sample population is made up of males aged 25–54.
Columns may not total because of rounding. CPS means Current
Population Survey. NHIS means National Health Interview Survey.
NHANES means National Health and Nutrition Examination Survey.
Sources: Authors’ calculations based on data from the U.S. Census
Bureau, March Current Population Surveys; and U.S. Department
of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Health Statistics, National Health
Interview Survey and National Health and Nutrition Examination
Survey.
X variables are attributable to changes in obesity, age,
race, and ethnicity.9 Differences attributable to changes
in the coefficients, on the other hand, are attributable
to the supply side and demand side factors described
previously.
3) Y04 − 05 − Y84 −85 = β104 − 05 ( X 04 − 05 − X 84 −85 )
0
+ (β104 − 05 − β184 −85 ) X 84 −85 + (β004 − 05 − β84
− 85 ).
The first term after the equals sign is the difference attributable to changes in the X values, and the
second two terms are the differences attributable to
changes in the coefficients. As written out in equation 3,
the change in individual characteristics between the
two periods is evaluated using the “returns” to these
characteristics that prevailed in the later period. If we
had done the subtraction the other way, we would get
a different answer.
Our approach is to examine how the changes in
individual characteristics that actually occurred between
1984–85 and 2004–05 would have been expected to
change the fraction of the population that was not
working, given the “conditions” that prevailed in both
the earlier and later periods. We can use equations 1
and 2 to predict how people with the characteristics
of those who existed in 2004–05 would have “behaved”
in 1984–85:
β184 −85 X 04 − 05 .
12
And we can use those same equations to predict how
people with the characteristics of those who existed
in 1984–85 would have “behaved” in 2004–05:
β104 − 05 X 84 −85 .
Suppose we imagine that the only thing that explains the increase in men’s non-employment is that
non-employment is higher among the morbidly obese
and that, in the later period, more men are morbidly
obese. Then, evaluating the effect of the increase in
morbid obesity using the “returns” to morbid obesity
that prevailed in the earlier period should yield the
exact increase in non-employment that we observe in
the data. Since, in fact, conditions, or “returns to characteristics,” may have changed, we can think of this
exercise as answering the following question: How
much of an increase in non-employment would we
have expected in 1984–85 if morbid obesity had increased to today’s levels under those conditions?
We present these calculations, allowing age,
race, and ethnicity characteristics to change in addition to obesity measures, and we allow for pairwise
interactions in these characteristics. Age may exacerbate the health problems associated with obesity—for
example, the knees of 30 year olds may not hurt among
either those of normal weight or the obese, but the
knees of 50 year olds may have suffered more wear
and tear among the obese but still be fairly pain free
among those of normal weight. And thus, we would
find that adjusting the data from the two periods to
have the same age–obesity profile explains more of
the change in non-employment over time. Obesity
may have different effects in different populations as
well as for different age groups. If obesity-related
health problems are more prevalent among blacks
and Hispanics, for example, then adjusting for the
obesity–age–race/ethnicity profile may explain more
of the changes over time. We present the results for
these different adjustments separately.
Our decompositions are similar to those presented in Lakdawalla, Bhattacharya, and Goldman (2004).
They examine how much of the increase in disability
rates across different age groups between 1984 and
1996 can be explained by the rise in obesity. They decompose the change in disability rates between 1984
and 1996 into:
[( O96 − O84 ) *( D90O − D90NO )]
+ [O90 *{( D96O − D84O ) − ( D96NO − D84NO )}],
1Q/2008, Economic Perspectives
where Oyr is the obesity rate in a given year and Dyr
is the disability rate in a given year, and where the
superscripts denote whether the disability rate is measured among the obese (O) or the nonobese (NO).
The first term in this expression is the amount of
increased disability we would have expected had obesity risen as it did between the two periods, but the effect
of obesity on disability was as it was in the interim
year—1990. The second term is the amount of increase
in disability that is due to the fact that disability among
the obese rose, holding constant obesity rates at the
level of the interim period. Using this decomposition,
Lakdawalla, Bhattacharya, and Goldman (2004) find
that 50 percent of the rise in disability for 18–29 year
olds; 25 percent for 30–39 year olds; 10 percent for
40–49 year olds; and nearly all for 50–59 year olds
can be explained by increases in obesity.10
This calculation combines the rise in disability
that comes from the increase in obesity and the rise in
disability that comes from changes in the effect of
obesity on disability. In our analysis that follows, we
focus on numbers that are similar to the first component—the amount by which non-employment would
have risen had obesity rates risen—but we show this
effect under the conditions of the earlier and later periods—that is, holding constant the effect of obesity
on non-employment at its level in the earlier period
and then at its level in the later period.
Table 4 presents the results of these simulations.
The first row shows actual non-employment rates,
which increased 2.2 percentage points, from 10.3 percent to 12.5 percent between 1984–85 and 2004–05.
The second row shows predicted non-employment rates
given the BMI distribution that existed in the other
period, using the coefficients for the period listed in
the column heading. For example, looking at the
second row of numbers, the first column tells us that
had the weight distribution that existed in 2004–05
occurred in 1984–85, we would have seen a non-employment rate of 10.4 percent in 1984–85—slightly
higher than the actual non-employment rate in that
period. Similarly, if the weight distribution that existed in 1984–85 occurred in 2004–05, we would expect
a non-employment rate of 12.3 percent—slightly lower
than the actual non-employment rate in that period.
The last two columns show us how much of the actual
change in non-employment between the two periods
can be explained by evaluating the change in characteristics listed on the leftmost column using the returns to those characteristics in the years given in the
column headings. So, about 3 percent of the increase
in non-employment can be explained by the rise in
obesity alone using the “returns” to obesity that prevailed in 1984–85. About 13 percent of the rise in
non-employment would be attributed to the increase
in obesity if we evaluated that increase using the “returns” that prevailed in 2004–05.11 This is consistent
with a story in which either supply side or demand
side deterrents to working for the obese are stronger
in 2004–05 than in 1984–85. For example, this could
occur if disability insurance takeup rates are higher
among the obese in the later period. However, if there
are other characteristics of obese workers that are
also correlated with non-employment but are not held
constant in these regressions, then those effects will
load onto the obesity coefficients here, leading us to
attribute either too little or too much of the changes to
changes in obesity. Furthermore, changes in the characteristics we use in our analysis—age, race, and ethnicity—may also be linked to changes in underlying
health. Finally, as discussed earlier, we want to include
interactions between age, race, ethnicity, and weight
Table 4
Actual and simulated average share of non-employed males and the percent
of actual change explained by given characteristics
1984–85
( - - - - - - - - - - - - - - - - - - - - - - - - percentage - - - - - - - - - - - - - - - - - - - - - - - )
Actual non-employment
Characteristics used in simulation
Weight categories
Weight categories, age polynomial
Weight categories, age, race, ethnicity (all interactions)
2004–05
Percent of actual increase
explained by characteristics
under conditions in:
10.3
12.5
10.4
10.7
11.4
12.3
11.8
11.8
1984–85
3.4
14.3
46.8
2004–05
12.5
31.6
33.9
Notes: The sample population is made up of males aged 25–54. The normal weight category is excluded from the weight categories. See the text
for further details.
Source: Authors’ calculations based on data from the U.S. Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Health Statistics, National Health Interview Survey.
Federal Reserve Bank of Chicago
13
measures. If it is not just the fraction of the population
that is morbidly obese that matters for non-employment, but rather the fraction that is older and morbidly obese, we want to capture that in our simulations.
Including a polynomial in age in our simulations
increases the amount of the increase in non-employment that we can explain to 14 percent using the earlier period and 32 percent using the later period. Once
we include age, race, and ethnicity in the models, more
of the increase in non-employment can be explained
using the returns to characteristics that prevailed in
1984–85 than those in 2004–05. Changes in these
characteristics can explain from 34 percent (using
2004–05 returns to characteristics) to 47 percent
(using 1984–85 returns to characteristics) of the increase in the non-employment rate.12
Changes in age, race, and ethnicity—which may
themselves be markers of changes in underlying health—
explain a larger share of the increase in non-employment between 1984–85 and 2004–05 than do changes
in obesity measures alone. However, (in results not
shown) adding obesity measures to simulations that
include age, race, and ethnicity controls increases the
amount of the predicted increase in non-employment
by 10 percentage points, regardless of which period
we use to evaluate the change.
These results suggest that changes in underlying
population characteristics may have played an important role in the increase in non-employment among
men of prime working age over the past 30 years.
Conclusion
This article examines the role of the increase in
obesity in changes in non-employment. Men of prime
working age have increased their non-employment rates
over the past 30 years, and disability rates have also
increased. Many have noted that this increase has
happened against a backdrop of generally improving
health in the U.S. population. However, obesity has
14
increased substantially over this period. Here, we have
tried to disentangle the changes that occurred in heath
and employment because of the increase in the fraction of the population that is obese from the changes
that are due to changes in the differences in outcomes
between obese and nonobese individuals. We find that,
while the morbidly obese have always been more likely to report musculoskeletal ailments and more likely
to report being disabled, their propensity to report ailments and disability has not statistically significantly
increased over time.
The results for non-employment are consistent
with those for health and disability. If the results had
shown that increases in obesity had little effect on health
and disability rates but had a large effect on employment, this would have pointed toward the importance
of demand side factors—such as efforts by employers
to avoid higher health care costs—in employment outcomes for the obese. However, since the results are
consistent for health, disability, and non-employment,
we cannot use these differences to infer the relative
importance of demand side or supply side effects.
For men of prime working age, changes in their
characteristics—including age, race, ethnicity, and
obesity levels—can explain a large portion (around
40 percent) of the increase in non-employment over
the period. The portion of the change in non-employment that is explained by changes in these characteristics is similar regardless of whether we evaluate the
change in characteristics using the returns to characteristics that prevailed in either the earlier period
(1984–85) or the later period (2004–05). This means
that under either the earlier or later labor market conditions, we would expect that these changes in characteristics would lead to a substantial increase in nonemployment. Similar to Lakdawalla, Bhattacharya,
and Goldman (2004), we find that the obesity epidemic
may be playing an important role in changing labor
market outcomes.
1Q/2008, Economic Perspectives
notes
See Barrow (2004); Anderson, Barrow, and Butcher (2005); and
Aaronson, Park, and Sullivan (2006) for trends in unemployment
rates and labor force participation.
1
A 2006 revision to Barrow and Butcher (2004) is available from
the authors upon request.
2
Body mass index = (weight in kilograms)/(height in meters squared).
3
Disability benefit award numbers are from the Social Security
Administration’s (SSA) Annual Statistical Supplement to the Social
Security Bulletin, 2005, and the noncivilian population figures
come from the monthly household data in Haver Analytics. Note
that disability awards have risen particularly among women.
