Can We Determine the Optimal Size of Government?, Cato Development Briefing Paper No. 7

8/4/2019 Can We Determine the Optimal Size of Government?, Cato Development Briefing Paper No. 7

1/12

the cato institute

1000 Massachusetts Avenue, N.W., Washington, D.C., 20001-5403

www.cato.org

Phone (202) 842-0200 Fax (202) 842-3490

S e p t e m b e r 1 4 , 2 0 1 1 n o . 7

Can We Determinethe Optimal Size of Government?

by James A. Kahn

T

he massive spending programs and new regula-tions adopted by many countries around theworld in response to the economic crisis of 2008

have drawn renewed attention to the role of governmentin the economy. Studies of the relationship between gov-ernment size and economic growth have come up with awide range of estimates of the optimal or growth-maxi-mizing size of government, ranging anywhere between 15and 30 percent of gross domestic product (GDP).

This paper argues that such an exercise is ill conceived.Modern growth economics suggests, first, that govern-ment policies leave their long-term impact primarily onthe level of economic activity, not the growth rate; and,

second, that the sources of this impact are multi-dimensional and not necessarily well measured by conventionalmeasures of size, such as the share of government spend

ing in GDP.In fact, measures of economic freedom more closely re-

late to per capita GDP than do simple measures of government spending. The evidence shows that governments aregenerally larger than optimal, but because the availabledata include primarily countries whose governments aretoo large, it cannot plausibly say what the ideal size of gov-ernment is. The data can realistically only say that smallergovernments are better, and suggest that the optimal sizeof government is smaller than what we observe today.

James A. Kahn is the Henry and Bertha Kressel Professor of Economics at Yeshiva University.

Executive Summary


2/12

2

Governmentpolicies leave

their long-termimpact primarily

on the levelof economic

activity, not thegrowth rate.

Introduction

The massive spending programs andnew regulations adopted by many countriesaround the world in response to the eco-

nomic crisis of 2008 have drawn renewedattention to the role of government in theeconomy. Recent research relating the size ofgovernment to the pace of economic growthhas suggested that the growth-maximizingsize of government may be as large as 30percent of gross domestic product (GDP).1This paper will argue that such a conclusionis not warranted by the evidence. Moreover,much of this research asks the wrong ques-tions with the wrong data, and consequentlycannot hope to come up with a plausible an-

swer to the question of what is the optimalsize of government.

The first part of this paper briefly reviewsthe literature on economic growth. Two im-portant messages from this body of researchare, first, that government policies leavetheir long-term impact primarily on the lev-el of economic activity, not the growth rate;and, second, that the sources of this impactare multi-dimensional and not necessarilywell measured by conventional measuresof size, such as the share of government

spending in GDP.The paper then examines the available

evidence on the impact of governmentsizewhich is broadly defined in terms ofeconomic freedomon per capita GDP. Theconclusion is that governments are gener-ally too large: that is, that greater economicfreedom results in a higher average standardof living. As with the earlier studies, this evi-dence cannot plausibly say what the idealsize of government is, only that virtually allgovernments are larger than optimal from

the standpoint of maximizing economicactivity. Because the available data includeprimarily countries whose governments aretoo large, econometricslike a lost hikertrying to descend a mountaincan realisti-cally only say what direction to go (down),not how far one must go to reach the desti-nation.2

Lessons from Researchon Growth

The study of economic growth experi-enced a resurgence in the 1980s. For the most

part, this work reaffirmed the basic conclu-sion of Nobel Prizewinning economistRobert Solow that economic policies have alasting impact primarily on the levelof eco-nomic activity, not the growth rate.3 Differ-ences in growth rates between two countriestend to be transitory. Longer-term growthtrends are largely a function of worldwidetechnological advance. Hard as it is to be-lieve, poor countries, on average, grow aboutas fast as rich countries, although there isa much wider dispersion of growth rates at

the poor end of the distribution. Changes ingovernment policies (or other events, suchas changes in commodity prices) may resultin temporary bursts of growth or periods ofrelative stagnation, but eventually countriestend to level off to a normal rate of growth,albeit at different levels. Sometimes, as in thecase of postwar Japan or modern-day Chinathese growth episodes can last decades, espe-cially if a country starts from a very low levelof per capita GDP. (This suggests that majorpolicy changes can produce higher growth

for a substantial period of time, but thereare relatively few such examples.) It is a mis-take, however, to extrapolate current growthrates. Those who predicted, in the 1980s,that Japan would surpass the United Statesin per capita GDP learned this the hard way,yet many so-called experts are now predict-ing that Chinas total GDP will surpass thatof the United States by 2030.4

