University of Richmond
UR Scholarship Repository
Political Science Faculty Publications Political Science
Spring 2013
University of Richmond, md[email protected]du
Follow this and additional works at: h8p://scholarship.richmond.edu/polisci-faculty-publications
Part of the Business Law, Public Responsibility, and Ethics Commons, Economic Policy
Commons, International Business Commons, and the Models and Methods Commons
7is Article is brought to you for free and open access by the Political Science at UR Scholarship Repository. It has been accepted for inclusion in
Political Science Faculty Publications by an authorized administrator of UR Scholarship Repository. For more information, please contact
scholarshiprepository@richmond.edu.
Recommended Citation
Da8a, Monti Narayan, and Kevin Bales. "Slavery Is Bad for Business: Analyzing the Impact of Slavery on National Economies." e
Brown Journal of World Aairs 19, no. 2 (Spring/Summer 2013): 205-23.
205
S/S
2013
• ,
205
Copyright © 2013 by the Brown Journal of World Aairs
Slavery is Bad for Business:
Analyzing the Impact of Slavery on
National Economies
Assistant Professor
M N D
University of Richmond
P and modern-day slavery is reach-
ing a tipping point—it is coming to be understood as a global problem with
economic and policy implications far beyond simple reports of cross-border
human tracking. A decade ago most educated citizens considered slavery a
phenomenon of the past, relegated to history textbooks. Today a strong narra-
tive has reached global proportions: activists, epistemic communities, NGOs,
IGOs, and governments are acknowledging the scope and extent of slavery in
the twenty-rst century. One need only point to Nicholas Kristof and Sheryl
Wu Dunn’s bestseller Half the Sky, President Barack Obama’s 2012 speech at the
Clinton Global Intiative, or the awareness that celebrities such as Mira Sorvino
and Jada Pinkett Smith are raising about the subject to illustrate how far the
antislavery movement has progressed.
1
Yet despite such civic mindedness, surprisingly little data and empirically
driven research exist on slavery today. Although some headway has been made
in estimating its prevalence—most notably in the form of contributions by
Kevin Bales and by the International Labor Organization (ILO)—apart from a
Professor
K B
University of Hull
M N D is assistant professor of political science at the University of Richmond. He is
the author of Anti-Americanism and the Rise of World Opinion: Consequences for the U.S. National Interest
(Cambridge University Press, forthcoming). He is developing several projects on human tracking and
modern-day slavery with Kevin Bales of Free the Slaves and Helen Sworne and Siobhan Miles of Chab Dai.
K B is professor of contemporary slavery at the Wilberforce Institute for the Study of Slavery and
Emancipation at the University of Hull. He is the author of Disposable People: New Slavery in the Global
Economy (University of California Press, 1999) and a number of other books on contemporary slavery.
He is currently completing a book on the relationship between slavery and environmental destruction
and global warming and a project on global forced marriage with Jody Sarich.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
206
207
as jus cogens, a norm from which no derogation is ever permitted. As an illegal
economic activity, it is restricted, at least nominally, in all nation-states. On a
moral level, no rational politician would condone slavery today. Indeed, nor-
matively, the twenty-rst century is an age in which it is socially fashionable to
wear T-shirts or display bumper stickers that promote the antislavery movement.
On the other hand, contemporary slavery, like its historical antecedent,
yields certain economic benets, albeit for a very small portion of the global
economy. e ILO estimates that about $44 billion USD per year is made in
slavery-derived prots. Although this is a drop in the bucket of global economic
output, valued at $74 trillion, it suggests that, at least for illicit markets, there
are incentives to perpetute slavery.
7
e prots of slavery ow primarily to
slaveholders, who benet from global markets and global prices. For example,
a criminal who uses slaves to produce a cotton crop will sell his cotton at the
same price as his neighbor who does not exploit slaves. ey both receive the
market price, one which is ultimately set in the global commodity exchanges.
While both farmers receive the same price for their cotton, the slave-using farmer
will have a higher prot margin due to his much lower labor cost, amounting
to bare subsistence for his workers. With a guaranteed market price, he has no
incentive to sell at a lower rate. As with many other criminal enterprises such
as drug tracking, the estimated prot margins found in contemporary slavery
are sometimes astoundingly high.
8
Prot margins vary in contemporary slavery. At one end of the spectrum
is the commercial sexual exploitation of enslaved women. Such an illegal—and
extremely brutal—working-class brothel business has been found to generate
prot margins in the order of 850 percent in the booming economy of ailand,
a locality with high levels of demand.
As Table 1 shows, the monthly prots from slavery in a working-class
brothel in a small town in western ailand are about $72,000. Annually, this
yields a net prot of $864,000. Brothels serving richer clients or sex tourists
might be expected to fare better. e fact that there are dozens of brothels in
this small town alone suggests that enslavement for sexual exploitation is a
multimillion-dollar business in the region.
At the other end of the spectrum are traditional forms of enslavement that
tend to be part of a country’s local economy. Brick kilns that rely on slave labor
in rural Pakistan, for example, generate around 25 percent prot, compared to
10–15 percent for those kilns that pay their workers. At the same time, hereditary
forms of collateral debt bondage with similar prot margins, practiced in agricul-
tural slavery in northern India, remain extensive.
9
Adding to the protability of
rough estimate of how many slaves exist in the world today (21 to 27 million),
scholars and policy makers know little about the risk factors—let alone the busi-
ness impact—that contemporary slavery has on the global community.
2
Indeed,
most extant research, athough useful, is qualitative, not allowing for statistical
models.
3
To what extent is slavery empirically bad for business? For whom is
the business of slavery protable, and for whom is it economically burdensome?
is article, using a novel dataset, demonstrates that slavery is empirically
bad for business. Building upon the work of Robert Smith, our analysis exam-
ines the relationship between the prevalence of slavery in a country (in terms of
the proportion of the population enslaved) and several economic measures (the
United Nations Human Development Index,
growth domestic product in terms of purchas-
ing power parity, access to nancial services,
and the Gini coecient).
4
In each instance,
controlling for alternative explanations, greater
levels of slavery are associated with a decline in economic growth and human
development. e ndings imply that beyond the morality of the issue, slavery
is objectively harmful for total economic output and social development. is
article begins with a discussion of how slavery is protable for slaveholders and
then proceeds with a discussion of how it undermines social and economic
output at the macro level.
conteMPorary slavery: ProfitaBle for slaveholders
A signicant amount of historical and economic research has examined the
slavery of the past and clearly demonstrates its protability for slaveholders.
5
As a legal enterprise, slavery in the past generated ancillary economic activity in
the form of insurance, transport, advertising, loans, credit, mortgages, and so
on. Governments could also tax proceeds from slave-trading and slave use and
regulate the activities and procedures of slavery. Although a moral evil, histori-
cally slavery was good for business.
6
Contemporary slavery, primarily due to its universal illegality, is a far dif-
ferent creature. e fundamental nature of slave work has not changed dramati-
cally; it is still dirty, dangerous, and demeaning. Most slaves work in agriculture,
mining or other such extractive industries, assembly, food processing, manu-
facturing, or what might be called “personal (forced) services” such as domestic
service or commercial sexual exploitation. It is a type of work that tends to be
hidden because it is illegal in all countries and condemned in international law
Slavery is objectively harmful
for total economic output
and social development.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
206
207
as jus cogens, a norm from which no derogation is ever permitted. As an illegal
economic activity, it is restricted, at least nominally, in all nation-states. On a
moral level, no rational politician would condone slavery today. Indeed, nor-
matively, the twenty-rst century is an age in which it is socially fashionable to
wear T-shirts or display bumper stickers that promote the antislavery movement.
On the other hand, contemporary slavery, like its historical antecedent,
yields certain economic benets, albeit for a very small portion of the global
economy. e ILO estimates that about $44 billion USD per year is made in
slavery-derived prots. Although this is a drop in the bucket of global economic
output, valued at $74 trillion, it suggests that, at least for illicit markets, there
are incentives to perpetute slavery.
7
e prots of slavery ow primarily to
slaveholders, who benet from global markets and global prices. For example,
a criminal who uses slaves to produce a cotton crop will sell his cotton at the
same price as his neighbor who does not exploit slaves. ey both receive the
market price, one which is ultimately set in the global commodity exchanges.
While both farmers receive the same price for their cotton, the slave-using farmer
will have a higher prot margin due to his much lower labor cost, amounting
to bare subsistence for his workers. With a guaranteed market price, he has no
incentive to sell at a lower rate. As with many other criminal enterprises such
as drug tracking, the estimated prot margins found in contemporary slavery
are sometimes astoundingly high.
