The Becker-Posner Blog

Why Are Tax Burdens So Different in Different Developed Countries?--Posner

There are striking differences in tax burdens across nations, as explained in a recent report by the Organisation for Economic Co-Operation and Development. Measuring the tax burden in 2006 as the percentage of gross domestic product that is collected in taxes, the report arrays 20 countries from top to bottom. At the top is Sweden, with a tax burden of 50.1 percent; at the bottom is South Korea, with a tax burden of 26.8 percent. The United States is near the bottom, with 28.2 percent, and between it and South Korea are Greece and Japan, each with 27.4 percent. Next below Sweden is Denmark, with 49 percent, France,with 44.5 percent, and Norway, with 43.6 percent. The middle range is illustrated by Britain with 37.4 percent, Spain with 36.7 percent, and Germany with 35.7 percent. In all 20 countries except the Netherlands, the tax burden has increased since 1975, though in some countries, such as the United States, the increase has been slight--only 2.6 percent. In others, however--Denmark Greece, Italy, Portugal, South Korea, Spain, and Turkey--it has exceeded 10 percent. Spain's increase has been the greatest, at 18.3 percent, followed by Italy's at 17.3 percent and Turkey's at 16.5 percent. The OECD report explains that the increase in tax burden is due to increased revenues from "direct" taxes--income (including payroll) and corporate taxes--rather than from "indirect" taxes such as VAT, sales taxes, and other excise taxes. Even though most countries, including the United States, have cut income and corporate tax rates, the cuts have been more than offset by increases in income and corporate profits; of course the cuts may have helped generate those increases. The OECD favors indirect taxes because they tax only consumption, whereas direct taxes tax income that is saved, and thus discourage investment. The increase in the tax base for direct taxes explains the mechanism by which the tax burden has grown but not why it has grown--why in other words the demand for government spending has grown. The OECD speculates that the cause is increased demand for social services such as pensions and health care. The curious thing about the OECD data is that prosperity, economic growth, and other measures of economic well-being do not seem closely correlated with the tax burden. The variance across countries in tax burden is very great, yet one finds troubled economies, such as those of Japan and Greece, near the bottom of the tax-burden distribution--of course Japan is a very wealthy country, as Greece is not, but Japan's economic performance has been disappointing in recent decades. And one finds some high-performing economies, such as those of Sweden, Norway, and Finland at the top of the distribution, or (as in the case of the Netherlands, Spain, and the United Kingdom) in the middle. However, there is some negative correlation between economic performance and the tax burden; for Ireland, Switzerland, and the United States are low on the distribution, while typically low-performing Western European countries cluster in the upper half. One would think that the tax burden, especially but not only when it is created mainly by direct taxes, would have a strong negative effect on economic well-being. (Perhaps it does, when other factors affecting economic well-being are adjusted for, which I have not attempted to do.) If government is less efficient than private enterprise, the more economic activity that is performed by government rather than by the private sector the less productive the economy as a whole should be; and the higher the tax burden, the greater the amount of economic activity performed by government. To the extent, moreover, that variance in tax burden across countries reflects variance in marginal rates of taxing income and corporate profits, we would expect the high tax-burden countries to be less productive, because the higher the tax on income, the greater the incentive to substitute leisure (which is untaxed) for work and to expend resources (and create economic distortions) in an effort to reduce the tax bite. But there is an important difference between the actual production of economic goods and services by government, on the one hand, and transfer payments on the other. The effect of taxes on the behavior of the taxed entity is the same, but the effect on the efficiency of production is different. In the United Kingdom (which nevertheless has a high-performing economy), the government produces medical services; the National Health Service is the employer of the vast majority of doctors and other health professionals and owns most of the hospitals and other health care facilities in the U.K.; only about 8 percent of the U.K.’s population is served by private health providers. In contrast, the U.S. Medicare and Medicaid programs transfer vast amounts of public money to health care providers, but the providers are mostly private. The transfers come with strings attached, of course, and some of those strings induce inefficient behavior by the recipients. Nevertheless, a U.S. National Health Service on the English model would undoubtedly be highly inefficient compared to our admittedly highly imperfect private provision of health care. Transfer socialism is not as inefficient as means-of-production socialism. To the extent that the growth in government spending is a growth in transfers rather than in government ownership of producers, the impact on economic growth and prosperity may be small, especially since the growth in transfers has coincided with the deregulation movement, which has resulted in privatization of significant areas of traditional public ownership, less regulation of the economy, and, as I mentioned, lower direct-tax rates. There thus appears to be a kind of balance, in which the efficiency-reducing effects of greater government spending are contained by reductions in direct-tax rates, by increased privatization and deregulation, and by channeling increased tax revenues mainly into transfer programs rather than into government production of goods and services.