The implication is that growth rates areonly a transitory indicator of economicpolicies. Strong growth may indicate im

proved policies, but the improvement may befrom a very low level, and the growth maynot persist more than a few years.5 Zimba-bwe could probably experience double-digitgrowth with a few relatively modest reformssimply because it is starting at such a lowlevel and with such horrendous economicpolicies; but without a major transformation,


3/12

3

A long-runnegative impactof governmentoutlays on percapita GDP findsupport in theliterature.

that growth will peter out long before thecountry even begins to close its gap witheven its more prosperous (but still relativelypoor) neighbors, such as South Africa andNamibia, never mind with more advanced

economies.More generally, countries can be thoughtof as either in a balanced growth state, inwhich per capita GDP growth fluctuatesaround the normal long-term rate of abouttwo percent annually, or in transitionmeaning sustained growth at an above orbelow normal rate until they reach the bal-anced growth state. China is obviously intransition; West Germany and Japan werein transition during their decades-long re-covery from World War II. Most of Western

Europe, the United States, and Canada weremore or less in balanced growth over thepast 20 years up until 2007, when the finan-cial crisis hit and government policies devi-ated wildly from the past.

When countries do level off to normalgrowth rates, they can be at widely differentlevels of GDP per capita, depending on gov-ernment policies and other fundamentals.This is what economists refer to as condi-tional convergence: countries with similarpolicies, institutions, and demographics will

eventually end up at similar levels of GDP percapita.6 Thus, Western European countriesappear to have converged to a level about70 to 75 percent of U.S. per capita GDP (al-though with Spain and Italy somewhat low-er). Canada and Australia, whose fundamen-tals lie somewhere between those of WesternEurope and the United States, converged to80 to 85 percent of U.S. per capita GDP.

All of this suggests that the relationshipbetween per capita GDP and economic poli-cies is not easily captured by looking at cor-

relations at a point in time, unless countrieshave had stable policies for long enough tohave converged to their balanced growthpaths. The relationship of policies to GDPgrowth is even more complex, as this essen-tially requires countries to be in transition,and therefore their starting points will comeinto play. China can experience double-digit

growth despite a large government presencein the economy simply because it startedfrom an impoverished state with even moregovernment control.

Given this discussion, we will focus on

the relationship between levels of GDP percapita and government policies for coun-tries that have had relatively stable policies(at least until 2007). Figure 1 looks at the re-lationship between government outlays as ashare of GDP and GDP per capita for 18 advanced economies. It uses the 1996 spend-ing share with 2008 per capita GDP data,the idea being that after the volatile 1970sand 1980s, government spending had set-tled down to a steady share in most of thesecountries by the mid-1990s, and any transi-

tional effects would have dissipated by 2008.Even for this relatively homogeneous group,however, while there is a negative relation-ship, it is not especially strong or tight.7

The weak relationship in Figure 1 shouldnot be surprising. Many factors influenceper capita GDP other than the size of govern-ment, even among broadly similar advancedeconomies. Policies related to human capital(education and immigration, for example)can vary widely. Countries may differ in theirtax systems (progressivity, capital taxation,

social insurance) quite apart from the scaleof government. Governments of similar sizemay differ in the scope of their activities. Andeven advanced economies may be influencedby history and still be in transition, even ifnot as dramatically as China in the last de-cade or Japan after World War II. Finally, wecannot rule out reverse causalitywealthiercountries may demand more or fewer gov-ernment services.