8
Prot margins vary in contemporary slavery. At one end of the spectrum
is the commercial sexual exploitation of enslaved women. Such an illegal—and
extremely brutal—working-class brothel business has been found to generate
prot margins in the order of 850 percent in the booming economy of ailand,
a locality with high levels of demand.
As Table 1 shows, the monthly prots from slavery in a working-class
brothel in a small town in western ailand are about $72,000. Annually, this
yields a net prot of $864,000. Brothels serving richer clients or sex tourists
might be expected to fare better. e fact that there are dozens of brothels in
this small town alone suggests that enslavement for sexual exploitation is a
multimillion-dollar business in the region.
At the other end of the spectrum are traditional forms of enslavement that
tend to be part of a country’s local economy. Brick kilns that rely on slave labor
in rural Pakistan, for example, generate around 25 percent prot, compared to
10–15 percent for those kilns that pay their workers. At the same time, hereditary
forms of collateral debt bondage with similar prot margins, practiced in agricul-
tural slavery in northern India, remain extensive.
9
Adding to the protability of
rough estimate of how many slaves exist in the world today (21 to 27 million),
scholars and policy makers know little about the risk factors—let alone the busi-
ness impact—that contemporary slavery has on the global community.
2
Indeed,
most extant research, athough useful, is qualitative, not allowing for statistical
models.
3
To what extent is slavery empirically bad for business? For whom is
the business of slavery protable, and for whom is it economically burdensome?
is article, using a novel dataset, demonstrates that slavery is empirically
bad for business. Building upon the work of Robert Smith, our analysis exam-
ines the relationship between the prevalence of slavery in a country (in terms of
the proportion of the population enslaved) and several economic measures (the
United Nations Human Development Index,
growth domestic product in terms of purchas-
ing power parity, access to nancial services,
and the Gini coecient).
4
In each instance,
controlling for alternative explanations, greater
levels of slavery are associated with a decline in economic growth and human
development. e ndings imply that beyond the morality of the issue, slavery
is objectively harmful for total economic output and social development. is
article begins with a discussion of how slavery is protable for slaveholders and
then proceeds with a discussion of how it undermines social and economic
output at the macro level.
conteMPorary slavery: ProfitaBle for slaveholders
A signicant amount of historical and economic research has examined the
slavery of the past and clearly demonstrates its protability for slaveholders.
5
As a legal enterprise, slavery in the past generated ancillary economic activity in
the form of insurance, transport, advertising, loans, credit, mortgages, and so
on. Governments could also tax proceeds from slave-trading and slave use and
regulate the activities and procedures of slavery. Although a moral evil, histori-
cally slavery was good for business.
6
Contemporary slavery, primarily due to its universal illegality, is a far dif-
ferent creature. e fundamental nature of slave work has not changed dramati-
cally; it is still dirty, dangerous, and demeaning. Most slaves work in agriculture,
mining or other such extractive industries, assembly, food processing, manu-
facturing, or what might be called “personal (forced) services” such as domestic
service or commercial sexual exploitation. It is a type of work that tends to be
hidden because it is illegal in all countries and condemned in international law
Slavery is objectively harmful
for total economic output
and social development.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
208
209
of enslaved people since replacement can be more cost eective than care. e
cost of slaves is now so low that they are often seen as disposable inputs into
criminal enterprises rather than as capital investments.
tainted Goods
Although slavery might benet slaveholders, it negatively aects everyone else—
ranging from slaves to the global consumer—in dierent ways. e negative
impact for global consumers is not economic but moral and political. Slavery does
not reduce or increase prices but does involve consumers in illegal and immoral
activities. Many countries including the United States prohibit the importation
of slave-made goods, and case law going back to the eighteenth century allows
the conscation of goods that circumstantially may have been involved in the
slave trade. Adding to this is the repugnance expressed by most consumers at the
idea of buying goods made by slaves, especially child slaves—and that repugnance
is hard to avoid. Slaves are used in the production of many basic commodities
that ow from the poor Global South to the rich North, as well as of some
goods, such as tomatoes and oranges, produced and consumed in the North.
12
Originating from numerous countries, documented cases of slavery exist in the
production of textiles, gemstones, apparel, and agricultural products.
13
Coee
is sometimes grown using slave labor, and some sugar is harvested by slaves. In
the Congo, for example, armed gangs enslave local people and force them to dig
minerals such as coltan, cassiterite, wolframite, and tungsten. ese minerals
are then sold to exporters who send them to Europe and Asia where they are
used to make components needed for the production of cell phones, comput-
ers, and other electronics.
14
Slavery also creeps into nancial investments when
pension funds or mutual funds carry stock in companies that subcontract other
companies that use slave labor.
15
ese modern, scattered, small-scale, and criminal slave-based enterprises
contrast markedly with the relatively much larger economic organization of
historical slavery, in which, as evil as it was, the economic benets of the legal
slave trade were diused throughout local and regional economies. In the past,
it could be assumed that a sizable proportion of commodities such as sugar and
cotton was produced by slaves and that the prots from such enterprises would
be cycled back through local, regional, and national economies along with the
prots from other legitimate businesses.
16
Today, however, the situation is more complicated in that it is slaveholders
who primarily benet from contempary slavery. Consider the cocoa industry
contemporary slavery, regardless of the type of slavery, is the exceptionally low cost
of acquiring slaves.
10
For most of human history, slaves were expensive. In 1850,
for example, the average price of an “average” slave in the American South (a
“prime eld hand” meaning a young man capable of heavy agricultural work) was
between $1,000 and $1,200 in 1850 dollars, which equates to around $45,000
in 2013 dollars.
11
Today it could be argued that slaves are less expensive today
than at any time in the past. In a marketplace in the central region of the Ivory
Coast, one researcher easily purchased two 19-year-old agricultural workers for
about $30 each. In India, bonded laborers are able to report precise sums that
represent the acquisition cost required to enslave their family (often the debt
being taken some generations previously and passed down)—ranging from as
little as $10 to over $100. In Brazil desperate urban workers place themselves
in the hands of recruiters only to nd themselves held under violence far in the
forests and charged with a “debt” amounting to a few hundred dollars. Even in
ailand, where teenagers are sold into prostitution and a premium is charged
for their virginity and attractiveness, prices rarely go over $1000.
Table 1. Monthly income and expenditures (in baht) for the always prospering
brothel.U.S. gures adjusted for ination for 2012. Source: Kevin Bales, Disposable
People: New Slavery in the Global Economy (University of California Press, 1999).
Low initial acquisition costs means that slave owners neglect the welfare
Table 1. Monthly Income & Expenditure (in Baht) for the Always Prospering Brothel
Costs (per month)
Income (per month)
Rent
5,000
Commercial sex*
1,050,000
Utilities & bills
2,000
Rent paid by prostitutes
600,000
Food & drink
45,000
Sale of condoms
70,000
Pimp’s salary
7,000
Sale of drinks
672,000
Cashier
7,000
Virgin premium
50,000
Cook
5,000
“Interest” on debt-bond
15,000
Bribes
6,000
Payments to taxis etc.
12,000
Beer & whisky
168,000
TOTAL COSTS
257,000
TOTAL INCOME
2,457,000
(In US dollars)
($8,389)
MONTHLY PROFIT
2,200,000
(In U.S. dollars)
($71,813)
*Average 14 clients per day at 125 baht per client for 20 prostitutes for 30 days.
Source: Disposable People: New Slavery in the Global Economy, University of California
Press, 1999. U.S. figures adjusted for inflation for the year 2012.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
208
209
of enslaved people since replacement can be more cost eective than care. e
cost of slaves is now so low that they are often seen as disposable inputs into
criminal enterprises rather than as capital investments.
tainted Goods
Although slavery might benet slaveholders, it negatively aects everyone else—
ranging from slaves to the global consumer—in dierent ways. e negative
impact for global consumers is not economic but moral and political. Slavery does
not reduce or increase prices but does involve consumers in illegal and immoral
activities. Many countries including the United States prohibit the importation
of slave-made goods, and case law going back to the eighteenth century allows
the conscation of goods that circumstantially may have been involved in the
slave trade. Adding to this is the repugnance expressed by most consumers at the
idea of buying goods made by slaves, especially child slaves—and that repugnance
is hard to avoid. Slaves are used in the production of many basic commodities
that ow from the poor Global South to the rich North, as well as of some
goods, such as tomatoes and oranges, produced and consumed in the North.
12
Originating from numerous countries, documented cases of slavery exist in the
production of textiles, gemstones, apparel, and agricultural products.