Disability insurance pays benefits to an individual and certain family members, provided that the individual is “insured”—meaning
that the person has worked long enough and paid social security
taxes. The increase in women’s labor supply presumably increased
the pool of eligible workers. See www.ssa.gov/disability/.
asked about personal care or routine needs disability. In 1997, the
wording of the disability question also changed. Previously the
personal care question read, “Because of any impairment or health
problem, does ___ need the help of other persons with personal
care needs, such as eating, bathing, dressing, or getting around this
home?” After 1996, however, the question read, “Because of a mental,
physical, or emotional problem, does ____ need the help of other
persons with personal care needs, such as eating, bathing, dressing,
or getting around this home?”
4
The substantial increase in musculoskeletal conditions in 1995 is
due to a different sampling methodology used by the Social Security
Administration. Prior to 1995, the SSA only included awards allowed
after the initial determination. Since many musculoskeletal conditions are denied initially and awarded later after an appeals process,
the pre-1995 sample understates the share of musculoskeletal awards
relative to the post-1995 sample that includes awards granted after
the appeals process.
5
Specifications include age and age squared, as well as indicator
variables for black, other, and Hispanic ethnicity.
6
Prior to 1997, only respondents who had a major activity limitation were asked if they needed assistance with personal care or routine need tasks. Individuals older than 60 years, however, were not
screened and were automatically asked about any potential disability.
In 1997, respondents were no longer screened and everyone was
7
Federal Reserve Bank of Chicago
The former will be changes in the characteristics of the population
(the Xs) and the latter changes in coefficients (the βs).
8
Specifically, weight categories (underweight, overweight, obese,
and morbidly obese), age, age squared, black, other, Hispanic ethnicity, and interactions between the weight categories and the other
demographic variables (age and race) are included in the X values.
9
10
Their analysis also includes women.
This is because the point estimate for the coefficient on morbid
obesity is higher in the later years; however, just as in the results
for health conditions presented earlier, this difference is not statistically significant.
11
We find similar results if we decompose the change in routine needs
disabilities. Because of the change in survey questions regarding
routine needs disabilities, we perform this analysis for changes from
1984–85 through 1995–96 and from 1996–97 through 2004–05.
Changes in weight categories, age, race, and ethnicity can explain
about a third of the increase in routine needs disabilities between
1984–85 and 1995–96, using the “returns” to these characteristics
that prevailed in either time period. Changes in these characteristics
between 1996–97 and 2004–05 explain about 33 percent of the increase in routine needs disabilities using the “returns” to characteristics that prevailed in 1996–97 and about 42 percent of the
increase using “returns” that prevailed in 2004–05.
12
15
REFERENCES
Aaronson, Daniel, Kyung-Hong Park, and Daniel
Sullivan, 2006, “The decline in teen labor force participation,” Economic Perspectives, Federal Reserve
Bank of Chicago, Vol. 30, No. 1, First Quarter, pp. 2–18.
Anderson, Katharine, Lisa Barrow, and Kristin F.
Butcher, 2005, “Implications of changes in men’s and
women’s labor force participation for real compensation growth and inflation,” Topics in Economic
Analysis and Policy, Vol. 5, No. 1, article 7.
Autor, David H., and Mark G. Duggan, 2003, “The
rise in the disability rolls and the decline in unemployment,” Quarterly Journal of Economics, Vol. 118, No. 1,
February, pp. 157–205.
Barrow, Lisa, 2004, “Is the official unemployment
rate misleading? A look at labor market statistics over
the business cycle,” Economic Perspectives, Federal
Reserve Bank of Chicago, Vol. 28, No. 2, Second
Quarter, pp. 21–35.
Barrow, Lisa, and Kristin F. Butcher, 2004, “Not
working: Demographic changes, policy changes, and
the distribution of weeks (not) worked,” Federal
Reserve Bank of Chicago, working paper,
No. WP-2004-23.
Blinder, Alan S., 1973, “Wage discrimination: Reduced
form and structural estimates,” Journal of Human
Resources, Vol. 8, No. 4, Autumn, pp. 436–455.
16
Carpenter, Christopher S., 2006, “The effects of
employment protection for obese people,” Industrial
Relations, Vol. 45, No. 3, July, pp. 393–415.
Cawley, John, and Sheldon Danziger, 2004, “Obesity as a barrier to the transition from welfare to work,”
National Bureau of Economic Research, working
paper, No. 10508, May.
Cutler, David, and Elizabeth Richardson, 1997,
“Measuring the health of the U.S. population,” Brookings
Papers on Economic Activity: Microeconomics,
Vol. 1997, pp. 217–282.
Juhn, Chinhui, Kevin M. Murphy, and Robert H.
Topel, 2002, “Current unemployment, historically
contemplated,” Brookings Papers on Economic
Activity, Vol. 2002, No. 1, pp. 79–135.
Lakdawalla, Darius, Jayanta Bhattacharya, and
Dana P. Goldman, 2004, “Are the young becoming
more disabled?,” Health Affairs, Vol. 23, No. 1,
January/February, pp. 168–176.
Oaxaca, Ronald, 1973, “Male–female wage differentials
in urban labor markets,” International Economic
Review, Vol. 14, No. 3, October, pp. 693–709.
Philipson, Tomas J., and Richard A. Posner, 1999,
“The long-run growth in obesity as a function of technological change,” National Bureau of Economic
Research, working paper, No. 7423, November.
1Q/2008, Economic Perspectives
Avoiding a meltdown: Managing the value of small change
François R. Velde
Introduction and summary
In 2007, the American bald eagle, a symbol of our nation, was removed from the threatened species list. But
another American icon (or two) might well take its
place on the list. On December 14, 2006, the United
States Mint announced new regulations “to limit the
exportation, melting, or treatment” of the American
penny and nickel coins. The purpose of these regulations is “to safeguard against a potential shortage of
these coins in circulation.” The regulations make it
illegal to export, melt, or treat one-cent and five-cent
coins of the United States, except in some cases or
with the Secretary of the Treasury’s explicit permission.1
Our pennies and nickels, it turns out, are threatened with extinction by melting. Why that is the case,
and what can be done about it, is the subject of this
article. As inflation erodes the value of money, a coin
of a given denomination (say, one cent or five cents)
loses value. But coins are made of a physical material
whose intrinsic value is usually low relative to the
value of the coin, yet not negligible. Every now and
then, we reach a point where the market value of the
coin (its purchasing power) drops close to or below
the intrinsic value of the materials used to make it.
Our pennies and nickels have now reached that point.
This has two consequences. One is that the Mint
is producing these coins at a loss. It now costs 1.67 cents
to make a penny and 5.97 cents to make a nickel. The
other is that it can be profitable to melt down the coins
and recycle their metal content. The Mint’s regulations
were announced because we are close to the melt-down
point for pennies and nickels.
The problem we are now facing is infrequent but,
in many ways, a very old one. Seven hundred years
ago (in the statute of 1299), the Parliament of England
enacted that “no good money of silver, of the king’s
coin or other, nor any silver in plate or otherwise,
should go forth or be carried out of the Realm or out
Federal Reserve Bank of Chicago
of the King’s power into foreign parts without especial leave from the king” (Ruding, 1817–19, Vol. 1,
p. 385). This was the first of many such prohibitions—
sometimes under penalty of death.
These prohibitions were passed at a time when
money was different from ours, that is, when it was
made of precious metals like gold and silver. Our money
does not derive its value from its intrinsic content,
which should be immaterial. In this article, I will first
explain how a medieval problem can reappear in modern
times. I will provide a quick overview of the history
of American coinage, highlighting earlier instances of
such problems, in particular the coin shortages of 1964
and 1965, and what solutions were adopted then. I will
then discuss possible remedies to our current situation.
Historical background
The economy needs money to operate. Money is
commonly described as having two functions: a unit
of account in which prices and obligations are denominated and a medium of exchange in actual transactions.
The two functions are logically distinct, but typically
the unit of account has been tied to an actual medium
of exchange. Coined money—that is, standardized
quantities of metal shaped into a convenient form for
everyday use—was introduced in Europe in the sixth
century BC, and has served as a medium of exchange
for almost all of subsequent history; and the unit of
account has consequently been tied to the metal or
metals coined.
In medieval Europe, where our modern system
has its roots,2 the metal was silver; the coin was the
penny, made of silver alloyed with a little copper for
François Velde is a senior economist in the Economic
Research Department at the Federal Reserve Bank of
Chicago. The author thanks his colleagues in the Economic
Research Department for helpful comments.
17
convenience. Governments set the standards by deciding how much metal went into a penny. The quantity of money was determined by the private sector, in
the following way. If more money was needed, metal
was brought to the mint and transformed into new coins,
usually for a fee called seigniorage. If less money was
needed, money was melted down and the metal turned
to other uses. The signal for minting or melting was
given by the price level (the inverse of the value of
money): If silver in the form of coins was too cheap,
it was profitable to turn it into bullion; if it was too
expensive, it was profitable to sell bullion to the mint
and acquire new coins. These two actions (and the
equivalent actions of importing and exporting coins)
served to regulate the price level.
The system worked well with one coin. But the
growing needs of trade led to the introduction of larger silver coins and later even more valuable gold coins;
and with multiple coins, the system does not work as
well. The reason is that smaller coins are more expensive to make, in proportion to value, than larger ones.
Mints were not subsidized and had to recover their
production costs. This created a wider gap between
minting and melting points for small coins than for
large coins: The value of small coins had to go up higher
before minting new ones became profitable (net of
production costs). This led to a dilemma. If the mint
bought silver for the same nominal price whether it
paid in large or small coins, small coins had to contain less silver relative to their value and large coins
might disappear and be melted down for their content. If the mint made all coins full-bodied, but charged
more for small coins, large coins would be produced
but not small coins, even when they were needed.
The Middle Ages were plagued with difficulties
in maintaining an adequate supply of all denominations (Sargent and Velde, 2002). One common response
to a shortage of one denomination was to prohibit the
melting or exporting of the coins in short supply. The
English statute of 1299 was an early example. It was
followed within a few years by many other such statutes—a clear indication that such measures were difficult to enforce and had limited effect.
Another short-term solution was to debase the
coin in short supply. Debasing a coin meant reducing
its intrinsic content—for example, putting less silver
in each penny. For a given market value of silver, debasement of one denomination can make it profitable
to mint it again. In the case of medieval England, the
cycle of melting prohibitions begun in 1299 led to a
debasement of silver money in 1343. A debasement
would restore the supply of the scarce denomination
for a while, but inevitably shortages reappeared and
18
further debasements followed. This repeated process
led over time to pennies containing less silver and
more copper, so much so that by the late eighteenth
century, British (and American) pennies were made
of pure copper.
A long-term solution was to return to the singlecoin system, preserving the traditional minting and
melting mechanism for one large gold coin and making the other coins token—that is, worth substantially
more as money than as metal. The large coin pegged
the value of the unit of account to a particular commodity, as in any commodity money system. Smaller
denominations, however, were fiduciary; that is, their
value in circulation was significantly higher than that
of their intrinsic content. Their value came not from
their content, but from a policy of convertibility: The
authorities stood ready to exchange subsidiary coinage for gold coins, and vice versa. The provision of
token coins was then left to the government, which
bought and sold token coins on demand and made a
profit from the substantial difference between face
value and content. This is called the gold standard,
and it became the norm, after much experimentation,
in most countries by the end of the nineteenth century.