Despite these caveats, a long-run negativeimpact of government outlays on per capita

GDP finds support in the literature. Harvardeconomist Robert Barro, for example, con-trols for a number of other factors and findsthat even a narrow measure of governmentspending relative to GDP has an adverse im-pact on subsequent growth.8 James Gwart-ney et al. find a strong negative correlationbetween government outlays and subse-


4/12

4

Much of thedifference in

hours of workper capita

between theUnited Statesand Europe is

attributable tothe differences intheir tax systems.

quent growth in data that cover a somewhat

larger set of countries over a longer time pe-riod.9 Even if the true structural long-termrelationship is between government spend-ing and the level of per capita GDP, we couldsee growth effects for a substantial period oftime if countries with high spending tendalso to have increased the size or scope oftheir governments. The growth effects reflectthe transition to a level of GDP per capita be-low where it would have been without thegrowth in government.

The Size of Governmentand Labor Supply

The most straightforward mechanismlinking the size of the government budgetto the level of income per capita is throughthe impact of marginal tax rates on labor

supply. One way to think of this is that GDP

can be decomposed into two componentsoutput per hour (labor productivity) and to-tal hours of work. Many studies have shownthat European economies have similar pro-ductivity to the United States but much low-er hours of work. But here again, a simplenotion of size may be inadequate. Even twocountries with exactly the same ratio of gov-ernment expenditure to GDP may have dif-ferent income-tax structures that give rise todifferent levels of employment. One countrymay have a relatively flat tax, so that mar-

ginal rates are not much higher than averagerates, whereas another country may have avery progressive structure that penalizes extra work effort on the margin. With similaraverage tax rates, economic theory predictsthat the country with the more progressivesystem will have lower hours of employ-ment.

GDPpercapita2008(Thosandsof$U

S)

Government Outlays, 1996 (Share of GDP)

Figure 1

Government Spending and Per Capita GDP

Source: Authors work, based on data from the Fraser Institute, http://freetheworld.com.

Note: Data come from the OECD, in year 2000 $US, PPP adjusted. Government outlays includes transfers.


5/12

5

A major factoraccountingfor poverty isinefficiency.

This prediction is borne out by a compar-ison between the United States and WesternEuropean economies. Nobel Prizewinningeconomist Edward Prescott argues thatmuch of the difference in hours of work

per capita between the United States andcontinental Europe is attributable to thedifferences in their tax systems.10 This in-cludes not just income taxes, however, butalso their unemployment and social secu-rity systems.11 Countries such as France andGermany have relatively high marginal taxrates and generous unemployment insur-ance. While having similar productivity toworkers in the United States, the French andGermans work only about 70 percent of thehours of their U.S. counterparts; hence the

difference in GDP per capita.It is sometimes argued that the lower la-

bor supply in Europe is the result of culturaldifferences (or tastes), not taxation or regula-tion. Some have even gone so far to say thatthe European preference for leisure is supe-rior to the American inclination toward hardwork. Economist Juliet Schor argues thatAmericans propensity for hard work is likethe bad outcome of a Prisoners Dilemmaifwe could all agree to work less we would allbe happier.12 In that scenario, government-

mandated vacation time and workweeklimits make us all better off. Proponents ofthe European model, however, have not ex-plained the source of the market failure thatwould make individual voluntary decisionsresult in an inferior outcome. Moreover,Schors argument is premised on the ques-tionable proposition that leisure time hasdeclined. Recent work by Mark Aguiar andErik Hurst has shown that once other non-work activities such as household and schoolwork are taken into account, Americans lei-

sure time has trended upward.13

The Size of Governmentand Productivity

Once we move beyond the advancedeconomies and look at income per capita

more broadly, the vast differences aroundthe world are no longer driven by labor sup-ply, but by productivity. That is, people inimpoverished countries have low incomesnot primarily because they do not work as

hard as those in advanced economies, or evenbecause they are not as educated or skilled.While differences in human capital play animportant role, a major factor accountingfor poverty is inefficiency. Whatever naturalresources or innate endowments a countrymay have, including its labor force and capi-tal stock, if it fails to use them efficiently itwill leave itself relatively poor compared tocountries that do make more efficient use oftheir resources.