13
Coee
is sometimes grown using slave labor, and some sugar is harvested by slaves. In
the Congo, for example, armed gangs enslave local people and force them to dig
minerals such as coltan, cassiterite, wolframite, and tungsten. ese minerals
are then sold to exporters who send them to Europe and Asia where they are
used to make components needed for the production of cell phones, comput-
ers, and other electronics.
14
Slavery also creeps into nancial investments when
pension funds or mutual funds carry stock in companies that subcontract other
companies that use slave labor.
15
ese modern, scattered, small-scale, and criminal slave-based enterprises
contrast markedly with the relatively much larger economic organization of
historical slavery, in which, as evil as it was, the economic benets of the legal
slave trade were diused throughout local and regional economies. In the past,
it could be assumed that a sizable proportion of commodities such as sugar and
cotton was produced by slaves and that the prots from such enterprises would
be cycled back through local, regional, and national economies along with the
prots from other legitimate businesses.
16
Today, however, the situation is more complicated in that it is slaveholders
who primarily benet from contempary slavery. Consider the cocoa industry
contemporary slavery, regardless of the type of slavery, is the exceptionally low cost
of acquiring slaves.
10
For most of human history, slaves were expensive. In 1850,
for example, the average price of an “average” slave in the American South (a
“prime eld hand” meaning a young man capable of heavy agricultural work) was
between $1,000 and $1,200 in 1850 dollars, which equates to around $45,000
in 2013 dollars.
11
Today it could be argued that slaves are less expensive today
than at any time in the past. In a marketplace in the central region of the Ivory
Coast, one researcher easily purchased two 19-year-old agricultural workers for
about $30 each. In India, bonded laborers are able to report precise sums that
represent the acquisition cost required to enslave their family (often the debt
being taken some generations previously and passed down)—ranging from as
little as $10 to over $100. In Brazil desperate urban workers place themselves
in the hands of recruiters only to nd themselves held under violence far in the
forests and charged with a “debt” amounting to a few hundred dollars. Even in
ailand, where teenagers are sold into prostitution and a premium is charged
for their virginity and attractiveness, prices rarely go over $1000.
Table 1. Monthly income and expenditures (in baht) for the always prospering
brothel.U.S. gures adjusted for ination for 2012. Source: Kevin Bales, Disposable
People: New Slavery in the Global Economy (University of California Press, 1999).
Low initial acquisition costs means that slave owners neglect the welfare
Table 1. Monthly Income & Expenditure (in Baht) for the Always Prospering Brothel
Costs (per month)
Income (per month)
Rent
5,000
Commercial sex*
1,050,000
Utilities & bills
2,000
Rent paid by prostitutes
600,000
Food & drink
45,000
Sale of condoms
70,000
Pimp’s salary
7,000
Sale of drinks
672,000
Cashier
7,000
Virgin premium
50,000
Cook
5,000
“Interest” on debt-bond
15,000
Bribes
6,000
Payments to taxis etc.
12,000
Beer & whisky
168,000
TOTAL COSTS
257,000
TOTAL INCOME
2,457,000
(In US dollars)
($8,389)
MONTHLY PROFIT
2,200,000
(In U.S. dollars)
($71,813)
*Average 14 clients per day at 125 baht per client for 20 prostitutes for 30 days.
Source: Disposable People: New Slavery in the Global Economy, University of California
Press, 1999. U.S. figures adjusted for inflation for the year 2012.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
210
211
Slavery generates a downward spiral of social and economic stuntedness. It is a
social and economic cancer.
Only very recently did scholars begin to examine the impact of slavery on
human development. In 2009 Robert Smith was among the rst to analyze a
number of factors that were thought to aect social and economic development
as measured by the United Nations Human Development Index; these factors
included corruption, regime type, internal conict, national debt, geography,
and culture.
20
For the rst time in a study of contemporary nation-states, Smith
included measures of slavery in the analysis of human development using an
earlier version of data from Kevin Bales of Free the Slaves, the center for the
modern abolitionist movement.
21
Smith’s study explored the impact that slavery might have on human de-
velopment and provided insight as to how slavery might t with other factors
that are known to aect development. Given the large amount of work that
has been done to explore determinants of development in the Global South, it
was remarkable that, as Smith explains, “When the full range of the covariates
are retained, and when the four categories of slavery are used to classify the re-
gions, the resulting ... statistics imply that slavery explains much of the regional
variability in the Human Development Index rank scores.”
22
Smith went on to
identify the policy implications:
Among the covariates studied here, slavery (debt bondage, forced labor,
forced prostitution, chattel slavery) and the lack of political freedom
explain much of the variability that is between regions and corruption
explains much of the variability among countries within a region.
Additionally, countries with higher values of conict and social unrest
and higher values of national debt have signicantly worse Human
Development Index rank. To enhance human development countries
should eliminate debt bondage and other forms of slavery, move toward
fully democratic political systems, and eliminate corruption.
23
Clearly there can be mutually reinforcing relationships among these factors.
ese interrelationships point to the circular and reinforcing nature of human
rights violations. In terms of human development and economic development,
the result from Smith’s analysis is clear: slavery impedes economies and societ-
ies even though the proportion of slaves in the global population is small, their
productive capacity is small, and their economic value, as measured by price,
has reached an all-time low.
in the twenty-rst century. In the Ivory Coast, there are about 600,000 small
family farms that grow cocoa, producing just under half of the world’s supply.
17
Young men, normally from neighboring countries such as Mali, migrate looking
for work. In remote rural areas, some are tricked and enslaved into working on
farms that grow a range of crops, including cocoa for export. While no one knows
for certain how many of these farms use slaves, the percentage of farms using
slaves is likely to be in the single digits.
18
When the farmers sell their cocoa to
wholesalers, the slave-grown cocoa is mixed with the free-labor cocoa, and there
is currently no way to tell the two apart.
e farmer who exploits slaves and the
farmer that does not will receive exactly
the same price for their cocoa at the farm
gate. In this way, the prots of slavery
are retained by the slaveholder and are not passed up along the product chain.
e cocoa that is exported from the Ivory Coast to Europe and North America
to be made into chocolate, used in other foodstus, or made into cosmetics is
also mixed with cocoa from other countries; thus, it becomes even harder to
distinguish the fraction that is tainted by slavery.
19
is concealment of slave-
made commodities, made possible through blending in global supply chains, also
applies to cotton, sugar, gold, shrimp, and a host of other products. What the
case of cocoa illustrates, and what applies to virtually all slave-based businesses,
is that prots from slavery tend to accrue solely to the criminal slaveholder and
are not passed up the product chain in the form of lower prices to consumers.
the effects of slavery on huMan develoPMent
Contemporary slavery yields a number of negative social and economic eects.
Slaves, subjects of violence and intimidation on a systematic basis, are unwilling
agents of economic stagnation. From a scal perspective, slave labor tends to
pull down the wages of free labor in the same area, thus lessening the disposable
incomes of free families. At the same time, slaves are not able to play a full role
in their local economy in that they have little or no purchasing power and so
they subsist on the barest necessities.
More generally, slaves, by the nature of their social position, have no means
to contribute to the economy except in minimal ways. ey are not at liberty to
purchase anything in a country’s markets. erefore, slaves, as a group—denied
the benets of human development in the form of education, access to politi-
cal participation, and self-direction—remain an untapped economic resource.
The prots of slavery are retained
by the slaveholder and are not
passed up along the product chain.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
210
211
Slavery generates a downward spiral of social and economic stuntedness. It is a
social and economic cancer.
Only very recently did scholars begin to examine the impact of slavery on
human development. In 2009 Robert Smith was among the rst to analyze a
number of factors that were thought to aect social and economic development
as measured by the United Nations Human Development Index; these factors
included corruption, regime type, internal conict, national debt, geography,
and culture.
20
For the rst time in a study of contemporary nation-states, Smith
included measures of slavery in the analysis of human development using an
earlier version of data from Kevin Bales of Free the Slaves, the center for the
modern abolitionist movement.
21
Smith’s study explored the impact that slavery might have on human de-
velopment and provided insight as to how slavery might t with other factors
that are known to aect development. Given the large amount of work that
has been done to explore determinants of development in the Global South, it
was remarkable that, as Smith explains, “When the full range of the covariates
are retained, and when the four categories of slavery are used to classify the re-
gions, the resulting ... statistics imply that slavery explains much of the regional
variability in the Human Development Index rank scores.”
22
Smith went on to
identify the policy implications:
Among the covariates studied here, slavery (debt bondage, forced labor,
forced prostitution, chattel slavery) and the lack of political freedom
explain much of the variability that is between regions and corruption
explains much of the variability among countries within a region.