The U.S. monetary system, which Congress has
sole power to regulate, began in 1792 as a bimetallic
system.3 This is a system in which silver and gold
coins are provided by the minting and melting mechanism, and both coins play the same role as anchor of
the monetary system. The founding fathers did not innovate at all in monetary matters. The bimetallic system
was commonplace in Europe (though not in the mother
country of Great Britain). The very mixed record of
paper money during the colonial era, as well as the
decidedly disastrous experience of the Continental
money issued to finance the American Revolution,
had predisposed the U.S. government to adhere to a
commodity money system throughout the denomination structure. Even the smallest coins, the cent and
half cent, were made of copper but were not token,
which led to various problems. Copper was not that
valuable and a cent’s worth of copper was inconveniently
large. Also, the world price of copper was volatile
(because of its military uses), and it was difficult to
maintain the cent at a fixed parity of 100:1 with the
silver dollar. A similar problem arose from fluctuations in the relative price of gold to silver, which led
to periods when no gold coin or no silver coin was
minted and which prompted one debasement in 1834.
Prompted by the same forces as other countries,
the U.S. gradually moved to a gold standard by making the smaller coins token. The first step was in
1853, when the silver content of quarters and dimes
1Q/2008, Economic Perspectives
was reduced relative to the silver content of the dollar
coin. This proved difficult to enact, as there was reluctance on the part of many legislators to issue token
money. Then, in 1873, silver dollars ceased to be minted on demand. The market value of silver fell substantially so that silver dollars became tokens too. The U.S.
formally adopted the gold standard in 1900. Smaller
coins were made of silver (the quarter and dime), nickel and copper (the nickel), or a copper alloy (the penny). The value of the silver in a quarter was around
10 cents. A quarter was worth 25 cents because the
U.S. Department of the Treasury was always willing
to exchange 40 of those coins for a gold $10 coin.
When the Federal Reserve System was created in
1914, the U.S. remained on a gold standard because
Federal Reserve notes were redeemable on demand
into gold at a fixed parity of $20.67 per ounce. The
Great Depression, as well as the perceived need to increase the money supply without constraints to stimulate the economy, led to drastic changes. The gold
content of the dollar was reduced by 40 percent, private holdings of gold by U.S. citizens were prohibited,
and the Federal Reserve notes ceased to be redeemable on demand. The U.S. was on its way to a fiat
money system (one in which money has value by fiat,
that is, because the monetary authority or the government decrees it). After World War II, the Bretton Woods
system restored a semblance of the gold standard,
with foreign currencies convertible into dollars and
dollars convertible into gold for foreigners. This lasted until 1971, when President Nixon closed the gold
window and permanently severed the tie between the
dollar and any commodity.
What about smaller denominations? In 1934, the
Silver Purchase Act was passed, requiring the Treasury
to purchase silver with the goal of reaching either a
market price equal to its “monetary price” of $1.29 or
a certain proportion of the monetary stock. The reasons
for this action were complex: The issue of silver certificates in exchange for the silver purchased was to
provide an additional avenue for increasing the money
supply. Also, strong pressures from western states where
silver was mined played a role in the legislation.
The market price of silver in late 1933 was 44 cents
an ounce, and during the following years, the U.S.
Department of the Treasury bought silver at abovemarket prices, between 50 cents and 77.5 cents an
ounce, and after 1946 at 90.5 cents an ounce, accumulating a stockpile of 3,200 million ounces. By 1955,
however, the world price of silver had risen to the
Treasury’s purchase price, and the Treasury began
selling its silver, as it was authorized to do under
existing legislation. Prices remained pegged at the
Federal Reserve Bank of Chicago
Treasury’s price of 90.5 cents an ounce, and the
stockpile of silver that was not held to back silver
certificates dwindled until November 1961, when
President Kennedy stopped the sales. The price of
silver started rising, and it reached the monetary
price in September 1963.
At that price, the metallic content of dimes, quarters, half dollars, and dollars was exactly equal to their
face value. Anyone needing silver for industrial uses
could readily buy it on the commodities market as
bullion or buy it from the banking system in the form
of coins and melt them down.4 As world supply and
demand factors kept exerting upward pressure on prices,
the U.S. monetary stockpile was drawn down, in various ways, either by redemption of silver certificates
or else by the United States Mint working overtime
to meet the “demand” for quarters and dimes. In early
1963, Treasury officials estimated that their silver
supply would last 20 years. But the demand for subsidiary coinage proved unexpectedly strong, and the
Mint’s annual production quadrupled from 1963 to
1964. This was attributed initially to the growing use
of vending machines, but it became clear that much
of this demand was speculative: The public was buying the Treasury’s stockpile at $1.29 an ounce in expectation of exhausting it and seeing the market price
rise above the value they had paid. The Senate held
hearings on the question in April and August of 1964
but came to no conclusion. The Treasury conducted
its own studies and recommended in February 1965
that the silver content of subsidiary coinage be reduced
or eliminated. In the end, following the recommendation
of the Treasury studies, President Johnson proposed
to Congress new legislation in June: It was swiftly
voted into law and signed as the Coinage Act of 1965.5
The new law provided for the minting of the quarters and dimes made of copper and nickel (or cupronickel) that we know. The half dollar was replaced
with a 40 percent silver core clad in copper and nickel.6
The new quarters were issued in November 1965, by
which time the reports of coin shortages had disappeared; dimes and half dollars followed in March 1966.
At the signing ceremony on July 23, 1965,
President Johnson made curious remarks: “Some
have asked whether our silver coins will disappear.
The answer is very definitely no. … If anybody has
any idea of hoarding our silver coins, let me say this.
Treasury has a lot of silver on hand, and it can be,
and it will be used to keep the price of silver in line
with its value in our present silver coin. There will be
no profit in holding them out of circulation for the
value of their silver content.”7
19
Indeed, the government’s intention was not to replace silver dimes and quarters with cupronickel
dimes and quarters, but only to reduce global demand
for silver by removing the United States Mint from
the ranks of the buyers. But keeping the existing stock
of silver dimes and quarters in circulation was possible only if the price of silver did not rise above $1.29
an ounce. To achieve this, the Treasury had two means.
One was its large stockpile of silver. The other was
the authority given by the Coinage Act of 1965 to prohibit the melting and exportation of coins when necessary. The Treasury used both means in succession.
First, for two years it sold silver at $1.29 an ounce, the
price at which a quarter’s content was worth 25 cents.
The silver stockpile went from 1,200 million ounces
in 1964 to 350 million in 1967. Then, using its new
powers under the Coinage Act of 1965, it banned the
melting, treatment, and export of silver dimes and quarters on May 20, 1967. Soon after, the Treasury stopped
supplying silver at a fixed price on July 14, 1967, the
day on which silver became “just another metal.”
The prohibition met with some negative reactions
in Congress, where two representatives introduced
bills to repeal it, without success.8 Although the ban
was enforced and resulted in several indictments,9 it
did not prevent the disappearance of silver quarters
and dimes from circulation. Silver half dollars had
virtually disappeared from circulation by early 1966,
and there were already reports of “culling” by consumers, that is, people picking out silver coins from
their change and paying out only clad quarters.10 By
June 1968, the Treasury was itself melting silver quarters in its vaults, using new electronic sorting machines.11
In August 1968, it was reported that dealers were
paying 12 percent above face value for silver quarters
and dimes. The combined forces of the Treasury and
private speculators rapidly removed the silver coinage from circulation, making the ban moot. It was
lifted in June 1969.
The provisions of the Coinage Act of 1965 were
used a second time—this time to protect the penny.
The peg to gold had ended in August 1971. Inflation
was rampant, and commodity prices were exploding.
On April 1, 1974, the price of copper reached a record
of $1.40 per pound. At the time, 154 pennies contained
one pound of copper. Although copper prices fell back
somewhat, the demand for pennies rose to suspiciously high levels. The Treasury concluded that hoarding
was under way in expectation that it would become
profitable to melt pennies, and it announced the ban
on April 18, 1974.
A few months later, Public Law 93-441 (31 USC
5112(c)) granted to the Secretary of the Treasury the
20
power to change the proportion of zinc and copper in
pennies to ensure adequate supplies. This gave the
Treasury the option to replace copper with zinc in the
composition of the penny, at its discretion. Copper
prices stayed below the penny’s melting point in subsequent years, so the ban was lifted in June 1978 without any further action. Soon, however, copper prices
rose again and hit another record of $1.44 per pound
on February 12, 1980. The Treasury briefly considered
another ban, but instead used its statutory authority to
change the composition of the penny, almost reversing the proportions. The Mint announced in June
1981 that, instead of 95 percent copper and 5 percent
tin and zinc, pennies would be primarily zinc with a
coating of copper; production started early the following year. As in 1965, no effort was made to retire
the older coins: They were allowed to remain in circulation side by side with the new pennies.
Most people do not know that all pennies are not
the same. Lincoln’s profile has been unchanged since
1909. But take a penny dated 1983 or later and scratch
its surface; you will see the shiny white zinc underneath the copper coating. As for the nickel, its size
and composition have not changed since 1866. The
effort to maintain the outward appearance of the coinage suggests the importance of habits in our attitudes
toward coinage and currency.
The current situation
Between 1982 and 2004, the price of copper
surged to the level of $1.50 per pound a few times,
briefly. But in late 2004 it reached that level once
more and has not come down since. Other commodities have surged in value as well, notably zinc and
nickel. Table 1 shows the current value of the metal
contained in U.S. coins.
The values shown in table 1 do not properly measure the profit to be made by melting down the coins.
It would be necessary to subtract melting and refining
costs (scrap copper is worth about 20 percent less than
high-grade copper whose price is used in table 1).
Collecting and shipping the coins for melting would
impose additional costs, and those costs would be relatively larger for the smaller denominations, since
digging a penny or a nickel out of a sofa requires the
same effort.
Nevertheless, in 2006 some businesses became
interested in the activity and inquired with the Mint
about the legality of melting down coins. One firm in a
midwestern state even began buying pennies from banks
and sorting them to extract pre-1982 copper pennies.
When the regulations were issued in December 2006,
the Treasury had good reason to think that melting
pennies and nickels was close to being profitable.
1Q/2008, Economic Perspectives
the market value of the tokens we use in physical transactions should be of no consequence
Intrinsic value and composition of U.S. coins, 2007
to their value. The problem of small change
Coin
Composition
Intrinsic value
was a difficult one to solve under a commodi
(percent
(percent of
ty money regime, but in a fiat money regime
of metal)
face value)
shortages of small change should not occur.
Penny
95 zinc, 5 copper
69.7
The value of pennies and nickels has reached
Penny (pre-1982)
95 copper, 5 zinc and tin
209.5
the floor set by their intrinsic content. We are
Nickel
75 copper, 25 nickel
136.2
printing our money on needlessly expensive
Dime, quarter, and
Susan B. Anthony dollar
75 copper, 25 nickel
20.9
material.