The government may have both a positive

and negative role in encouraging efficiency.A positive contribution may come from thegovernments production of public goods,including, for example, law enforcement,national defense, and some kinds of infra-structure. On the other hand, governmentsmay distort private decisions and push econ-omies away from efficient outcomes. Com-mon interventions include imposing restric-tions on international trade, or on labormarkets (such as minimum-wage laws andrestrictions on firing employees), and direct

government ownership and control of majorindustries. The latter is particularly commonin countries reliant on natural resource ex-traction.

Robert Hall and Charles Jones argue thatsocial infrastructure, by which they meaninstitutions to protect the output of in-dividual productive units from diversion,(e.g., confiscatory taxation, expropriation,and thievery) is a key determinant of a coun-trys level of productivity.14 They note thatwhile government can protect against diver-

sion, it is also in practice a primary agent ofdiversion.15 This conundrum is sometimesrendered as Quis custodiet ipsos custo-des? (Who watches the watchmen?) It is aplausible explanation for the apparent biastoward overly large and intrusive govern-ments throughout the world. We have toprovide government with sufficient powers


6/12

6

Whateverthe optimalgovernmentsize may be,

the normal stateof affairs is for

governmentsto grow beyond

that.

(taxation, military, legal) to do the neces-sary things: national defense, courts, police,and enforcement of contracts and propertyrights. Once so endowed, the temptation todo more or to take advantage is irresistible.16

William Easterly similarly emphasizes theimportance of maintaining incentives forindividuals to invest, work, and take risks.17He writes that governments can avoid kill-ing growth by avoiding . . . high inflation,high black market premiums, high budgetdeficits, strongly negative real interest rates,restrictions on free trade, excessive red tape,and inadequate public services. Only the lastof these represents a positive role for govern-ment, at least for services the private sectoris unable to provide; the rest are symptoms

of government excessfiscal irresponsibility,corruption, or just incompetence.18

The inescapable conclusion of this litera-ture is that while government can and doesplay a positive role in facilitating productiveactivity, it also has a natural tendency to di- vert resources toward itself, either via out-right corruption, excessive taxation, or thegradual accretion of bureaucratic interven-tion. So whatever the optimal governmentsize may be, the normal state of affairs is forgovernments to grow beyond that, to a point

where, on the margin, it exerts a negative in-fluence. Barros finding, mentioned earlier,of a negative effect of government spendingon economic growth (once initial conditionsand a variety of other variables are controlledfor) can be seen as evidence for this proposi-tion.19

Measuring theSize of Government

The two measures of government sizemost researchers use are the share of govern-ment purchases (in other words, governmentspending excluding transfer payments) inGDP and the ratio of total government ex-penditures to GDP. The argument for themore narrow measure is that, first, trans-fer payments, such as Social Security, do

not count toward GDP; and, second, thatmerely taking money from one person andtransferring it to another does not representgovernment economic activity in the sameway as, for example, the government taking

that money and providing a governmentservice, building a road, or purchasing mili-tary equipment.

Transfer payments impact on overalleconomic performance may be smaller thanthat of government purchases, but bothtypes of spending will increase average (andpresumably marginal) tax rates. Moreover,as discussed above, some government expen-dituresnational defense, police, and thecourt system being obvious examplesarefor public goods that provide some benefit

and may enhance economic performanceThe social benefits of transfers are morenebulous: possibly reduced poverty and in-equality, reduced income volatility, or socialstability.20 And frequently a large compo-nent of transfer payments is interest on thedebt, which just represents delayed paymentfor previous government purchases.

But neither of these measures adequatelycaptures the magnitude of governments rolein the economy. Many government actionscost little or nothing, but have a potentially

large impact on the economy. The primarybudget cost of government regulations andtrade restrictions is for enforcement, butthat is only a small fraction of their econom-ic impact. If the government puts a quotaon steel imports, restricts employers abilityto fire their workers, or nationalizes a majorindustry, it may have little or no impact onexpenditures yet have a significant effect onproductivity or GDP.

There have been several efforts to quantifythese broader notions of government size.