Additionally, countries with higher values of conict and social unrest
and higher values of national debt have signicantly worse Human
Development Index rank. To enhance human development countries
should eliminate debt bondage and other forms of slavery, move toward
fully democratic political systems, and eliminate corruption.
23
Clearly there can be mutually reinforcing relationships among these factors.
ese interrelationships point to the circular and reinforcing nature of human
rights violations. In terms of human development and economic development,
the result from Smith’s analysis is clear: slavery impedes economies and societ-
ies even though the proportion of slaves in the global population is small, their
productive capacity is small, and their economic value, as measured by price,
has reached an all-time low.
in the twenty-rst century. In the Ivory Coast, there are about 600,000 small
family farms that grow cocoa, producing just under half of the world’s supply.
17
Young men, normally from neighboring countries such as Mali, migrate looking
for work. In remote rural areas, some are tricked and enslaved into working on
farms that grow a range of crops, including cocoa for export. While no one knows
for certain how many of these farms use slaves, the percentage of farms using
slaves is likely to be in the single digits.
18
When the farmers sell their cocoa to
wholesalers, the slave-grown cocoa is mixed with the free-labor cocoa, and there
is currently no way to tell the two apart.
e farmer who exploits slaves and the
farmer that does not will receive exactly
the same price for their cocoa at the farm
gate. In this way, the prots of slavery
are retained by the slaveholder and are not passed up along the product chain.
e cocoa that is exported from the Ivory Coast to Europe and North America
to be made into chocolate, used in other foodstus, or made into cosmetics is
also mixed with cocoa from other countries; thus, it becomes even harder to
distinguish the fraction that is tainted by slavery.
19
is concealment of slave-
made commodities, made possible through blending in global supply chains, also
applies to cotton, sugar, gold, shrimp, and a host of other products. What the
case of cocoa illustrates, and what applies to virtually all slave-based businesses,
is that prots from slavery tend to accrue solely to the criminal slaveholder and
are not passed up the product chain in the form of lower prices to consumers.
the effects of slavery on huMan develoPMent
Contemporary slavery yields a number of negative social and economic eects.
Slaves, subjects of violence and intimidation on a systematic basis, are unwilling
agents of economic stagnation. From a scal perspective, slave labor tends to
pull down the wages of free labor in the same area, thus lessening the disposable
incomes of free families. At the same time, slaves are not able to play a full role
in their local economy in that they have little or no purchasing power and so
they subsist on the barest necessities.
More generally, slaves, by the nature of their social position, have no means
to contribute to the economy except in minimal ways. ey are not at liberty to
purchase anything in a country’s markets. erefore, slaves, as a group—denied
the benets of human development in the form of education, access to politi-
cal participation, and self-direction—remain an untapped economic resource.
The prots of slavery are retained
by the slaveholder and are not
passed up along the product chain.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
212
213
the 0.001-level and explains about 55 percent of the variance we observe (R
2
=
0.55). is nding substantiates Smith’s reasoning that slavery is a useful factor in
understanding variation in social and economic development around the globe.
Figure 1. e relationship between the Human Development Index and slavery.
Of course, correlation does not imply causation. Simply because there is
a strong bivariate relationship between HDI and slavery does not mean that
the latter causes the former. Indeed, the relationship may be spurious. In this
instance, to control for the possibility of a spurious relationship, it is important to
account for other competing explanations that may shed light on cross-national
variation in HDI, independent of slavery. If, while holding these other factors
constant, the relationship between slavery and HDI maintains statistical signi-
cance, then we have greater support for our hypothesis. In this light we include
three control variables, the rst of which accounts for cross-national variation
in literacy rates, given that scholarship suggests that education is a powerful
predictor of human development and economic growth.
28
We measure education
in terms of literacy (on a scale from 0 percent to 100 percent) from the United
Nations Development Program.
29
e second variable examines cross-national
variation in regime type, given that some scholars argue that higher levels of
democratization lead to greater levels of productivity and growth.
30
We obtain
data for this measure from Freedom House, which ranks all the countries of
the relationshiP BetWeen slavery and the huMan develoPMent index
e next step in this research agenda is to explore the empirical relationship be-
tween slavery and human development. A challenge arises, however, in obtaining
data to approximate the prevalence of slavery around the globe. Estimates vary
based on diering approaches that use secondary source estimates and country-
level, random sample surveys. e ILO estimates that there are 21 million
slaves in the world today. Based on the data used by this article, however, this
gure is somewhat higher, upwards of 27 million.
24
Part of the reason for the
discrepancy is in the transparency of the research methods made available.
25
As
scholarship on human tracking and modern-day slavery develops over time,
the complete sharing of data and methods used will help resolve the dierence
observed in the number of slaves.
We begin by exploring the relationship between the prevalence of slavery
and the UN Human Development Index (HDI). e HDI, as the UN reports,
“is a summary measure of human development. It measures the average achieve-
ments in a country in three basic dimensions of human development: a long and
healthy life (health), access to knowledge (education), and a decent standard of
living (income).”
26
HDI is measured on an interval scale, from zero (represent-
ing “low development”) to one (indicating “high development”). We use data
from the year 2012. For the 162 countries in our dataset for which we have a
corresponding measure of slavery prevalence, we nd that the measure of HDI
has considerable range, with a minimum value of 3.04 percent (for the country
of Somalia) and a maximum value of 95.5 percent (for Norway). Countries on
average have an HDI score of 66.7 percent, which indicates “reasonably high
development.”
27
Our rst hypothesis is higher levels of the prevalence of slavery lead to lower
levels of HDI, all else being equal. In other words, we believe that slavery is a
powerful predictive factor in understanding cross-national variation in human
development. We test this in two ways, rst by observing the graphical relation-
ship between the two variables (HDI and Prevalence) as Figure 1 illustrates.
In Figure 1, our measure of slavery is on the x axis (in which we have
taken the log of this measure so as to normalize its distribution) and the HDI
measure is on the y axis. Each dot in Figure 1 represents one of the countries
in our dataset. Note the downward sloping shape of the graph, which suggests
that, as we hypothesized, higher levels of slavery correlate with lower levels of
HDI. Conversely, lower levels of slavery correspond to higher levels of HDI. A
bivariate regression analysis yields that this nding is statistically signicant at
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
212
213
the 0.001-level and explains about 55 percent of the variance we observe (R
2
=
0.55). is nding substantiates Smith’s reasoning that slavery is a useful factor in
understanding variation in social and economic development around the globe.
Figure 1. e relationship between the Human Development Index and slavery.
Of course, correlation does not imply causation. Simply because there is
a strong bivariate relationship between HDI and slavery does not mean that
the latter causes the former. Indeed, the relationship may be spurious. In this
instance, to control for the possibility of a spurious relationship, it is important to
account for other competing explanations that may shed light on cross-national
variation in HDI, independent of slavery. If, while holding these other factors
constant, the relationship between slavery and HDI maintains statistical signi-
cance, then we have greater support for our hypothesis. In this light we include
three control variables, the rst of which accounts for cross-national variation
in literacy rates, given that scholarship suggests that education is a powerful
predictor of human development and economic growth.
28
We measure education
in terms of literacy (on a scale from 0 percent to 100 percent) from the United
Nations Development Program.
29
e second variable examines cross-national
variation in regime type, given that some scholars argue that higher levels of
democratization lead to greater levels of productivity and growth.
30
We obtain
data for this measure from Freedom House, which ranks all the countries of
the relationshiP BetWeen slavery and the huMan develoPMent index
e next step in this research agenda is to explore the empirical relationship be-
tween slavery and human development. A challenge arises, however, in obtaining
data to approximate the prevalence of slavery around the globe. Estimates vary
based on diering approaches that use secondary source estimates and country-
level, random sample surveys. e ILO estimates that there are 21 million
slaves in the world today. Based on the data used by this article, however, this
gure is somewhat higher, upwards of 27 million.
24
Part of the reason for the
discrepancy is in the transparency of the research methods made available.
25
As
scholarship on human tracking and modern-day slavery develops over time,
the complete sharing of data and methods used will help resolve the dierence
observed in the number of slaves.
We begin by exploring the relationship between the prevalence of slavery
and the UN Human Development Index (HDI). e HDI, as the UN reports,
“is a summary measure of human development. It measures the average achieve-
ments in a country in three basic dimensions of human development: a long and
healthy life (health), access to knowledge (education), and a decent standard of
living (income).”
26
HDI is measured on an interval scale, from zero (represent-
ing “low development”) to one (indicating “high development”). We use data
from the year 2012. For the 162 countries in our dataset for which we have a
corresponding measure of slavery prevalence, we nd that the measure of HDI
has considerable range, with a minimum value of 3.04 percent (for the country
of Somalia) and a maximum value of 95.5 percent (for Norway). Countries on
average have an HDI score of 66.7 percent, which indicates “reasonably high
development.”