Golden dollar
88.5 copper, 2 nickel, and
The historical overview also shows that
3.5 manganese
5.7
this problem is not new. Figure 1 shows the
Note: These are data as of November 14, 2007.
value, as a percentage of face value, of the inSources: Author’s calculations based on data from the United States Mint
trinsic content of coins minted every year
and Haver Analytics.
since 1825 for three denominations. For all
three types of coins (the penny, nickel, and
quarter), there is over time a general upward
The nature of the problem
trend; every time the value comes close to 100 perThis brief historical overview frames the probcent, it becomes necessary to change the composilem, which is something of a paradox.
tion, which has the effect of abruptly lowering the
A fiat money system is one in which money has
value of the intrinsic content. The quarter began at
value by fiat, that is, because someone said “let it be
100 percent because it was a full-bodied coin, but in
so.” Economists like to describe money in their mod1853 it became a subsidiary coin, overvalued relative
els as “intrinsically useless pieces of colored paper”
to its silver content.12 Figure 1 clearly shows what
because the challenge for monetary economics is to
happened in 1965, when its composition was changed
explain the value of such objects. For objects that are
from silver to cupronickel. The line for the nickel
not intrinsically useless, we have standard price theodisplays a sharp uptick during World War II. At that
ry. For claims on objects that are not intrinsically usetime, nickel being needed for the war effort, nickels
less, we have finance theory.
were made of silver. These coins swiftly disappeared
Since at least 1971, the U.S. has operated under a
in the early 1960s when the price of silver began to
pure fiat money system, in which the intrinsic value
rise. Finally, the penny’s line falls sharply in 1982
of the objects used as a medium of exchange should
with the switch from copper to zinc.
not matter. This is in stark contrast with the commodThe authority that the Secretary of the Treasury is
ity money regime of 1900. In that regime, the intrinusing today to prohibit the melting and exportation of
sic content of coins provided a floor below which the
pennies and nickels was granted during the shortage
value of coins could not fall, and minting on demand
of quarters and dimes in 1964–65. This authority was
provided a ceiling above which it could not rise. The
used to protect pennies in 1974. In each instance
gap between floor and ceiling was usually fairly
when the intrinsic value of the coin exceeded its face
small. Under a fiat money regime, the ceiling is revalue, the long-term solution was to change the commoved, as there is no minting on demand. The floor is
position of the threatened coin.
normally of no consideration because no one pays
Logic suggests, and history shows, that prohibimuch attention to the content of coins (copper pentions on melting will not solve the problem. If it is renies and zinc pennies circulate at par, although the
ally profitable to melt pennies or nickels, people will
content of the former is twice as valuable as the condo it. The ban imposed in 1967 was lifted in 1969 betent of the latter). The stock of money, and its value,
cause the coins it was designed to protect had disapis determined not by minting and melting, but by the
peared. Such stopgap measures at best increase the costs
monetary authority’s policy. In this respect there is no
of melting by a small amount—the probability of bedifference between notes and coins. The value of a
ing caught times the penalties imposed. Devoting enough
dollar bill has nothing to do with its alternative uses
law enforcement resources to increase the probability
as wallpaper or insulating material. Pennies and nickof catching penny smelters hardly seems worthwhile.
els are like notes, except they are made of something
Alternatively, speculators can simply hoard the coins
more durable than paper.
and incur time and storage costs as they wait for the
Now that all our currency is fiduciary (that is,
regulations to be repealed. Those costs are real, but
with a market value higher than the intrinsic value),
Table 1
Federal Reserve Bank of Chicago
21
figure 1
Intrinsic value of U.S. coins, 1825–2006
percentage of face value
Penny
Nickel
Quarter
100
80
60
40
20
0
1840
’60
’80
1900
’20
’40
’60
’80
2000
Note: No quarters were produced from 1861 through 1870.
Sources: Author’s calculations based on data from the United States Mint; and the U.S. Department of the Interior, U.S. Geological Survey.
they are modest compared with potential movements
in commodity prices.
What drives this long-term trend in the intrinsic
content of coins? Inflation is the answer. Although it
was not much of a force in the nineteenth century (the
price level was about the same in 1913 as in 1825), in
the twentieth century it has been the main culprit. Money
steadily loses its value relative to other goods, including the goods with which it is made. In other words,
the floor on the value of coins is always creeping up,
however slowly. In countries with high levels of inflation, the process can be rapid, and coins become
obsolete in a matter of a few years. Our relatively low
inflation in the U.S. means that these problems occur
relatively infrequently, but they do occur.
The upward trend can be accelerated if metals
rise in price faster than other goods. Figure 2 plots
the real price of several metals that have been used in
coins, deflated by the Consumer Price Index. The evidence is rather mixed. For some metals, such as aluminum, the secular trend is clearly downward. For
other metals, there are long cycles—for example, the
rise in the 1970s and the fall in the 1980s and 1990s.
Since 2000, however, all metals have shown a sharp
increase. The recent surge in commodity prices may
arguably be speculative, and prices could well come
down again, lowering the floor for a while. But as
long as inflation is positive, the real value of a penny
22
(which is always $0.01 in nominal terms) will fall
relative to goods and services. When zinc replaced
copper in the manufacture of pennies in 1982, the respite gained was relatively brief, since zinc was only
half as costly as copper. Since zinc pennies were introduced, the real value of the penny (as measured by
inflation) has fallen by half. Even if commodity prices stabilize, a 2 percent annual inflation rate will reduce the real value of the penny by another one-third
over the next 20 years, and the problem will inevitably return unless another metal is found to replace
those used in pennies and nickels.
Replacing the metal is not easy. As the law currently stands, the United States Mint has no authority
to change the composition of the nickel and can only
use copper and zinc for pennies. The Mint is nevertheless investigating alternatives. Finding a cheap
metal is not enough: It must be easy to mint and must
not present health risks, be allergenic, or wear out too
quickly in circulation. Other countries, such as Canada,
the United Kingdom, and those of the eurozone, have
found steel a convenient substitute for other metals
in the one-cent coin. Steel was used for the U.S.
penny during World War II and was considered as
an alternative in the 1970s. It has the advantage of
being cheaper than other metals that have been used
historically, such as aluminum (which was also considered in the 1970s), tin, and lead. New Zealand’s
1Q/2008, Economic Perspectives
figure 2
Real price of metals used in U.S. coins, 1825–2006
index, 1913 = 100, log scale
900
800
700
600
500
400
Gold
Silver
Zinc
Copper
Nickel
Tin
Aluminum
300
200
100
90
80
70
60
50
40
30
20
10
1840
’60
’80
1900
’20
’40
’60
’80
2000
Notes: The data plotted are three-year moving averages. The values are deflated by the Consumer Price Index. There are no data available
for silver from 1860 through 1871.
Sources: Author’s calculations based on data from the U.S. Department of the Interior, U.S. Geological Survey; and Carter et al. (2006).
coinage now consists solely of steel cores, plated for
aesthetic reasons with other metals and produced by
the Royal Canadian Mint.
Even if a suitable metal is found, however, it will
be difficult to produce pennies without taking a loss
because production costs other than the metal were
already 66 percent of face value in 2004 (see table 2).
The Royal Canadian Mint is able to produce its penny for 0.8 cents.13
two pennies from the saucer next to the register and
hands them to me, and I return them to the saucer.
The transaction is the same as if the cashier rounded to
$2.00, except for a little side game between me and
the cashier involving copper-colored tokens.
That I and the cashier are willing to give away
the pennies in the saucer suggests that the penny isn’t
worth much. One way to see this is to measure the
Should we eliminate the penny?
A simpler alternative is to let the penny
melt out of existence. After all, do we need
the penny?
The penny’s role in our economy is not as
a medium of exchange. There is nothing that a
penny buys: Dime stores have long ago been
replaced by dollar stores. Almost no coin-operated machinery accepts it.14 We don’t even
use it truly to make change. It is merely a
symbolic counter to simulate remainders of a
division by five in retail transactions. When I
buy a cup of coffee and the price comes out to
$1.98, I give two dollar bills, the cashier takes
Federal Reserve Bank of Chicago
Table 2
U.S. coin production costs and profits, 2004
Coin
Penny
Nickel
Dime
Quarter
Half dollar
Golden dollar
Costs
Metal
Other
Total
( - - - - percent of face value - - - - )
27
56
9
9
9
2
66
35
22
20
25
19
93
91
31
29
34
21
Profits
($millions)
2
6
170
424
2
4
Notes: Metal cost is based on average metal prices for 2004. The profits are
calculated from coin production numbers for 2004.
Sources: Author’s calculations based on data from the United States Mint,
United States Mint Annual Report 2004; and the U.S. Department of the
Interior, U.S. Geological Survey.
23
What figure 3 does show is that there is a
wide range across countries in terms of the
Value of the smallest circulating coin compared across
value of their smallest denomination. That is
countries, 2003
in part because, in recent years, a number of
value measured in seconds of a worker’s time
countries have abandoned their smallest de20
nominations. In Australia and New Zealand,
�
whose dollars are comparable in value to the
18
�
U.S. dollar, one-cent coins were also made es16
�
sentially of copper. In 1987, the rise of copper
�
prices made the one-cent coin unprofitable to
14
mint. Instead of changing the content, New
�
12
Zealand stopped producing its one-cent and
10
�
�
two-cent coins (worth about 0.5 cents and one
�
�
�
cent in U.S. currency, respectively) in March
�
8
�
�
�
�
1989, and they ceased to be legal tender in
6
�
�
April 1990. The coins were bought back by
�
�
4
�
the Reserve Bank of New Zealand and melted
� �
�
� � ��
�
down for scrap metal.15 Australia followed
�
�
�
2
��
�
suit, stopping production of the coins in
�
0
�
�
August 1990 and issuance in February 1992.16
2
3
4
5 6 7 8 9 10
20
30
gross domestic product per capita in thousands of Geary-Khamis dollars
New Zealand went further in 2006: Existing
five-cent, ten-cent, 20-cent, and 50-cent coins
Notes: The sample includes the OECD (Organization for Economic
ceased to be legal tender, and all but the fiveCooperation and Development) countries plus other European countries.
For details on Geary-Khamis dollars, see Maddison (1995).
cent denomination were replaced with smaller
Sources: Author’s calculations based on data from the Organization for
and cheaper coins of plated steel.
Economic Cooperation and Development, International Labor Organization,
national mint websites, and Maddison (1995).
In the eurozone, the smallest euro denominations are the one-cent (currently worth
about 1.5 cents in U.S. currency) and two-cent
coins. Each country can mint its own coins (with a compenny with the value of time. Median weekly earnmon European obverse and nationally designed reverse),
ings for wage earners and salaried workers are $675.
and all coins are legal tender throughout the eurozone.
Assuming a 40-hour workweek, it takes most U.S.