The social infrastructure index of Hall andJones, for example, emphasizes indicators ocorruption and diversion of resources.21 Forthe purposes of this paper, I will use the Fra-ser InstitutesEconomic Freedom of the World index, which encompasses measures of barriersto international trade, labor market freedomand currency soundness, among other vari-


7/12

7

Greater economfreedom resultsin higher GDPper capita, withno maximumin sight.

ables, in addition to government spending.22The relation between this index and GDP percapita is illustrated in Figure 2 for the same

group of countries included in Figure 1.This broader (inverse) measure of the

size of government clearly shows a stron-ger relationship with GDP per capita than just the ratio of government spending toGDP. It also shows no obvious leveling offor hump shape to suggest that there is anoptimal level of economic freedom withinthe observed range. The data simply suggestthat more economic freedom leads to great-er prosperity.

Moreover, even looking beyond these 18

countries, there are few examples of govern-ments that are smaller (i.e., that allow moreeconomic freedom) than the United States,23so there is no basis for concluding that thereis an optimal level of economic freedom orsize of government. Until we find substantialnumbers of countries with greater freedombut declining living standards, we can only

conclude that greater economic freedom re-sults in higher GDP per capita, with no max-imum in sight.24

Of course, reverse causality still may bean issue, even using the freedom index from1995, given the strong persistence of relativeincome levels. Perhaps wealthier societies de-mand more economic freedom. For example,relative income levels from 1985 still line upwell with economic freedom in 1995. To ad-dress these concerns, Figure 3 looks at chang-es in the index from 19851995 in relationto per capita GDP growth from 19952005. Again we see a strong positive relationship.The reverse causality argument is much hard-

er to make here, since it would require growthin the later period to have caused the increasein economic freedom 10 years earlier.25

Figure 3 further buttresses the view thateconomic freedom, which is really a broaderand more inclusive inverse measure of gov-ernment size, has a strong positive long-runinfluence on the level of GDP per capita.

Figure 2

GDP Per Capita and Economic Freedom


Note: GDP per capita is as in Figure 1, in year 2000 $US, PPP adjusted.

GDPpercapita2008(Thousandsof$US)

Economic Freedom Index, 1995


8/12

8

If the bias ofgovernment isalways toward

more spending,perhaps there

is too muchspending on corefunctions as well.

ConclusionMost political theorists have long under-

stood the necessity of some form of govern-ment to protect individuals and property,provide public goods, and enforce privatecontracts. They have also recognized the al-most inexorable tendency of governmentsto become overgrown and corrupt, and thus,on the margin, to exert a negative impact onproductive economic activity. Efforts to dis-cover the optimal size of government have

been plagued by issues of measurement(what is the right measure of size?) andmethodology (what is the right evidence tolook at and how should it be examined?).

On the measurement front, this paperhas argued for going beyond the size of thegovernment budget to include such factorsas regulations, price controls, and trade re-

strictions. Regarding methodology, it hasargued for approaches that take into ac-count simultaneity (the problem of dis-entangling cause-and-effect) and the mul-tidimensional nature of the problem, andwhich look for lasting effects on the level,rather than the growth rate, of economic ac-tivity. Surveying the literature, and lookingdirectly at broader measures of governmentsize, while the evidence does not allow us todetermine what the optimal size of govern-ment is, it does clearly indicate that, for the

most part, the governments we observe aretoo largeat least from the point of view ofmaximizing GDP per capita.

What is the ideal size of government? TheGwartney et al. estimate that the U.S. gov-ernment spends about 15 percent of GDPfor core functions provides some guid-ance, but even that may be misleading.26

Figure 3

Economic Freedom and Growth (Changes)

RealG

DPGrowthpercapita(percent),

1995

2005

Change in Economic Freedom, 19851995



9/12

9

Surely governments face tradeoffs, and tothe extent that they do too much of onething, they may do too little of another.27Alternatively, if the bias of government is al-ways toward more spending, perhaps there

is too much spending on core functions aswell. Once we accept that the vagaries of thepolitical process do not result in efficientlevels of spending in some areas, there is noreason to presume that the observed levelsof spending in any area are just right.28 Sowhile the gap between the 15 percent ofGDP on core functions and overall outlaysof nearly 40 percent may provide a roughmeasure of unproductive outlays, we can-not therefore infer that 15 percent is theright size of government. Such a determina-

tion requires a cost-benefit analysis that is,unfortunately, rarely a part of the politicalprocess.