27
Our rst hypothesis is higher levels of the prevalence of slavery lead to lower
levels of HDI, all else being equal. In other words, we believe that slavery is a
powerful predictive factor in understanding cross-national variation in human
development. We test this in two ways, rst by observing the graphical relation-
ship between the two variables (HDI and Prevalence) as Figure 1 illustrates.
In Figure 1, our measure of slavery is on the x axis (in which we have
taken the log of this measure so as to normalize its distribution) and the HDI
measure is on the y axis. Each dot in Figure 1 represents one of the countries
in our dataset. Note the downward sloping shape of the graph, which suggests
that, as we hypothesized, higher levels of slavery correlate with lower levels of
HDI. Conversely, lower levels of slavery correspond to higher levels of HDI. A
bivariate regression analysis yields that this nding is statistically signicant at
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
214
215
variables, literacy and regime type, also hold statistical signicance.
Yet HDI does not capture total economic output, which prompts another
question: What is the relationship between slavery and overall economic output?
at is, to what extent does slavery inuence the average person’s ability to make
a decent living on the macro level?
An excellent measure for economic output is gross domestic product (GDP)
per capita in terms of purchasing power parity.
33
Countries vary considerably
in terms of how much the average citizen makes. Following the logic of Smith,
we argue that higher levels of slavery predict lower levels of GDP, all else being
equal. is is our second hypothesis. We obtain data for GDP per capita from
the World Bank’s “World Development Indicators” using data from the year
2011.
34
GDP per capita varies considerably, with a minimum value of $277 (for
Somalia), a maximum value of $82,978 (for Qatar), and a standard deviation
of $14,537. Figure 2 illustrates the relationship between slavery and GDP and
casts support for our second hypothesis.
Figure 2. e relationship between gross domestic product (GDP) and slavery.
Like the information presented in Figure 1, the data in Figure 2 illustrates
a downward-sloping, negative relationship between the outcome we seek to
understand (GDP per capita) and the key explanatory variable of interest (the
0 20000 40000 60000 80000
Gross Domestic Product
-10 -8 -6 -4 -2
log of Prevalence Measure
GDP (purchasing power parity) Fitted values
the globe on a linear scale from 2 (“highly free”) to 14 (“not free).
31
e third
variable captures cross-national variation in geography, given that dierences
in human development and productivity may be explained at least in part by
geographic diversity.
32
We account for this with dummy variables for Western
Europe, North America, Central and Eastern Europe, Asia, Latin America, and
the Middle East, using Sub-Saharan Africa as our baseline.
Slavery (log) -4.487***
(0.532)
Literacy 0.546***
(0.0343)
Regime Type -0.615***
(0.176)
Western Europe -2.698
(1.750)
North America -4.395
(13.85)
Central and Eastern Europe -2.293
(1.945)
Asia 1.926
(1.688)
Latin America -1.334
(1.756)
e Middle East -1.295
(1.682)
Constant -2.232
(4.014)
N 160
R
2
0.84
*p < 0.05, **p < 0.01, ***p < 0.001
Table 2. Human development indicators (robust standard errors in parentheses).
Table 2 presents the results of our rst multivariate model, in which
we regress HDI on our slavery measure while controlling for literacy rates,
regime type, and geographic variation. e results of Table 2 lend support for
our hypothesis. Holding literacy rates, regime type, and geography constant,
a one-percent increase in the prevalence of slavery is associated with a 0.045
percent decrease in human development, other things being equal. is nding
is signicant at the 0.001-level and lends credence to Smith’s claims. Two other
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
214
215
variables, literacy and regime type, also hold statistical signicance.
Yet HDI does not capture total economic output, which prompts another
question: What is the relationship between slavery and overall economic output?
at is, to what extent does slavery inuence the average person’s ability to make
a decent living on the macro level?
An excellent measure for economic output is gross domestic product (GDP)
per capita in terms of purchasing power parity.
33
Countries vary considerably
in terms of how much the average citizen makes. Following the logic of Smith,
we argue that higher levels of slavery predict lower levels of GDP, all else being
equal. is is our second hypothesis. We obtain data for GDP per capita from
the World Bank’s “World Development Indicators” using data from the year
2011.
34
GDP per capita varies considerably, with a minimum value of $277 (for
Somalia), a maximum value of $82,978 (for Qatar), and a standard deviation
of $14,537. Figure 2 illustrates the relationship between slavery and GDP and
casts support for our second hypothesis.
Figure 2. e relationship between gross domestic product (GDP) and slavery.
Like the information presented in Figure 1, the data in Figure 2 illustrates
a downward-sloping, negative relationship between the outcome we seek to
understand (GDP per capita) and the key explanatory variable of interest (the
0 20000 40000 60000 80000
Gross Domestic Product
-10 -8 -6 -4 -2
log of Prevalence Measure
GDP (purchasing power parity) Fitted values
the globe on a linear scale from 2 (“highly free”) to 14 (“not free).
31
e third
variable captures cross-national variation in geography, given that dierences
in human development and productivity may be explained at least in part by
geographic diversity.
32
We account for this with dummy variables for Western
Europe, North America, Central and Eastern Europe, Asia, Latin America, and
the Middle East, using Sub-Saharan Africa as our baseline.
Slavery (log) -4.487***
(0.532)
Literacy 0.546***
(0.0343)
Regime Type -0.615***
(0.176)
Western Europe -2.698
(1.750)
North America -4.395
(13.85)
Central and Eastern Europe -2.293
(1.945)
Asia 1.926
(1.688)
Latin America -1.334
(1.756)
e Middle East -1.295
(1.682)
Constant -2.232
(4.014)
N 160
R
2
0.84
*p < 0.05, **p < 0.01, ***p < 0.001
Table 2. Human development indicators (robust standard errors in parentheses).
Table 2 presents the results of our rst multivariate model, in which
we regress HDI on our slavery measure while controlling for literacy rates,
regime type, and geographic variation. e results of Table 2 lend support for
our hypothesis. Holding literacy rates, regime type, and geography constant,
a one-percent increase in the prevalence of slavery is associated with a 0.045
percent decrease in human development, other things being equal. is nding
is signicant at the 0.001-level and lends credence to Smith’s claims. Two other
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
216
217
ing our understanding of the inuence of slavery on GDP? ere is no perfect
answer to this question given the lack of data on the size of untraceable criminal
economic activity in each country, but in the case of slavery this uncounted
output should not dramatically alter the results
presented for two reasons. First, in virtually all
economies, the estimated value of slavery within
the criminal economy is much lower than the
sums generated by drugs, weapons, gambling,
and general corruption. Second, a signicant
part of slave-produced goods and economic activity are counted in GDP, but
without the fact of slave origin being known. Note that cocoa, cotton, iron,
shrimp and sh, and minerals for electronics can all have slave origin and still
be counted in national output. e key point here is that it is the slavery that
is hidden, not the export goods or the prots made on them.
Regardless of whether HDI or GDP per capita is used, a story begins to
emerge: slavery is associated with outcomes that are harmful not only for eco-
nomic and social development (as HDI illustrates) but also for economic output
per citizen (as GDP per capita illustrates). ese are macro estimates. We are not
arguing that the lives of most families and individuals are directly aected by
slavery—far from it. However, from a systemic perspective, the evidence suggests
that slavery signicantly depresses social and economic output at the aggregate
level. Based on these data, we nd further reason to substantiate Smith’s logic that
slavery yields serious negative outcomes of global proportions despite the small
ancillary benets that a small population of contemporary slaveholders reap.
Let us press on with Smith’s logic by exploring two more indicators of so-
cial productivity. One measure of interest is the access to nancial services that
the average citizen can obtain in a given country. Access to nancial services
is crucial for upward social mobility. Without access to credit, there is often
little the average family can do to improve its lot in life. We obtain data for this
measure from scholarship by Patrick Honohan, aliated with the World Bank,
Trinity College Dublin, and the Center for Economic and Political Research.
35
Honohan’s research has generated a cross-national measure of access to nancial
services that ranks on a scale from 0 (no access) to 100 (full access). e de-
scriptive statistics for this measure are interesting. e country with the lowest
ranking is Kyrgyzstan (which received a 1), and the country with the highest
is the Netherlands (which received a 100). e standard deviation is about 28
points, indicating a great deal of cross-national variation in this measure.