Two countries, the Netherlands and Finland, opted
workers no more than two seconds to earn a penny.
not to issue one-cent and two-cent coins at all, and
Rounding transaction prices to the nearest five cents
they officially encourage rounding to the nearest five
would save more than the time we spend fishing for
cents within their borders. Outside of the eurozone,
pennies in our pockets or wallets.
the Czech Republic and Slovakia have recently elimiA comparison with other countries is instructive.
nated their two smallest coins, and Hungary plans to
Figure 3 compares the values of the smallest circulatdo so next year.
ing coins in about 30 countries—mostly the OECD
The penny is disappearing of its own accord in
(Organization for Economic Cooperation and Develeconomic terms. Various interest groups (for examopment) countries plus other European countries. The
ple, zinc producers, charities, and the state of Illinois)
values of each coin are again measured in the time it
can point to continued support for the penny shown
takes to earn it at the average wage in manufacturing.
in polls.17 But the United States Mint’s annual output
The values are plotted as a function of gross domestic
of pennies, nickels, dimes, and quarters as a ratio of
product (GDP) per capita, measured in Geary-Khamis
GDP tells a different story (see figure 4). While the
dollars (Maddison, 1995). There seems to be a small
relative importance of 25-cent coin output has been
negative relationship. However, this relation is not
stable over the past 30 years, that of the other coins
very robust and is largely due to the recent adoption
has been declining steadily. Relative to GDP, the outof the euro as the common currency by the relatively
put of pennies is 12 percent of what it was in 1982.
rich European countries (the same figure in 1999,
The trend is not much better for the nickel.
right before the introduction of the euro, shows no
So a penny isn’t worth much and the quantities
significant relationship between the value of small
produced are declining relative to GDP, but we still
coins and GDP per capita).
figure 3
24
1Q/2008, Economic Perspectives
figure 4
Ratio of U.S. coin production to gross domestic product, 1946–2006
log scale
–3
10
Penny
Nickel
Dime
Quarter
–4
10
–5
10
–6
10
1950
’60
’70
’80
’90
2000
Note: The data plotted are three-year moving averages.
Sources: Author’s calculations based on data from the United States Mint and U.S. Bureau of Economic Analysis.
produce a lot of them. Since 1982, the Mint has produced 910 pennies for every man, woman, and child
in America. It estimates that 100 billion pennies currently circulate. In 2006, the Mint used 20,000 tons
of zinc, worth $60 million, to produce pennies. Even
if the Mint (and the taxpayer) were not losing money
on this activity, it would be fair to ask whether all that
zinc might be put to better use than manufacturing
throwaway tokens.
The declining value of the penny is not a temporary phenomenon. It is a trend driven by several factors. One, noted previously, is inflation. The penny
has been part of our denomination structure since the
beginning, in 1792, but the price level has gone up by
a factor of 20 in the past century: A penny today is
worth one-twentieth of a penny before World War I.
If people got by without coins as small as 0.05 cents
back then, we can probably do so today. A second
factor is that, even in the absence of inflation, a penny
means less over time because we are becoming richer. As productivity grows, a penny will be worth ever
less of our time because our time is more productive.
A third factor is the replacement of cash (coins and
notes) by other means of payment, notably electronic
ones. Just as there was a boom in the demand for
coins in the 1950s and 1960s because of the spread of
Federal Reserve Bank of Chicago
coin-operated machinery, we can expect technological change to affect the demand for coins in the future.
These factors together tell us that the penny will
disappear sooner or later, as did the farthing (onequarter of a penny) and the ha’penny (one-half of a
penny) of medieval England, and our own half cent,
last minted in 1857.
Moreover, the experience of other countries suggests that there are few problems involved in doing so.
The Reserve Bank of New Zealand has not found any
evidence of inflation or upward rounding since it withdrew its one-cent and two-cent coins. The Royal
Canadian Mint recently published survey results indicating that small retailers were vastly in favor of removing the penny, and consumers were split on the issue.
Current legislative proposals
As I noted earlier, the solution to our problem of
small change is constitutionally vested in the hands
of Congress, and some legislation is on the agenda.
Two bills were introduced in Congress in early
August 2007.18 Both bills confer on the Secretary of
the Treasury the power to “prescribe the weight and
the composition” of existing denominations, considering “such factors that the Secretary considers, in the
Secretary’s sole discretion, to be appropriate.” A third
25
bill introduced in October 2007 includes a similar
provision.19
Delegating such power to the Secretary of the
Treasury would represent a significant change. Ever
since the Coinage Act of 1792,20 Congress has retained
for itself the exercise of its constitutional powers to
“coin money, regulate the value thereof.” There were
good reasons for the founding fathers to assign such
powers to Congress. Under a commodity money system
(the only system they could conceive for our country),
setting the weight and composition of coins is the essence of monetary policy and is therefore an extremely
important power. Recent European history, with which
they were familiar, gave them reason to be wary of
handing over monetary policy to the executive branch.
But things have changed. The composition of
coins is not central to monetary policy anymore. Under
a fiat system, it is a purely technical issue, whose
only potential consequence for the legislature is the
profit or loss made on coining.
Profit on coinage, of course, is not negligible.
Table 2 (p. 23) shows that, on some coins, the profits
can be substantial. The high figure for the quarter reflects the success of the “state quarters” program, which
has generated $3.2 billion in “above-average” profits
on this denomination over eight years.
But profits can rapidly turn into losses. The United
States Mint made a small profit ($5 million) on pennies
and nickels in fiscal year 2005, but this turned into a
loss of $33 million in fiscal year 2006 and a loss of
almost $100 million in fiscal year 2007.
Congress therefore retains an interest in the issue
of coin composition, but it could nevertheless delegate
the details to the executive branch (namely, the U.S.
Department of the Treasury) because the issue is purely
technical and because action in the executive branch
will be timelier than passing new legislation each time.21
A medieval solution to a medieval problem
In a recent Chicago Fed Letter, I made a different
proposal.22 Starting from the observation that there
are many pennies in circulation but they are not really
needed as one-cent coins and inspired by medieval
debasements, I proposed that the prohibition on melting should be repealed and that pennies should henceforth be worth five cents.
In this proposal, the existing nickels would disappear and be melted down, which seems likely to be
their fate under any conceivable proposal. Pennies
would then be recycled as five-cent coins, avoiding
the need to design and produce a new coin (a lengthy
process). Since the Mint has produced about seven
26
times as many pennies as nickels in the last 20 years,
there should be enough pennies to serve as five-cent
coins for a while.
The new value would be easily established by the
monetary authority standing ready to exchange 20
pennies for a dollar bill, instead of 100 pennies presently. It is true that vending machines and other coinoperated equipment currently accepting nickels would
have to be modified to accept pennies as five cents.
But such modifications may be unavoidable if the
nickel in its current form is doomed.
I call this a medieval solution because medieval
debasements were sometimes carried out in this manner.
When a coin was threatened by melting, as is now the
case with our penny, there were two ways to debase
it: One was to mint it with less metal than before, and
the other was to increase its face value. Thus, in 1269
Venice increased the face value of its grosso coin from
26 to 28, and again in 1282 to 32, each time leaving
its composition unchanged. As I recently found out,
the idea also has precedent in U.S. history. During the
1965 silver coinage crisis, Congressman Craig Hosmer
(a Republican from California) proposed to “arbitrarily double the value of existing silver coins” in order
to save them from being melted down.23
The proposal would require everyone to ignore
the inscription on the penny that says “one cent.” But
there is also precedent for U.S. coins being worth
more than what is written on them. In 1834, when the
gold–silver ratio was adjusted, half eagles minted before that date and bearing the inscription “5 D” (five
dollars) were declared to be “receivable in all payments
at the rate of 94 and 8/10ths of a cent per pennyweight,”
which works out to $5.33 for a full-weight coin.24 This
was nothing else than a debasement, albeit a relatively modest one.
Would such a measure be inflationary? The estimated stock of pennies is 100 billion, so increasing
their value to five cents would add $4 billion to the
money supply, which represents 0.5 percent of the
monetary base or 0.3 percent of M1 (a monetary aggregate composed of currency and demand deposits).
This is a modest addition. The average monthly increase in the monetary base over the past three years
has been about $2 billion; the monthly standard deviation of M1 is about $6 billion over the same period.
Thus, an addition of $4 billion would fall well within
the range of typical monthly variations in the money
supply. The one-time increase would also be offset by
reduced issues of other coins and thus unlikely to
have a noticeable impact.
1Q/2008, Economic Perspectives
Conclusion
To prevent a shortage of small change, the U.S.
Department of the Treasury recently enacted regulations to prohibit melting and exportation of pennies
and other coins. The threat of shortage arises because
pennies and nickels are made of inappropriately expensive material, and there is or soon will be a profit
to be made from transferring their content to alternative uses.
There is $1 billion worth of resources sitting in
cash registers, jars, and sofas across the United States.
It makes little sense to keep replenishing them, and
the regulations hold little promise of forestalling the
inevitable very long. The traditional solution since
medieval times is to “debase” the threatened coin, that
is, make it of a cheaper material or assign it a higher
face value, either of which requires congressional action. But the current situation may well prompt a more
general debate on whether such small denominations
are worth saving—a debate that is ongoing in many
other industrialized countries.
notes
The regulations became permanent on April 16, 2007, and now
constitute 31 CFR Part 82 (Federal Register, April 16, 2007). By
law (31 USC 5111 (d1)), the Secretary of the Treasury “may prohibit or limit the exportation, melting, or treatment of United States
coins when the Secretary decides the prohibition or limitation is
necessary to protect the coinage of the United States.” One of the
exceptions to the regulations allows Federal Reserve Banks and
depository institutions to continue exporting coins for circulation
in “dollarized” countries, such as Ecuador and Panama.
1
The word “penny” itself goes back at least to the ninth century.
2
See Carothers (1930).
3
This neglects refining costs: Coins consisted of silver at 90 percent
purity mixed with copper.
The break in the quarter series in figure 1 is related to another
shortage of small change—this one prompted by the introduction
of fiat money in the form of “greenbacks,” notes that were not redeemable into gold or silver. During the subsequent period of inflation, from 1861 through 1870, the dollar price of silver made it
unprofitable to mint silver quarters, while existing quarters were
hoarded or melted.
12
13
Branswell (2007).
One notable exception is the acceptance of pennies in automatic
toll lanes on Illinois roads. Until a few years ago parking meters in
downtown Hilo, HI, accepted pennies, but parking is now free.
14
15
APN News and Media (1990).
16
Glover (1992).
17
Hagenbaugh (2006).
18
HR 3330 and S 1986.
19
HR 3956.
Times Mirror Company (1965).
20
1 Statutes at Large 246.
Cabeen (1967).
21
4
The Coinage Act of 1965 is also known as Public Law 89-81 (79
Stat. 254).
5
The silver core was abandoned in 1971; at the same time the
Eisenhower dollar, also made of copper and nickel, was introduced
to replace the silver dollar discontinued in 1964.
6
7
8
Three Manhattan jewelry technicians were arrested in December
1967. Three men were arrested near Tucson, AZ, with two tons of
dimes and quarters and a small smelter in April 1968; two men
were arrested in Brooklyn and arraigned in December 1968
(Laurence 1968; Dow Jones and Company, 1968a, b).
9
10
The Reserve Bank of New Zealand is vested with the power to
“determine the denominations, form, design, content, weight, and
composition of its bank notes and coins,” according to the Reserve
Bank of New Zealand Act 1989, s. 25(2). Thus, the recent decision
to abolish the five-cent denomination was taken by the Reserve
Bank of New Zealand, without any legislation.