Notes1. For example, Dimitar Chobanov and Adri-ana Mladenova find an optimum of 25 percentfor government outlays as a share of GDP, whileRichard K. Vedder and Lowell E. Gallaway get29 percent. See Dimitar Chobanov and Adri-ana Mladenova, What Is the Optimum Size ofGovernment? Institute for Market Economics,

Working Paper, 2009; and Richard K. Vedderand Lowell E. Gallaway, Government Size andEconomic Growth, Joint Economic Committee,1998. Taking a different approach, Vito Tanziconcludes that 30 percent of GDP is sufficientto finance all legitimate interventions by thestate, whereas James Gwartney et al. obtain afigure of 15 percent. See Vito Tanzi, The Eco-nomic Role of the State in the 21st Century,Cato Journal 25, no. 3 (Fall 2005); James Gwart-ney, Randall Holcombe, and Robert Lawson,The Scope of Government and the Wealth ofNations, Cato Journal18, no. 2 (1998), 16390;and Vito Tanzi and Ludger Schuknecht, PublicSpending in the 20th Century: A Global Perspective(Cambridge: Cambridge University Press, 2000).

2. There may well be countries whose govern-ments are too smallcountries reputed to be innear anarchy, such as Afghanistan or Somalia,come to mindbut even this is unclear. Datafrom such countries tend to be of poor quality ornonexistent.

3. See Robert Solow, A Contribution to the

Theory of Economic Growth, Quarterly Journalof Economics 70, no. 1 (1956), 6594. See also thediscussions in Robert Hall and Charles Jones,Why Do Some Countries Produce So MuchMore Output per Worker than Others? Quar-terly Journal of Economics 114, no. 1 (1999), 83116; and Robert Barro and Xavier Sala-i-Martin,

Convergence, Journal of Political Economy 100,no. 1 (1992), 22351.

4. See, for example, experts referenced in DavidBarboza, China Passes Japan as Second-LargestEconomy, New York Times, August 15, 2010.This would require Chinas GDP to grow nearly6 percent faster than U.S. GDP annually over a20-year periodthe same sort of nave extrapo-lation that had been applied to Japan a genera-tion earlier being applied to a country with heavygovernment control and whose recent growth ispremised on unreliable data.

5. This is not an original argument, of course, it

is just frequently overlooked in the literature onthe optimal size of government. See, for example,Hall and Jones; Jan-Egburt Sturm and Jakob deHaan, How Robust is the Relationship betweenEconomic Freedom and Economic Growth?

Applied Economics33, no. 7 (2001), 83944; andStephen Easton and Michael Walker, Income,Growth and Economic Freedom,American Eco-nomic Review 87, no. 2 (May 1997).

6. See Barro and Sala-i-Martin.

7. The results are not sensitive to the choice ofyears for either variable. The sample of countries

in this chart was designed to include relativelyadvanced economies (as opposed to those thatare in a transition phase and likely to look verydifferent in a decade or two) not heavily relianton oil exports. Natural resource extraction ingeneral, but especially oil, can present a mislead-ing picture of GDP per capita.

8. Robert Barro, Economic Growth in a CrossSection of Countries, Quarterly Journal of Eco-nomics 106, no. 2 (1991), 40743.

9. Gwartney et al. (1998).

10. Edward C. Prescott, Why Do Americans

Work So Much More Than Europeans?FederalReserve Bank of Minneapolis Quarterly Review 28,no. 1 (2004), 213.

11. See also Lars Ljungqvist and Thomas Sar-gent, Do Taxes Explain European Employment?Indivisible Labor, Human Capital, Lotteries,and Savings, NBER Macroeconomics Annual21(2006), 181246, and Prescotts response therein.

12. Juliet Schor, The Overworked American (New


10/12

10

York: Basic Books, 1993).

13. Mark Aguiar and Erik Hurst, MeasuringTrends in Leisure: The Allocation of Time overFive Decades, Quarterly Journal of Economics122no. 3 (2007), 9691006.