Our third hypothesis is higher levels of slavery will be correlated with lower
The key point here is that it
is the slavery that is hidden,
not the export goods or
the prots made on them.
prevalence of slavery by country). A bivariate regression analysis reveals that
the nding is statistically signicant at the 0.001-level and explains about 39
percent of the variance we observe (R
2
= 0.39). A multivariate regression analysis
(using the same control variables as we did for the HDI hypothesis) yields the
following output.
Slavery (log) -5260.3***
(734.8)
Literacy 143.0***
(41.38)
Regime Type -148.8
(364.4)
Western Europe 2753.9
(3429.1)
North America -3239.6
(2187.6)
Central and Eastern Europe 1475.0
(2631.1)
Asia -2496.7
(2249.3)
Latin America -3109.1
(2531.4)
e Middle East 204.8
(4837.5)
Constant -31747.7***
(7178.9)
N 160
R
2
0.44
*p < 0.05, **p < 0.01, ***p < 0.001
Table 3. Gross domestic product (robust standard errors in parentheses).
Table 3 sheds light on the extent to which the prevalence of slavery is asso-
ciated with GDP per capita. Holding literacy rates, regime type, and geography
constant, a one-percent increase in the prevalance of slavery is associated with
a $52.6 decrease in per capita GDP. us, a 10-percent increase corresponds to
a $526 decrease in GDP per capita—not a small sum.
At this point, it would be fair to ask if the measurement of slavery’s impact
on GDP is suciently robust given that the criminal market is not normally
included in any computation of GDP. Since the proceeds of the underground
economy are untraceable and thus untaxable, how might their existence be alter-
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
216
217
ing our understanding of the inuence of slavery on GDP? ere is no perfect
answer to this question given the lack of data on the size of untraceable criminal
economic activity in each country, but in the case of slavery this uncounted
output should not dramatically alter the results
presented for two reasons. First, in virtually all
economies, the estimated value of slavery within
the criminal economy is much lower than the
sums generated by drugs, weapons, gambling,
and general corruption. Second, a signicant
part of slave-produced goods and economic activity are counted in GDP, but
without the fact of slave origin being known. Note that cocoa, cotton, iron,
shrimp and sh, and minerals for electronics can all have slave origin and still
be counted in national output. e key point here is that it is the slavery that
is hidden, not the export goods or the prots made on them.
Regardless of whether HDI or GDP per capita is used, a story begins to
emerge: slavery is associated with outcomes that are harmful not only for eco-
nomic and social development (as HDI illustrates) but also for economic output
per citizen (as GDP per capita illustrates). ese are macro estimates. We are not
arguing that the lives of most families and individuals are directly aected by
slavery—far from it. However, from a systemic perspective, the evidence suggests
that slavery signicantly depresses social and economic output at the aggregate
level. Based on these data, we nd further reason to substantiate Smith’s logic that
slavery yields serious negative outcomes of global proportions despite the small
ancillary benets that a small population of contemporary slaveholders reap.
Let us press on with Smith’s logic by exploring two more indicators of so-
cial productivity. One measure of interest is the access to nancial services that
the average citizen can obtain in a given country. Access to nancial services
is crucial for upward social mobility. Without access to credit, there is often
little the average family can do to improve its lot in life. We obtain data for this
measure from scholarship by Patrick Honohan, aliated with the World Bank,
Trinity College Dublin, and the Center for Economic and Political Research.
35
Honohan’s research has generated a cross-national measure of access to nancial
services that ranks on a scale from 0 (no access) to 100 (full access). e de-
scriptive statistics for this measure are interesting. e country with the lowest
ranking is Kyrgyzstan (which received a 1), and the country with the highest
is the Netherlands (which received a 100). e standard deviation is about 28
points, indicating a great deal of cross-national variation in this measure.
Our third hypothesis is higher levels of slavery will be correlated with lower
The key point here is that it
is the slavery that is hidden,
not the export goods or
the prots made on them.
prevalence of slavery by country). A bivariate regression analysis reveals that
the nding is statistically signicant at the 0.001-level and explains about 39
percent of the variance we observe (R
2
= 0.39). A multivariate regression analysis
(using the same control variables as we did for the HDI hypothesis) yields the
following output.
Slavery (log) -5260.3***
(734.8)
Literacy 143.0***
(41.38)
Regime Type -148.8
(364.4)
Western Europe 2753.9
(3429.1)
North America -3239.6
(2187.6)
Central and Eastern Europe 1475.0
(2631.1)
Asia -2496.7
(2249.3)
Latin America -3109.1
(2531.4)
e Middle East 204.8
(4837.5)
Constant -31747.7***
(7178.9)
N 160
R
2
0.44
*p < 0.05, **p < 0.01, ***p < 0.001
Table 3. Gross domestic product (robust standard errors in parentheses).
Table 3 sheds light on the extent to which the prevalence of slavery is asso-
ciated with GDP per capita. Holding literacy rates, regime type, and geography
constant, a one-percent increase in the prevalance of slavery is associated with
a $52.6 decrease in per capita GDP. us, a 10-percent increase corresponds to
a $526 decrease in GDP per capita—not a small sum.
At this point, it would be fair to ask if the measurement of slavery’s impact
on GDP is suciently robust given that the criminal market is not normally
included in any computation of GDP. Since the proceeds of the underground
economy are untraceable and thus untaxable, how might their existence be alter-
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
218
219
Slavery (log) -10.54***
(1.515)
Literacy 0.123
(0.0891)
Regime Type -2.514***
(0.428)
Western Europe 2.675
(4.213)
North America 20.83*
(10.33)
Central and Eastern Europe 5.492
(4.690)
Asia -3.856
(5.112)
Latin America -9.024
(5.099)
e Middle East -7.879
(5.356)
Constant -17.56*
(8.821)
N 139
R
2
0.64
*p < 0.05, **p < 0.01, ***p < 0.001
Table 4. Access to nancial services (robust standard errors in parentheses).
e last measure we explore is the level of economic equality/inequality in
the world today. ere is evidence of a growing gap between the Global South
and the North. As the rich countries get richer, the poor countries get poorer. Our
fourth hypothesis is that greater levels of slavery should be correlated with higher
levels of global inequality. e measure we use to test this is the Gini coecient,
the data for which we obtain from the World Bank.
36
e Gini coecient is a
measure of income inequality, on a scale from 0 to 100, where 100 represents
perfect inequality and 0 signies perfect equality. us, a lower score on the
Gini coecient scale is better than a higher score. Among the data we have,
the country with the highest amount of income inequality is Namibia (a score
of 63.9) and the country with the lowest income inequality is Denmark (24.7).
levels of access to nancial service. We nd graphical evidence of this in Figure 3.
Figure 3. e relationship between access to nancial services and slavery.
Similar to Figures 1 and 2, Figure 3 illustrates a downward-sloping, nega-
tive relationship between access to nancial services (on the y axis) and the
prevalence of slavery (on the x axis). A bivariate regression analysis informs us
that this nding is also statistically signicant at the 0.001-level and explains
about 49 percent of the variance we observe (R
2
= 0.49). A multivariate analysis,
as seen in Table 4, reveals that a one-percent increase in the prevalence of slavery
is associated with about a tenth of a unit decrease in access to nancial services.
us, a ten-percent increase in slavery corresponds to about a one-unit decrease in
access to nancial services, other things being equal.
Once again, the data points to the destructive social and economic conse-
quences of contemporary slavery, but it is important to reect on the circular
nature of this particular relationship as well—lack of access to credit can place
poor families in extreme vulnerability to enslavement in times of crisis, in which
individuals and families are removed from the credit market, cut o from ever
being credit consumers.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
218
219
Slavery (log) -10.54***
(1.515)
Literacy 0.123
(0.0891)
Regime Type -2.514***
(0.428)
Western Europe 2.675
(4.213)
North America 20.83*
(10.33)
Central and Eastern Europe 5.492
(4.690)
Asia -3.856
(5.112)
Latin America -9.024
(5.099)
e Middle East -7.879
(5.356)
Constant -17.56*
(8.821)
N 139
R
2
0.64
*p < 0.05, **p < 0.01, ***p < 0.001
Table 4. Access to nancial services (robust standard errors in parentheses).
e last measure we explore is the level of economic equality/inequality in
the world today. ere is evidence of a growing gap between the Global South
and the North. As the rich countries get richer, the poor countries get poorer. Our
fourth hypothesis is that greater levels of slavery should be correlated with higher
levels of global inequality. e measure we use to test this is the Gini coecient,
the data for which we obtain from the World Bank.
36
e Gini coecient is a
measure of income inequality, on a scale from 0 to 100, where 100 represents
perfect inequality and 0 signies perfect equality. us, a lower score on the
Gini coecient scale is better than a higher score. Among the data we have,
the country with the highest amount of income inequality is Namibia (a score
of 63.9) and the country with the lowest income inequality is Denmark (24.7).
levels of access to nancial service. We nd graphical evidence of this in Figure 3.