22
Velde (2007).
23
Foley (1965).
24
4 Statutes at Large 699, section 3.
Janssen (1966).
Times Mirror Company (1968).
11
Federal Reserve Bank of Chicago
27
REFERENCES
APN News and Media, 1990, “Small coins worth
$3m are sold as scrap,” New Zealand Herald,
April 26, p. 9.
Branswell, Jack, 2007, “Is penny lane a dead end?,”
Montreal Gazette, October 9, p. B1.
Cabeen, Richard McP., 1967, “Rules on handling
silver coins may be due for a change,” Chicago
Tribune, August 27, p. F8.
Carothers, Neil, 1930, Fractional Money: A History
of the Small Coins and Fractional Paper Currency of
the United States, New York: John Wiley and Sons.
Carter, Susan B., Scott Sigmund Gartner, Michael
R. Haines, Alan L. Olmstead, Richard Sutch, and
Gavin Wright (eds.), 2006, Historical Statistics of
the United States, Millennium ed., 5 vols., New York:
Cambridge University Press.
Dow Jones and Company, 1968a, “Two men are
seized, charged with illegal melting of coins,” Wall
Street Journal, December 5, p. 29.
Dow Jones and Company, 1968b, “Three men
charged with melting coins to reclaim silver,” Wall
Street Journal, April 30, p. 9.
Foley, Thomas J., 1965, “Silverless and lighter coins
asked by Johnson,” Los Angeles Times, June 4, p. 1.
Glover, Richard, 1992, “To coin a phrase, coppers
are a rum deal now,” Sydney Morning Herald,
January 30, p. 3.
28
Hagenbaugh, Barbara, 2006, “A penny saved could
become a penny spurned,” USA Today, July 7, p. B1.
Janssen, Richard F., 1966, “Gresham’s law faced by
Mint,” Wall Street Journal, February 16, p. 1.
Laurence, Michael, 1968, “Better than money.
Better than money?,” New York Times, August 25,
p. SM4.
Maddison, Angus, 1995, Monitoring the World
Economy, 1820–1992, Paris: Development Center
of the Organization for Economic Cooperation and
Development.
Ruding, Rogers, 1817–19, Annals of the Coinage of
Great Britain and its Dependencies, 3 vols., London:
Nichols, Son, and Bentley.
Sargent, Thomas J., and François R. Velde, 2002,
The Big Problem of Small Change, Princeton, NJ:
Princeton University Press.
Times Mirror Company, 1968, “Silver quarters go
to meet their melter,” Los Angeles Times, June 27,
p. E24.
Times Mirror Company, 1965, “New ‘sandwich’
coins bill signed by Johnson,” Los Angeles Times,
July 24, p. 3.
Velde, François, 2007, “What’s a penny (or a nickel)
really worth?,” Chicago Fed Letter, Federal Reserve
Bank of Chicago, No. 235a, February.
1Q/2008, Economic Perspectives
Corruption and innovation
Marcelo Veracierto
Introduction and summary
In this article, I illustrate how corruption can lower
the rate of product innovation in an industry. This is
important because, if many industries are subject to
corrupt practices, the lower rate of innovation would
result in a lower growth rate for the whole economy.1
Actually, the view that corruption is closely related to
economic development is widely held in practice: Poor
African countries, such as Kenya and Zaire, are commonly believed to lose a considerable fraction of their
gross domestic product (GDP) to corruption activities.
Figure 1 illustrates the extent of this perception. It
plots 2004 GDP per capita levels from the Penn World
Table against the 2004 Corruption Perception Index
constructed by Transparency International.2 Since a
Corruption Perception Index number close to zero indicates no corruption, figure 1 shows a clear negative
relation between corruption and economic development.
While a negative correlation between corruption
and GDP per capita levels is highly suggestive of an
actual link, it is not conclusive evidence. It may be the
case that corruption is closely related to other variables,
such as political instability, the extent of violence, or
the combativeness of unions, among other factors, and
that these other variables are the ones generating poor
economic development outcomes. In addition, GDP
per capita levels may be affecting corruption levels
and not the other way round. To complicate matters
further, the negative correlation between corruption
indexes and GDP per capita levels could be a mere
artifact: It may well be the case that low GDP per
capita levels are biasing the subjective perception of
corruption reported by survey respondents. To disentangle the effects of corruption on economic development, further analysis is needed.
In this article, I provide theoretical grounds for
pursuing such an analysis: In particular, I explore the
strategic interactions between producers and corrupt
Federal Reserve Bank of Chicago
officials. The basic corruption scenario considered involves three agents: an innovator, an incumbent producer, and a corrupt government official. The innovator
wants to enter business by potentially paying a bribe;
the incumbent producer wants to preclude the entry
of the innovator by potentially paying a bribe; and the
corrupt official decides on allowing the entry of the
innovator based on the bribes received. Key elements
of the game are that the government official can make
successive take-it-or-leave-it bribe offers to the producers and that the central government can never verify
the actual payment of a bribe (with some probability,
the central government can detect that the entry permit
was misallocated but cannot prove the actual amount
of the bribe paid). Under these assumptions and within certain ranges, I show that the amount of bribes
that the government official can collect can be very
responsive to small changes in the probability of detection or in the penalties imposed. In fact, the bribe
payments are shown to be a discontinuous function
of those variables. Since the resources devoted to innovation are continuously and inversely related to the
bribes that producers must pay, this means that the
amount of resources devoted to innovation is a discontinuous function of the probability of detecting
corruption and of the penalties imposed.
The rest of this article is organized as follows. In
the next section, I discuss the related literature. Then,
I describe the corruption game and characterize its
solution. Next, I analyze the implications of the corruption game for innovation decisions. Finally, I draw
some conclusions about my findings.
Marcelo Veracierto is a senior economist in the Economic
Research Department at the Federal Reserve Bank of
Chicago. The author thanks Marco Bassetto, Craig Furfine,
and seminar participants at the Federal Reserve Bank of
Chicago for their comments.
29
figure 1
Corruption and real gross domestic product per capita
Corruption Perception Index
14
12
10
8
6
4
2
0
0
10,000
20,000
30,000
40,000
50,000
60,000
real gross domestic product per capita, U.S. dollars
Notes: The Corruption Perception Index reported here is actually defined as 14 minus the Corruption Perception Index constructed by
Transparency International. The transformation is made to associate low values for the index with low levels of corruption.
Sources: Author’s calculations based on data from the University of Pennsylvania, Center for International Comparisons, Penn World Table;
and Transparency International, Corruption Perception Index.
Related literature
Systematic empirical evidence about the relationship between corruption and economic development
is hard to come by. A notable exception is the study
by Mauro (1995). Using Business International
Corporation’s indexes on corruption, red tape, and
efficiency of the judicial system over the period
1980–83 (now incorporated into the Economist Intelligence Unit), Mauro was able to estimate the direct
effects of corruption on economic development. He
found that corruption lowers investment, even controlling for other determinants of investment and endogeneity effects. The magnitude of the effect is quite
significant. Mauro found that a one standard deviation
improvement in the corruption index is associated
with an increase in investment of 2.9 percent of GDP.
This means, for example, that “if Bangladesh were to
improve the integrity and efficiency of its bureaucracy to the level of that of Uruguay, its investment rate
would rise by almost five percentage points, and its
yearly GDP growth rate would rise by over half a
percentage point” (Mauro, 1995, p. 705).
30
On the theoretical side, the literature has proceeded along two lines. One, following Becker and Stigler
(1974), used a principal–agent approach. In particular,
it focused on the incentives that the central government (the principal) can give a government official
(the agent) to make him behave honestly. Another
strand, following Shleifer and Vishny (1993), took
the corrupt behavior of government officials as a given and analyzed the consequences that their behavior
has on resource allocation. In this approach, corrupted officials are modeled as monopolistic suppliers of
a government good (such as a passport, an import license, the right to use a road, etc.) that is supposed to
be supplied at a prespecified price. The corrupt official overcharges the government good to maximize
his total revenues.
More recently, Acemoglu and Verdier (2000) took
a broader approach. They considered a static economy in which producers can choose to pay a cost in order to produce with a clean technology (otherwise,
their production process pollutes the environment).
The government wants to tax polluters and subsidize
clean producers in order to reduce the associated negative externality. However, it must rely on officials to
1Q/2008, Economic Perspectives
inspect the producers and determine their pollution
status. The officials are assumed to be corrupt: Through
bribes they are able to grab an exogenous share of the
surplus, which is assumed to be equal to the sum of
the tax and the subsidy that the official can potentially
charge. As a consequence, the government faces an
important trade-off between taxation and corruption:
It wants to tax polluters, but in order to detect them it
must rely on corrupted officials that consume resources.
In this environment, Acemoglu and Verdier (2000) characterize the optimal amount of taxation/corruption.
While Acemoglu and Verdier (2000) were able to
analyze the optimal taxation/corruption policy of the
government, in order to do so they had to simplify the
interaction between the government officials and the
producers to a reduced form. My contribution to this
literature is to spell out that interaction in an explicit
game and analyze its implications in detail. Since
Djankov et al. (2002) report that there are large differences across countries in the regulation of entry and
that this type of regulation is associated with sharply
higher levels of corruption, I formulate the corruption
game in the context of entry decisions to an industry.3
The corruption game
The corruption game is as follows. Consider the
case of a product line that is supplied by a single producer—the incumbent. The value of supplying the
product line is given by V. In addition, there is a potential producer that has just created a new product
generation—the innovator. If the innovator is allowed
to supply the new product, the incumbent will be driven
out of the market. As a consequence, the innovator
would obtain the value V and the incumbent would
lose it. Entry is regulated: The innovator must receive
permission from the government to enter business. The
reason for the regulation is that the innovator may produce with a technology that pollutes the environment.
The government is willing to grant the entry permit to
the innovator only if the new production technology
is clean. However, the government must send a government official to determine whether the new technology pollutes or not. Once the government official
inspects the new technology, its pollution status becomes fully known to him. After the official learns
the pollution status of the new technology, he must
report it to the central government. If the official reports that the new technology pollutes the environment, the innovator is precluded from producing but
faces no additional penalties.4
The government official is corrupt. He has the
ability of misrepresenting to the government the true
pollution status of the new technology. This allows
Federal Reserve Bank of Chicago
him to try to extract a bribe, either from the incumbent or the innovator, in determining which report to
make to the central government. For simplicity, I assume that the pollution status of the new technology
is fully known to both the innovator and the incumbent. This means that once the government official
inspects the new technology, its pollution status becomes common knowledge to the three parties—the
incumbent, the innovator, and the official.
The government never observes the actual bribe
payment received by the official. However, once the
official makes his report, with probability ϕ the government independently learns about the true pollution
status of the new technology. If the official is found
to have granted an entry permit to a polluter, there are
penalties involved. In particular, the official is fined
pV, while the innovator is fined mV. If the official is
found to have rejected an entry permit to a clean innovator, there are also penalties involved: The official
is fined pV, and the incumbent is fined mV.