14. See Hall and Jones.

15. This paradox is of course an old idea goingback to Platos Republic, and is analyzed exten-sively in James Buchanan, The Limits of Liberty:

Between Anarchy and Leviathan (Chicago: Univer-sity of Chicago Press, 1977).

16. Gwartney et al. (1998) estimate that expen-ditures for core functions of governmentpro-tections of persons and property, national de-fense, education, monetary stability and physicalinfrastructure amount to no more than 15 per-cent of GDP.

17. William Easterly, The Elusive Quest for Growth(Cambridge: MIT Press, 2002).

18. Daron Acemoglu et al. delve even deeper toargue that the range of institutions we observecan be explained by colonial policies around theworld, which in turn were the result of local con-ditions, notably mortality rates that affected theabilities of Europeans to settle. See Daron Ac-emoglu, Simon Johnson, and James Robinson,The Colonial Origins of Comparative Devel-opment: An Empirical Investigation, American

Economic Review 91, no. 5 (2001), 13691401.

19. See Barro for the measure of spending thatexcludes transfer payments, defense, and educa-tion.

20. The idea that income inequality causescrime is often cited as a justification for redis-tribution, but evidence for this proposition ismixed at best.

21. See Hall and Jones. Corruption is stronglycorrelated to less economic freedom, and is a

broad measure of the size of government. See James Gwartney et al., Economic Freedom of thWorld: 2009 Annual Report (Vancouver, BC: Fraser Institute, 2009), p. 22.

22. Ibid. Data retrieved from http://freetheworld.com.

23. In 1995 there were only three: Hong KongSingapore, and New Zealand. In 2008, Switzer-land and Chile also had more economic freedomthan the United States. See James Gwartney

Joshua Hall, and Robert Lawson,Economic Free-dom of the World: 2010 Annual Report(VancouverFraser Institute, 2010).

24. Thus Chobanov and Mladenovas claim thata government outlay ratio of 25 percent of GDPmaximizes growth is questionable at best, sincethe smallest ratio in their sample exceeds 30 per-cent.

25. Others have found similarly. For exampleSturm and de Haan also find that the changein economic freedom is strongly related to eco-nomic growth. However, the level of economicfreedom is not related to growth. StephenEaston and Michael Walker find a significantpositive effect of economic freedom on the long-run level of per capita income.

26. Gwartney et al. (1998).

27. Milton Friedman made a similar point inan interview with Charlie Rose on December 262005. He said that as governments have ventured

into areas they have no real business being in,they end up doing a worse job with the activitiesthey should be undertaking.

28. It is noteworthy that as the redistributive activities of the U.S. government have grown, theshare of expenditures on national defense hasdiminishedfrom nearly 10 percent in 1960 toaround 5 percent today. It would be a remark-able coincidence if these expenditure levels weresomehow both optimal.


11/12

STUDIES IN THE CATO INSTITUTE

DEVELOPMENT POLICY BRIEFING PAPER SERIES

6. Why Is Africa Poor? by Greg Mills (December 6, 2010)

5. Freedom and Exchange in Communist Cuba by Yoani Snchez (June 16, 2010)

4. Socialism Kills: The Human Cost of Delayed Economic Reform in India bySwaminathan S. Anklesaria Aiyar (October 21, 2009)

3. Troubling Signs for South African Democracy under the ANC by Marian L. Tupy(April 25, 2007)

2. Kenyas Fight against Corruption: An Uneven Path to Political AccountabilitybyJohn Githongo (March 15, 2007)

1. A Second Look at Microfinance: The Sequence of Growth and Credit in EconomicHistoryby Thomas Dichter (February 15, 2007)

OTHER PUBLICATIONS ON DEVELOPMENT FROM THE CATO INSTITUTE

James Gwartney, Joshua Hall, and Robert Lawson, Economic Freedom of the World: 2010 AnnualReport(Vancouver: Fraser Institute and Cato Institute, 2010).

James Tooley, The Beautiful Tree (Washington: Cato Institute, 2009).

Jean-Pierre Chauffour, The Power of Freedom: Uniting Human Rights and Development (Washington:Cato Institute, 2009).