Figure 3. e relationship between access to nancial services and slavery.
Similar to Figures 1 and 2, Figure 3 illustrates a downward-sloping, nega-
tive relationship between access to nancial services (on the y axis) and the
prevalence of slavery (on the x axis). A bivariate regression analysis informs us
that this nding is also statistically signicant at the 0.001-level and explains
about 49 percent of the variance we observe (R
2
= 0.49). A multivariate analysis,
as seen in Table 4, reveals that a one-percent increase in the prevalence of slavery
is associated with about a tenth of a unit decrease in access to nancial services.
us, a ten-percent increase in slavery corresponds to about a one-unit decrease in
access to nancial services, other things being equal.
Once again, the data points to the destructive social and economic conse-
quences of contemporary slavery, but it is important to reect on the circular
nature of this particular relationship as well—lack of access to credit can place
poor families in extreme vulnerability to enslavement in times of crisis, in which
individuals and families are removed from the credit market, cut o from ever
being credit consumers.
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
220
221
utes little to national production. Slaves are normally not able to acquire assets
or access credit and just as it is for the free working poor, credit access and asset
acquisition are key determining factors for achievement of economic autonomy.
Except in the way they benet criminals, we assert that slaves, though practically
invisible, exert a strong, negative pull on local and national economies through
their lack of full economic participation.
At the same time, there is a fruitful area for future research in the posi-
tive impact on local, regional, and national economies that coincides with the
liberation of those in slavery. Recent longitudinal research by Free the Slaves
among villages in northern India that have high levels of hereditary debt-bondage
slavery demonstrates a dramatic increase in economic activity (as well as human
development measures) following liberation.
37
e eradication of slavery has
long been sought on moral and philisophical grounds; this new research suggests
that the rationale for liberation might be extended to economic ends as well.
While slavery tends to deaden economic activity, it appears that liberation may
be an important economic stimulus. is is important because the resources
needed to reduce global slavery, while relatively small when compared to other
social problems, are not forthcoming.
38
e idea that the eradication of slavery
might ultimately pay for itself through economic growth is one that needs
demonstration. If “slavery is bad for business” is a hypothesis worth exploring,
how much more so is its mirror assertion that freedom from slavery generates
economic growth?
notes
1. “President Barack Obama speech to Clinton Global Initiative,” Politico, September 25, 2012; “Human
tracking: Mira Sorvino ghts against modern-day slavery,” Politico, December 6, 2012; “Jada Pinkett
Smith Hosts Sex-Tracking Awareness Party,” CBS News, October 30, 2012.
2. For example, see: Kevin Bales, “International Labor Standards: Quality of Information and Measures
of Progress in Combating Forced Labor,” Comparative Labor Law and Policy 24, no. 2 (2004); “21 Million
People are Now Victims of Forced Labour, ILO says,” International Labour Organization, June 1, 2012.
3. For example, see: Elżbieta M. Goździak and Micah N. Bump, “Data and Research on Human
Trafcking: Bibliography of Research-Based Literature,” October 2008.
4. Robert B. Smith, “Global human development: accounting for its regional disparities,” Quality and
Quantity 43, no. 1 (2009).
5. For example, see: Robert Fogel, Without Consent or Contract: e Rise and Fall of American Slavery:
Evidence and Methods (New York: Norton, 1992); Robert Fogel and Stanley Engerman, Time on the Cross:
e Economics of American Slavery (New York: Norton, 1974).
6. Of course, in no way do we condone the slavery of the past, as we see the morality of the issue as
the most important consideration. Slavery is a bloody and costly issue over the long term and can destroy
the fabric of a nation.
7. If these estimates are correct, the monetary value of slavery represents 0.00059 percent of the global
economy.
A
W
Figure 4. e relationship between the Gini coecient and slavery.
Figure 4 suggests that Smith’s thinking bears out: there is an upward-
sloping, positive relationship between the Gini coecient (on the y axis) and
our measure of the prevalence of slavery (on the x axis). Higher levels of slavery
correlate with greater levels of economic inequality. Although this nding is
signicant at the 0.001-level, the R
2
compared to those of the other models is
quite modest, explaining only about seven percent of the variance we observe.
A multivariate analysis, similar to the results of Tables 2, 3, and 4, reveals that
our measure of slavery prevalence is signicant at the 0.10-level in its association
with the Gini coecient, holding literacy, regime type, and geography constant.
conclusion
While slavery may make high prots for a small number of slaveholders, we argue
that the practice tends to disproportionately depress a country’s economy. Since
the work of slaves is generally concentrated at the lowest end of the economic
ladder in basic, low-skill jobs that are dirty and dangerous, slave output contrib-
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
220
221
utes little to national production. Slaves are normally not able to acquire assets
or access credit and just as it is for the free working poor, credit access and asset
acquisition are key determining factors for achievement of economic autonomy.
Except in the way they benet criminals, we assert that slaves, though practically
invisible, exert a strong, negative pull on local and national economies through
their lack of full economic participation.
At the same time, there is a fruitful area for future research in the posi-
tive impact on local, regional, and national economies that coincides with the
liberation of those in slavery. Recent longitudinal research by Free the Slaves
among villages in northern India that have high levels of hereditary debt-bondage
slavery demonstrates a dramatic increase in economic activity (as well as human
development measures) following liberation.
37
e eradication of slavery has
long been sought on moral and philisophical grounds; this new research suggests
that the rationale for liberation might be extended to economic ends as well.
While slavery tends to deaden economic activity, it appears that liberation may
be an important economic stimulus. is is important because the resources
needed to reduce global slavery, while relatively small when compared to other
social problems, are not forthcoming.
38
e idea that the eradication of slavery
might ultimately pay for itself through economic growth is one that needs
demonstration. If “slavery is bad for business” is a hypothesis worth exploring,
how much more so is its mirror assertion that freedom from slavery generates
economic growth?
notes
1. “President Barack Obama speech to Clinton Global Initiative,” Politico, September 25, 2012; “Human
tracking: Mira Sorvino ghts against modern-day slavery,” Politico, December 6, 2012; “Jada Pinkett
Smith Hosts Sex-Tracking Awareness Party,” CBS News, October 30, 2012.
2. For example, see: Kevin Bales, “International Labor Standards: Quality of Information and Measures
of Progress in Combating Forced Labor,” Comparative Labor Law and Policy 24, no. 2 (2004); “21 Million
People are Now Victims of Forced Labour, ILO says,” International Labour Organization, June 1, 2012.
3. For example, see: Elżbieta M. Goździak and Micah N. Bump, “Data and Research on Human
Trafcking: Bibliography of Research-Based Literature,” October 2008.
4. Robert B. Smith, “Global human development: accounting for its regional disparities,” Quality and
Quantity 43, no. 1 (2009).
5. For example, see: Robert Fogel, Without Consent or Contract: e Rise and Fall of American Slavery:
Evidence and Methods (New York: Norton, 1992); Robert Fogel and Stanley Engerman, Time on the Cross:
e Economics of American Slavery (New York: Norton, 1974).
6. Of course, in no way do we condone the slavery of the past, as we see the morality of the issue as
the most important consideration. Slavery is a bloody and costly issue over the long term and can destroy
the fabric of a nation.
7. If these estimates are correct, the monetary value of slavery represents 0.00059 percent of the global
economy.
A
W
Figure 4. e relationship between the Gini coecient and slavery.
Figure 4 suggests that Smith’s thinking bears out: there is an upward-
sloping, positive relationship between the Gini coecient (on the y axis) and
our measure of the prevalence of slavery (on the x axis). Higher levels of slavery
correlate with greater levels of economic inequality. Although this nding is
signicant at the 0.001-level, the R
2
compared to those of the other models is
quite modest, explaining only about seven percent of the variance we observe.
A multivariate analysis, similar to the results of Tables 2, 3, and 4, reveals that
our measure of slavery prevalence is signicant at the 0.10-level in its association
with the Gini coecient, holding literacy, regime type, and geography constant.
conclusion
While slavery may make high prots for a small number of slaveholders, we argue
that the practice tends to disproportionately depress a country’s economy. Since
the work of slaves is generally concentrated at the lowest end of the economic
ladder in basic, low-skill jobs that are dirty and dangerous, slave output contrib-
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
222
223
development-indicators.