The official is assumed to be able to make takeit-or-leave-it offers. The key decision for the official
is whether to request a bribe and from whom. In what
follows, we will see that the best strategy for the official is to turn to the producer with the largest joint
surplus and make him a take-it-or-leave-it offer. However, for the producer with the largest joint surplus
to accept this bribe proposal, it must be credible that
the producer with the second largest joint surplus
would be willing to accept a bribe proposal if offered
one. This will require the second largest joint surplus
to be positive.
In principle, two possible scenarios must be considered: the scenario in which the innovator does not
pollute the environment and the scenario in which the
innovator does pollute the environment. However, the
two scenarios are completely symmetrical. In each
scenario there is a “legal” producer and an “illegal”
producer. In the case that the innovator pollutes, the
legal producer is the incumbent; in the case that the
innovator does not pollute, the legal producer is the
innovator. Moreover, the payoffs to each player in the
corruption game only depend on whether the bribes
are being extracted from the legal producer or the illegal producer (that is, the payoffs are independent of
the actual identity of the producers). Given this symmetry, in what follows I consider a single corruption
game that differentiates producers only according to
their legal status, with the understanding that the
identities of the legal and illegal producers are determined by the actual scenario taking place.
Figure 2 describes the sequence of moves for the
corruption game. In the first stage, the government
31
will reject any bribe request, since he
knows that the government official will
subsequently take the legal course of
action.
Observe that the payoff to the illegal
producer of reaching a deal with the government official is:
figure 2
Sequence of moves for corruption game
O
No bribes sought
Turn to I
O
Make offer to I
I rejects
Turn to L
O
Make offer to L
I
I accepts
L accepts
L rejects
O
O
Make offer to I
Make offer to L
I
L
L rejects
PI = V − BI − ϕ [V + mV],
L
L accepts
O allocates permit properly
I rejects
I accepts
O allocates permit properly
Notes: I refers to illegal producer; L to legal producer; and O to government
official. See the text for further details.
official must decide between three alternatives: 1) not
to seek bribes, 2) to initially seek a bribe from the illegal producer I, and 3) to initially seek a bribe from
the legal producer L. In the case that the official seeks
a bribe, he must decide how much to demand from
the producer he initially turns to (the continuum of
values for the bribe are represented as the base of the
triangles in figure 2). If the bribe request is accepted,
the game ends. Otherwise, the official turns to the
second producer and decides how much to demand
from him. The game ends after this point. If this bribe
request is rejected, the official assigns the production
permit to the legal producer, since he has nothing to
gain otherwise. In what follows, I analyze the way
that the corruption game is played.
First, observe that the government official always
has a larger joint surplus to share with the legal producer than with the illegal producer. The reason is that
the value of being the product leader is the same for
both types of producers, but there are penalties involved
if a deal with the illegal producer is subsequently detected. In addition, the value of not being the product
leader is the same for both types of producers (in particular, it is equal to zero). This means that the government official will always want to extract bribes from
the legal producer. However, for the legal producer to
be willing to pay such a bribe, it should be credible
that the government official would want to reach a
deal with the illegal producer in a second round of
negotiation. If this is not the case, the legal producer
32
where BI are the bribes paid. This payoff
is equal to the value of being the product
leader net of the bribe payment minus the
losses if the deal is detected, an event that
happens with probability ϕ. Since the
payoff to the illegal producer of rejecting
the bribe is zero, the largest bribe that the
government official would be able to extract from the illegal producer in a take-itor-leave-it offer is given by:5
1)
this case is
BI = (1 − ϕ)V − ϕmV.
The payoff to the government official in
PO = BI − ϕpV = (1− ϕ)V − ϕmV − ϕpV.
That is, it is the maximum bribe that the government
official could extract from the illegal producer minus
the penalty pV times the probability ϕ of being caught
by the central government.
The condition that this payoff PO is positive reduces to
2)
1− ϕ
> m + p.
ϕ
If this condition is not satisfied, it would be in the
best interest of the government official not to seek a
bribe from the illegal producer. Hence, the legal producer would reject any take-it-or-leave-it offer made
by the official and the legal course of action would
be taken. If the condition in equation 2 is satisfied,
the government official would be able to extract bribes
from the legal producer, since it becomes fully credible that he would subsequently want to reach a deal
with the illegal producer.
Observe that the payoff to the legal producer of
reaching a deal with the government official is:
PL = V − BL.
1Q/2008, Economic Perspectives
That is, it is equal to the value of being the product
leader net of the bribes paid. This payoff is nonrandom because if the central government independently
learned about the pollution status of the innovator, it
would conclude that the entry permit was correctly
allocated (recall that the central government can never
prove that a bribe payment took place). Also, the payoff to the legal producer of rejecting a bribe offer
from the government official is ϕV, since with probability ϕ the illegal action will be detected by the central government and the legal producer will become
the product leader. Hence, the largest bribe that the
government official will be able to extract from the
legal producer is: 6
3) BL = (1 − ϕ)V.
To summarize, the equilibrium of the corruption
game is as follows. If the condition in equation 2 is
violated, no bribes are paid. If the condition in equation 2 is satisfied, the government official extracts
from the legal producer the bribes given by equation 3.
In both cases, the official takes the legal course of
action. Figure 3 provides an illustration of the equilibrium outcome.
Innovation decisions
In this section, I describe in detail the industry in
which the incumbent and innovator of the previous
section operate. The purpose is to determine how
corruption affects the industry’s innovation rate.
The industry produces a product that comes in
many possible qualities. At each point in time, there
is a frontier version that dominates all previous ones.
A single producer has the patent to this version. He
drives all other producers out of the market and enjoys a profit flow equal to Π. However, he loses his
leading position whenever an innovator enters business with a quality improvement. In this case, the
incumbent is driven out of the market, and the innovator becomes the new industry leader, which provides him the profit flow Π.
Product innovations take place at an endogenously determined rate η. At every point in time there
are a large number of potential producers (innovators)
that invest in research and development (R&D) in order to create a new product generation. They all face
a same cost function r(η), which describes the costs
of generating an arrival rate equal to η.7 If an innovator succeeds in creating the new product generation,
he can apply for an entry permit. If the entry permit is
awarded, the innovator becomes the new industry
leader. However, entry is regulated as in the previous
Federal Reserve Bank of Chicago
figure 3
Equilibrium outcomes in parameter space
Bribes from innovator
No bribes
(1 − ϕ )
p+m
ϕ
Note: See the text for further details.
section. In particular, a government official is sent to
inspect the pollution status of the new technology. As
a result, the official, the incumbent producer, and the
innovator end up playing the corruption game described before. The probability that an entry application is inspected by a government official is equal to
γ, while the probability that an innovation pollutes is
equal to ξ.
The optimization problem of an innovator is then
the following:
4) max {η [ξNP + (1 − ξ) NC ] − r (η)},
where NP is the value of being an innovator that pollutes and NC is the value of being an innovator that
produces with a clean technology. That is, the innovator chooses the arrival rate η to maximize the expected value net of R&D costs. The optimal innovation
rate η is characterized by the following condition:
5) r′ (η) = ξNP + (1 − ξ) NC .
That is, the innovator equates marginal revenue to
marginal cost. In what follows, I sketch the main
properties of the optimal R&D investment decisions
both from an individual point of view and at the
industry level. The appendix provides a more
detailed analysis.
To start with, observe that the marginal cost
function r′ is strictly increasing. Thus, given fixed
values for NP and NC , there is a unique value of η that
satisfies equation 5. While an individual innovator
takes the values of NP and NC as given (since he is
competitive), these values actually depend on the industry-wide innovation rate η*. Moreover, they are
strictly decreasing in the industry-wide innovation
rate η*. The reason is that given all other parameter
values, an increase in η* decreases the expected
length of time over which a producer can retain the
leadership of a product line (that is, it increases the
33
rate at which future innovators will drive him out of
the market). Thus, the expected value
N = ξN P + (1 − ξ ) N C
in the right-hand side of equation 5 is strictly decreasing in η*. At equilibrium, the industry-wide innovation rate η* that innovators take as given (and that determines the expected value N ) must be identical to
the one they choose from their individual perspective.
That is, at equilibrium we must have that the innovation rate satisfies:
6) r ′ ( η* ) = N ( η* ) .
Since the left-hand side of equation 6 is strictly increasing in η* and the right-hand side of equation 6 is
strictly decreasing in η*, there is a unique value of η*
that satisfies this equation. That is, there is a unique industry equilibrium. Figure 4 illustrates this equilibrium.
We are interested in how the equilibrium innovation rate η* is affected by changes in different parameter values. While the appendix provides a formal
analysis, the results are quite intuitive. We saw in the
previous section that the penalties to the government
official and illegal producer (p and m, respectively)
affect whether bribes are paid or not but do not affect
the magnitude of the bribes. In particular, if the condition in equation 2 is satisfied, bribes are paid. However, p and m do not enter equation 3, which describes
the equilibrium bribes BL that the government officials
are able to extract from the legal producers. This means
that as long as p + m > (1 − ϕ)/ϕ, the expected value
N is independent of those penalties; but as soon as p + m
becomes equal to (1 − ϕ)/ϕ, the expected value N
plummets because now producers become subject to
bribes. Further, decreases in p + m have no additional
effects in N. The implications for the equilibrium
innovation rate are shown in figure 5. The curve N1
describes the expected value of innovating in the case
in which there are no bribes (that is, when p + m >
(1− ϕ)/ϕ), while the curve N 2 describes the expected
value of innovation when producers pay bribes (that
is, when p + m < (1 − ϕ)/ϕ). Since N 2 is lower than
N1 for every value of η, it follows that the equilibrium innovation rate with bribes η*2 must be lower than
the equilibrium innovation rate when there are no
bribes η1* . This leads to my main result: The effects
of penalties to corruption on equilibrium innovation
rates are highly nonlinear. In particular, small changes
in penalties p + m around the critical value (1 − ϕ)/ϕ
can lead to large changes in innovation rates, while
changes in penalties far from that critical value have
no effects. The discontinuous dependence of the equilibrium innovation rate η* on the total penalties p + m
is depicted in figure 6.
The effects on the equilibrium innovation rate of
changes in the probability of detecting corruption ϕ
and in the fraction of entry applications that get inspected γ are more complex, since they not only determine
whether bribes are paid, but also affect the position of
figure 5
*
*
Innovation rates with ( η2 ) and without ( η1 ) bribes
figure 4
Industry equilibrium
r'
r'
N
N
η*
innovation rate
Note: See the text for further details.
34
1
N2
η*
η*
2
1
innovation rate
Note: See the text for further details.
1Q/2008, Economic Perspectives
Sources:
figure 6
Innovation rate (η*) vs. corruption penalties (p + m)
innovation rate
0.985
0.980
0.975
0.970
0.965
0.960
0.955
0.950
0.945
0.940
0.00E+0
0.1
0.2
0.3
0.4
0.5 0.6
0.7
0.8
0.9
1.0
1.1
1.2 1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
corruption penalties
the curves N1 and N 2 in figure 5. A numerical analysis
of these effects is provided in the appendix.