James A. Dorn and Barun S. Mitra, eds.,Peter Bauer and the Economics of Prosperity (New Delhi:Liberty Institute and Cato Institute, 2009).

Bibek Debroy, Laveesh Bhandari, and Swaminathan Aiyar,Economic Freedom of the States of India2011 (New Delhi: Friedrich Naumann Foundation, Indicus Analytics, and Cato Institute, 2011).

Gregory Elacqua, Humberto Santos, Dante Contreras, and Felipe Salazar, Private SchoolChains in Chile: Do Better Schools Scale Up? Cato Institute Policy Analysis no. 682, August 16,2011.

Takis Michas, Putting Politics Above above Markets: Historical Background to the Greek Debt

Crisis, Cato Institute Working Paper no. 5, August 15, 2011.

Sallie James, Time to X Out the Ex-Im Bank, Cato Institute Trade Policy Analysis no. 47, July6, 2011.

Juan Carlos Hidalgo and Daniel Griswold, Trade Agreement Would Promote U.S. Exports andColombian Civil Society, Free Trade Bulletin no. 44, February 15, 2011.


12/12

Anne ApplebAumWAshington post

gurchArAn DAsFormerceo, procter& gAmble, inDiA

ArnolD hArbergeruniversityoF cAliForniAAt los Angeles

FreD hugolDmAn sAchs, AsiA

peDro-pAblo KuczynsKiFormerprime ministeroF peru

DeepAK lAluniversityoF cAliForniAAt los Angeles

JospierAFormerministeroF lAborAnDsociAl security, chile

The Center for Global Liberty and Prosperity was established to promote

a better understanding around the world of the benefits of market-liberal solutions to some of the most pressing problems faced by develop

ing nations. In particular, the center seeks to advance policies that protect hu

man rights, extend the range of personal choice, and support the central role of

economic freedom in ending poverty. Scholars in the center address a range o

economic development issues, including economic growth, international finan

cial crises, the informal economy, policy reform, the effectiveness of official aid

agencies, public pension privatization, the transition from socialism to the mar

ket, and globalization.

For more information on the Center for Global Liberty and Prosperity,visit www.cato.org/economicliberty/.

Nothing in this Development Briefing Paper should be construed as necessarily reflecting thviews of the Center for Global Liberty and Prosperity or the Cato Institute or as an attempt toaid or hinder the passage of any bill before Congress. Contact the Cato Institute for reprinpermission. Additional copies of Development Briefing Papers are $6 each ($3 for five omore). To order, call toll free (800) 767-1241 or write to the Cato Institute, 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001; phone (202) 842-0200; or fax (202) 842-3490

All policy studies can be viewed online at www.cato.org.

OtherStudieSOndevelOpmentfrOmtheCatO inStitute

BOardOfadviSerS

The Elephant That Became a Tiger: 20 Years of Economic Reform in India by Swaminathan S.Anklesaria Aiyar, Development Policy Analysis no. 13 (July 20, 2011)

Why Is Africa Poor? by Greg Mills, Development Policy Briefing Paper no. 6 (December 6, 2010)

Freedom and Exchange in Communist Cuba by Yoani Snchez, Development Policy Briefing

Paper no. 5 (June 16, 2010)

The State of Liberal Democracy in Africa: Resurgence or Retreat? by Tony Leon, DevelopmentPolicy Analysis no. 12 (April 26, 2010)

Reflections on Communism Twenty Years after the Fall of the Berlin Wall by Paul Hollander,Development Policy Analysis no. 11 (November 2, 2009)

Socialism Kills: The Human Cost of Delayed Economic Reform in India by Swaminathan S.Anklesaria Aiyar, Development Policy Briefing Paper no. 4 (October 21, 2009)

An International Monetary Fund Currency to Rival the Dollar? Why Special Drawing RightsCant Play That Role by Swaminathan S. Anklesaria Aiyar, Development Policy Analysis no. 10

(July 7, 2009)

The False Promise of Gleneagles: Misguided Priorities at the Heart of the New Push forAfrican Development by Marian L. Tupy, Development Policy Analysis no. 9 (April 24, 2009)

Can We Determine the Optimal Size of Government?, Cato Development Briefing Paper No. 7

Documents