35. Patrick Honohan, “Cross-Country Variation in Household Access to Financial Services” (prepared
for the conference “Access to Finance,” Washington, D.C., March 15–16, 2007), http://siteresources.
worldbank.org/INTFR/Resources/Cross_Country_Variation_In_Household_Access.pdf; Patrick Hono-
han, “Resume,” Trinity College Dublin, http://www.tcd.ie/Economics/sta/phonohan; “Home Page,”
Center for Economic and Policy Research, http://www.cepr.net.
36. “Gini Index,” World Bank, http://data.worldbank.org/indicator/SI.POV.GINI.
37. “Baseline study of slavery and poverty in Kukrouthi village, St. Ravidas Nagar District, Uttar Pradesh,
June 2010” (internal research report, Free the Slaves, Washington, D.C.); “Follow Up Study of Slavery
and Poverty in Kukrouthi village, St. Ravidas Nagar District, Uttar Pradesh, June 2012” (internal research
report, Free the Slaves, Washington, D.C.).
38. Bales, Ending Slavery: How We Free Today’s Slaves, 55.
8. For a full discussion of the nature of and prots derived from slave-based businesses, see: Kevin Bales,
Disposable People: New Slavery in the Global Economy (Berkeley: University of California Press, 1999).
9. Ibid.
10. e acquisition cost of slaves has decreased rapidly in the past 50 years. While once tending to be
treated as capital purchase investments, slaves today can be acquired for around $3,000 to $8,000 in the
richer developed countries, and as little as $20 in the poorest countries. For a full discussion of the fall in
acquisition cost of slaves, see: Kevin Bales, Disposable People.
11. We are indebted to Junius Rodriguez for helping us understand these comparisons.
12. Kevin Bales and Ron Soodalter, e Slave Next Door: Human Tracking and Slavery in America
Today (University of California Press, 2009), 44–47.
13. Ibid., 137–59.
14. Ibid.
15. Ibid.
16. Gene Dattel, “When Cotton Was King,” New York Times, March 26, 2011. For example, as Dat-
tel explains, “e number of slaves increased from 700,000 in 1787 to over 4 million on the eve of the
American Civil War; approximately 70 percent were involved in some way with cotton production. Indeed,
so closely tied were cotton and slavery that the price of a slave directly correlated to the price of cotton
(except during years of excessive speculation).”
17. Kevin Bales, Ending Slavery: How We Free Today’s Slaves (Berkeley: University of California Press,
2007), 184.
18. Ibid.
19. Ibid.
20. Robert B. Smith, “Global human development: accounting for its regional disparities,” Quality and
Quantity 43, no. 1 (2009); United Nations Human Development Index, United Nations Environmental
Programme, http://hdr.undp.org/en/statistics.
21. “Home Page,” Free the Slaves, http://www.freetheslaves.net.
22. Ibid., 25.
23. Ibid.
24. For the most recent published estimates of slavery from Bales, see: “e Social Psychology of
Modern Slavery,” Scientic American, April 24, 2002. More updated gures on the prevalence of slavery
are forthcoming.
25. To date, the ILO has not made its data publicly available nor has it submitted its estimates of slavery
to an academic journal for peer review.
26. “Frequently Asked Questions (FAQs),” UN Human Development Index (HDI), http://hdr.undp.
org/en/statistics/hdi.
27. Although we are indebted to the United Nations for the data it has made available in its Human
Development Index, we nd it normatively troublesome that some countries can be more “developed”
than others. is, in our opinion, smacks of classism, but that is a matter for another paper.
28. For example, see: Robert J. Barro, “Education and Economic Growth” (OECD, 2001), http://
www1.oecd.org/edu/country-studies/1825455.pdf.
29. “International Human Development Indicators,” United Nations Development Program, http://
hdrstats.undp.org/en/indicators/101406.html.
30. For example, see: Adam Przeworki, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando
Limongi, Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990
(Cambridge University Press, 2000).
31. “Home Page,” Freedom House, http://www.freedomhouse.org.
32. For example, see: John Luke Gallup, Jerey D. Sachs, and Andrew D. Mellinger, “Geography and
Economic Development,” (working paper, National Bureau of Economic Research), http://www.nber.
org/papers/w6849.
33. Purchasing power parity allows us a more standardized way of making comparisons in wealth across
countries.
34. “World Development Indicators,” World Bank, http://data.worldbank.org/data-catalog/world-
M N D K B
Slavery is Bad for Business
S/S
2013
• ,
222
223
development-indicators.
35. Patrick Honohan, “Cross-Country Variation in Household Access to Financial Services” (prepared
for the conference “Access to Finance,” Washington, D.C., March 15–16, 2007), http://siteresources.
worldbank.org/INTFR/Resources/Cross_Country_Variation_In_Household_Access.pdf; Patrick Hono-
han, “Resume,” Trinity College Dublin, http://www.tcd.ie/Economics/sta/phonohan; “Home Page,”
Center for Economic and Policy Research, http://www.cepr.net.
36. “Gini Index,” World Bank, http://data.worldbank.org/indicator/SI.POV.GINI.
37. “Baseline study of slavery and poverty in Kukrouthi village, St. Ravidas Nagar District, Uttar Pradesh,
June 2010” (internal research report, Free the Slaves, Washington, D.C.); “Follow Up Study of Slavery
and Poverty in Kukrouthi village, St. Ravidas Nagar District, Uttar Pradesh, June 2012” (internal research
report, Free the Slaves, Washington, D.C.).
38. Bales, Ending Slavery: How We Free Today’s Slaves, 55.
8. For a full discussion of the nature of and prots derived from slave-based businesses, see: Kevin Bales,
Disposable People: New Slavery in the Global Economy (Berkeley: University of California Press, 1999).
9. Ibid.
10. e acquisition cost of slaves has decreased rapidly in the past 50 years. While once tending to be
treated as capital purchase investments, slaves today can be acquired for around $3,000 to $8,000 in the
richer developed countries, and as little as $20 in the poorest countries. For a full discussion of the fall in
acquisition cost of slaves, see: Kevin Bales, Disposable People.
11. We are indebted to Junius Rodriguez for helping us understand these comparisons.
12. Kevin Bales and Ron Soodalter, e Slave Next Door: Human Tracking and Slavery in America
Today (University of California Press, 2009), 44–47.
13. Ibid., 137–59.
14. Ibid.
15. Ibid.
16. Gene Dattel, “When Cotton Was King,” New York Times, March 26, 2011. For example, as Dat-
tel explains, “e number of slaves increased from 700,000 in 1787 to over 4 million on the eve of the
American Civil War; approximately 70 percent were involved in some way with cotton production. Indeed,
so closely tied were cotton and slavery that the price of a slave directly correlated to the price of cotton
(except during years of excessive speculation).”
17. Kevin Bales, Ending Slavery: How We Free Today’s Slaves (Berkeley: University of California Press,
2007), 184.
18. Ibid.
19. Ibid.
20. Robert B. Smith, “Global human development: accounting for its regional disparities,” Quality and
Quantity 43, no. 1 (2009); United Nations Human Development Index, United Nations Environmental
Programme, http://hdr.undp.org/en/statistics.
21. “Home Page,” Free the Slaves, http://www.freetheslaves.net.
22. Ibid., 25.
23. Ibid.
24. For the most recent published estimates of slavery from Bales, see: “e Social Psychology of
Modern Slavery,” Scientic American, April 24, 2002. More updated gures on the prevalence of slavery
are forthcoming.
25. To date, the ILO has not made its data publicly available nor has it submitted its estimates of slavery
to an academic journal for peer review.
26. “Frequently Asked Questions (FAQs),” UN Human Development Index (HDI), http://hdr.undp.
org/en/statistics/hdi.
27. Although we are indebted to the United Nations for the data it has made available in its Human
Development Index, we nd it normatively troublesome that some countries can be more “developed”
than others. is, in our opinion, smacks of classism, but that is a matter for another paper.
28. For example, see: Robert J. Barro, “Education and Economic Growth” (OECD, 2001), http://
www1.oecd.org/edu/country-studies/1825455.pdf.
29. “International Human Development Indicators,” United Nations Development Program, http://
hdrstats.undp.org/en/indicators/101406.html.
30. For example, see: Adam Przeworki, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando
Limongi, Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990
(Cambridge University Press, 2000).
31. “Home Page,” Freedom House, http://www.freedomhouse.org.
32. For example, see: John Luke Gallup, Jerey D. Sachs, and Andrew D. Mellinger, “Geography and
Economic Development,” (working paper, National Bureau of Economic Research), http://www.nber.
org/papers/w6849.
33. Purchasing power parity allows us a more standardized way of making comparisons in wealth across
countries.
34. “World Development Indicators,” World Bank, http://data.worldbank.org/data-catalog/world-