Conclusion
I have illustrated how the rate of product innovation can be affected by changes in parameter values determining the amount of corruption in an industry. An
interesting result of the analysis is that, under certain
parameter ranges, small increases in the penalties to
corruption or the effectiveness of detection can result
in large increases in the amount of product innovation.
While I have not explicitly analyzed the effects
of innovation on economic development, it is safe to
speculate what those effects would be. To be specific,
consider Grossman and Helpman’s (1991) endogenous growth model. In that model, there is a continuum of product lines, each characterized by quality
ladders of fixed increments. In each product line,
Federal Reserve Bank of Chicago
there is always a leader producer that supplies the
frontier quality and drives all previous producers out
of the market. However, the arrival rate of innovators
is optimally determined in an R&D sector. Successful
innovators drive the incumbent leaders out of the market
and become the new product leaders. Thus, each product line has a similar structure as the industry considered in this article. Introducing a corruption game in
each product line would thus deliver similar results.
Since in Grossman and Helpman (1991) the growth
rate of the economy is determined by the endogenous
innovation rate, the effects of corruption found here
would translate into growth effects. In particular,
small increases in the penalties to corruption or the
effectiveness of detection can lead to jumps in the
growth rate of the economy. Thus, corruption has the
potential of grouping countries into two distinct development groups: fast- and slow-growing countries.
35
notes
While I do not explicitly analyze the links between corruption, innovation, and economic growth, I sketch them in some detail in the
conclusion.
1
The Penn World Table—maintained by the Center for International
Comparisons at the University of Pennsylvania—provides purchasing power parity and national income accounts converted to international prices for 188 countries for some or all of the years 1950–2004.
For further details, please see http://pwt.econ.upenn.edu/. Transparency
International is a global organization promoting anticorruption
policies. Its Corruption Perception Index ranks countries by the
perceived levels of corruption (frequency and/or size of bribes) in
the public and political sectors, as determined by expert assessment
and business opinion surveys. The Corruption Perception Index
can be downloaded from www.transparency.org.
2
For example, Djankov et al. (2002) report that to meet government
requirements for starting a business in 1999, an entrepreneur in
Italy needed to follow 16 different procedures, pay US$3,946 in
3
fees, and wait at least 62 business days to acquire the necessary
permits. In contrast, an entrepreneur in Canada only needed to follow two procedures, pay US$280, and wait for two days. An extended account of how entry regulation leads to corruption and
bureaucratic delays is provided by De Soto (1989). However, he
focuses on the Peruvian economy.
Introducing a fine to polluters would significantly complicate the
analysis of the corruption game without additional insights.
4
This bribe request makes the illegal producer indifferent between
accepting and rejecting it.
5
This bribe request makes the legal producer indifferent between
accepting and rejecting it.
6
This cost function is assumed to be increasing, differentiable, and
strictly convex. Moreover, r′ (0) = 0 and r′ (∞) = ∞.
7
APPENDIX: RESEARCH AND DEVELOPMENT DECISIONS AND INDUSTRY EQUILIBRIUM
Given the solution to the corruption game characterized
in the main text, we can proceed to write expressions for
NP and NC. The expected value of an innovator that does
not pollute NC is given by:
(1 + ϕ )
V if p + m >
ϕ
NC =
.
(1 − γ )V + γϕV , otherwise
Observe that when p + m >
(1 + ϕ ) ,
there are no
ϕ
bribes paid in the corruption game. Hence, the clean
innovator obtains the value V of becoming a leader
(1 + ϕ ) , bribes are paid
with certainty. When p + m <
ϕ
whenever the innovator gets inspected. As a consequence,
the innovator gets the full value V only if he is not inspected, an event that happens with probability (1 − γ).
With probability γ, the (clean) innovator is inspected and
obtains a value (net of bribes) of ϕV.
The expected value of an innovator that pollutes NP
is given by:
NP = (1 − γ) V.
The innovator that pollutes obtains the full value of becoming the leader V only if he is not inspected, which
happens with probability (1 − γ). With probability γ, the
innovator that pollutes is inspected and is precluded
from producing (recall that for every parameter specification the government official always takes the legal
course of action).
36
The value of being the industry leader V is given as
follows:
(1 − ϕ ) < p + m
∏ −ηV + ηξγV if
iV =
ϕ
,
∏ −ηV + ηξγϕV, otherwise
where i is the instantaneous interest rate. The flow value
of being the leader iV is given by Π, but with arrival rate
η, a new innovator enters the market, in which case the
profit flow Π is permanently lost. However, there are
(1− ϕ ) < p + m, the loss is
exceptions to this loss. When
ϕ
avoided when the new arrival pollutes and is inspected by
a government official, an event that happens with probability ξγ (in this case there are no bribes imposed and the
(1 − ϕ ) ,
entry permit is rejected). Also, when p + m <
ϕ
the loss is partly avoided when the new arrival pollutes
and is inspected by a government official (again, an
event that happens with probability ξγ). However, in this
case, the leader is only able to retain a fraction ϕ of the
value of being the leader V.
We are now ready to write the expected value of
creating a new product generation in equation 4 (p. 33):
N = ξN P + (1 − ξ ) N C .
This expected value depends on parameter values, since
the outcome of the corruption game varies depending on
them. As a consequence, I will index the expected value
Nj according to the parameter region j.
1Q/2008, Economic Perspectives
Parameter region 1( j = 1) :
A1)
ϕ
N1 ( η) = {ξ (1 − γ ) + (1 − ξ )}
Parameter region 2 ( j = 2 )
r ′ ( η*j ) = N j ( η*j ) ,
(1 − ϕ ) < p + m,
and that these arrival rates are ordered across parameter
regions as follows:
∏
.
i + η − ηξγ
A4) η*2 < η1*.
(1 − ϕ ) ,
: p+m<
ϕ
{
}
A2) N 2 ( η) = ξ (1 − γ ) + (1 − ξ ) (1 − γ ) + γϕ
∏
.
i + η − ηξγϕ
Observe that, in each parameter region j, the expected value Nj ( η) depends on the industry’s arrival
rate η, which is an endogenous variable of the model. In
particular, the expected values Nj ( η) depend negatively
on η. Also, it is straightforward to verify that for every
possible value of the arrival rate η, that
A3) N 2 ( η) < N1 ( η).
Observe that, since r is a convex function, r′ is increasing in η. This, together with the previously mentioned properties for the expected values N j ( η) , allows
us to establish that in each parameter region j there is a
unique equilibrium arrival rate η*j satisfying that
Sources:
As mentioned in the main text, this inequality leads
to the main result of the article. Fixing all other parameter
values, lower penalties on corruption p + m lead to lower
rates of innovation. However, the relation is highly nonlinear. Reductions in p + m have no effects on rates of
innovation as long as they leave the model within the
same parameter region. But once the edge of a parameter
region is approached, small reductions in p + m have
large effects as the equilibrium innovation rate η* jumps
from one region to the next.
The effects of the probability of detection ϕ and the
fraction of entry applications that get inspected γ are
more complex because they affect not only the length
of the parameter regions but also the position of the
expected values N1 and N 2 in figure 5 (p. 34). To
ease the presentation of these effects, in what follows
I complement the analysis with a numerical example.
It is important to point out that the example has no
empirical content, since parameter values are not
chosen to reproduce observations; it serves illustration
figure A1
Innovation rate (η ) vs. probability of detecting corruption (ϕ)
*
innovation rate
0.99
0.98
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
0.97
0.96
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
0.94
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
0.93 F
F
F
F
F
F
F
F
F
F
F
F
0.95
0.92
0.91
0.90
.00E+0 5.00E-2 .10 .15
.20
.25
.30
.35
.40
.45
.50
.55
.60
.65
.70
.75
.80
.85
.90 .95 1.0
probability of detecting corruption
Federal Reserve Bank of Chicago
37
Sources:
figure A2
Innovation rate (η*) vs. probability of inspection (γ)
innovation rate
1.00
0.80
0.60
0.40
0.20
0.00
.00E+0 5.00E-2 .10
.15
.20
.25
.30
.35
.40
.45
.50
.55
.60
.65
.70
.75 .80
.85
.90
.95 1.00
probability of inspection
purposes only. The example considered has the following parameter values: ξ = 0.5, γ = 0.1, ϕ = 0.5, p = 0.8,
m = 0, i = 0.04, Π = 1 (this is just a normalization), and
1
r ( η) = η2 .
2
Fixing all other parameters at their benchmark values,
figure A1 shows how the equilibrium innovation rate
depends on the probability of detecting corruption ϕ.
The figure shows that a higher detection probability ϕ
(weakly) increases the innovation rate of the industry.
However, the dependence is discontinuous, and once the
arrival rate jumps, it is unresponsive to further increases
in ϕ. These properties are general. We see from equations A1 and A2 that N j ( η) increases with ϕ when j = 2
but is independent of ϕ when j = 1. Moreover, an increase in ϕ can bring the economy from parameter region j = 2 to j = 1, entailing a jump in the arrival rate
from η*2 to η1* at the critical value for ϕ at which
38
p+m=
(1 − ϕ ) .
ϕ
Figure A2 shows how the equilibrium innovation
rate depends on the probability of inspection γ. The figure shows that a higher probability of inspection γ decreases the innovation rate of the industry in a continuous way. This is a general result. We see from equations
A1 and A2 that N j ( η) decreases with γ in each case j = 1, 2.
Since the functions depicted in figure 5 (p. 34) shift
down as γ increases, the intersections with r′(η) take
place at lower values of η*j , for each j = 1, 2. However,
changes in γ have no effect on the parameter region that
the economy lies on. Thus, while the innovation rate
decreases with γ, there are no points of discontinuity.
1Q/2008, Economic Perspectives
REFERENCES
Acemoglu, D., and T. Verdier, 2000, “The choice
between market failures and corruption,” American
Economic Review, Vol. 90, No. 1, March, pp. 194–211.
Grossman, G., and E. Helpman, 1991, “Quality
ladders in the theory of economic growth,” Review of
Economic Studies, Vol. 58, No. 1, January, pp. 43–61.
Becker, G., and G. Stigler, 1974, “Law enforcement,
malfeasance, and the compensation of enforcers,”
Journal of Legal Studies, Vol. 3, No. 1, January,
pp. 1–18.
Mauro, P., 1995, “Corruption and growth,” Quarterly
Journal of Economics, Vol. 110, No. 3, August,
pp. 681–712.
De Soto, H., 1989, The Other Path: The Invisible
Revolution in the Third World, New York: Harper
and Row.
Shleifer, A., and R. Vishny, 1993, “Corruption,”
Quarterly Journal of Economics, Vol. 108, No. 3,
August, pp. 599–617.
Djankov, S., R. La Porta, F. Lopez-de-Silanes, and
A. Shleifer, 2002, “The regulation of entry,” Quarterly
Journal of Economics, Vol. 117, No. 1, February,
pp. 1–37.
Federal Reserve Bank of Chicago
39
@FedFRASER