Tax Burdens-Becker

The burden of taxes to a country depends not only on the fraction of its gross domestic product GDP that are collected as tax revenue –the data shown in Posner's chart- but on many other factors as well. Since my comment is brief I will confine my discussion to the link between tax burdens, the level of government spending, and the structure and incidence of taxes. It is not possible to separate tax burdens from government spending. Obviously, as Posner makes clear, how governments spend their tax revenues makes an enormous difference to the functioning of an economy. In addition, however, the level of government spending also affects the tax burden. If spending exceeds the amount collected in taxes, the excess spending must be financed by an increase in government debt (I ignore inflationary printing of money). Interest payments on the higher government debt have to be financed by higher taxes in the future, so the full tax burden is determined not by tax revenues alone but also by government spending. Senator McCain has justified his initial opposition to the Bush tax cuts by indicating that they were not combined with cuts in government spending -in fact, just the opposite occurred. The tax burden depends in addition on the type of taxes used and their structure. What economists call the "excess burden" is measured by the difference between the cost to those paying taxes and the revenue collected by government. The excess burden is zero for a head tax, which is an equal tax per person, since the amounts paid to governments from such a tax equals the cost to taxpayers. Taxes on income do have an excess burden because they distort taxpayers' decisions toward greater leisure. The higher the marginal tax rate, the greater are these and other distortions induced in labor supply, and hence the greater the excess burden of income taxes. To reduce distortions, broader and flatter taxes are better because then marginal tax rates are lower. Rudy Giuliani has proposed a flat and rather broad income tax with a highest marginal tax rate of only 30 percent to complement the present complicated income tax system. Consumption taxes, such as value added taxes, have lower excess burdens than income taxes. Like an income tax, a general tax on consumption does discourage work in favor of leisure essentially because individuals can avoid both consumption and income taxes by taking additional leisure since leisure is not taxed. However, an income tax has other distortions as well since income is both taxed when received, and also taxed again when the savings out of income produces additional income. Income taxes in effect tax savings twice, while consumption taxes only tax savings once, when they are spent. In order to reduce this double taxation of savings from income taxes, the US and other countries allow families to save in ways that are free of income taxes until the savings are spent, such as through saving with IRAs. There is a natural tendency to assume that the burden of taxes falls on persons or companies that mail the tax checks into the government. To show why this is generally false, consider a 10 percent tax on capital that initially reduces returns on capital from say 8 percent to 7.2 percent. This initial impact is clearly on owners of this capital, who are generally wealthier than the average individual. Over time, however, the capital stock would fall because companies reduce their investments in reaction to the lowering of after-tax returns on investments due to the capital tax. As the capital stock falls, the after-tax return would begin to increase because the productivity of capital is higher when capital is scarcer relative to labor. The capital stock would continue to fall essentially until after-tax returns climb back up to the 8 percent level they were at before the tax on capital was imposed. Since studies confirm that in the long run owners of capital get about the same rate of return that they would have without any taxes on capital, who then pays the capital tax in the long run? The answer is not capital but labor because wages and earnings are lower when workers have less capital to work with. Owners of capital continue to send in the checks to pay a capital tax, but the negative response of investments to a capital tax shifts the burden of a capital tax away from capital to labor. That eventually labor pays a tax on capital even though it is placed on capital explains why economists generally oppose long-term taxes on capital even though in the short run capital taxes have many desirable properties. Investment tax credits, accelerated depreciation, and low taxes on capital gains are some of the ways that the effective long run tax on capital is reduced toward zero.

Slowing of Black Progress-Becker

Black Americans have made considerable progress during the past two decades in reaching top positions in government, business, and the military. Colin Powell was Chairman of the Joint Chiefs of Staff of the Armed Forces and also Secretary of State, the third highest position in government, and has been followed at State by Condoleezza Rice. Richard Parsons is CEO of Time Warner, Kenneth Chenault is CEO of American Express, Stan O'Neal until recently was CEO of Merrill Lynch, and black men and women are heads too of other major American corporations. Barack Obama is making a credible run at becoming the presidential nominee of the Democratic Party, and he is appealing not only to black and other minority voters, but also to a wide cross section of independent voters. Obama's successful pursuit of the American presidency has appeared to provide closure to "The American Dilemma": this country's unsatisfactory relations with African American descendants of slaves, even though Obama himself has a white mother and an African father. Yet the political, governmental, and business success during the past few years of small numbers of blacks does not accurately measure the progress of typical African-American men and women. Education, earnings, and health gaps between whites and blacks did significantly narrow during the 40-year period from the end of World War II until the late 1980's. However, from then until the present, progress of blacks relative to whites has essentially stopped, leaving a still sizable distance between the circumstances of whites and blacks. My colleague at the University of Chicago, Derek Neal, has documented many aspects of this slowdown in black progress (see his "Why has Black-White Skill Convergence Stopped?" in Handbook of the Economics of Education, Vol 1, 2006). He shows that the racial gap in average years of schooling for men in their late twenties was about 2 1/4 years in 1965, declined to less than a year in the 1980's, and basically remained at that level into this century. The schooling gap between young black and white women has been smaller than that for men, it also fell a lot until the mid-1980's, but if anything the gap has increased since then to become similar to the gap for young men. Related trends of considerable progress and then stagnation are found in racial gaps for high school and college graduation rates, and in teenager reading and math tests scores. Earnings of blacks and whites with the same years of schooling show similar patterns: convergence until the late 1980's, and mainly stable since then, although there are increased racial gaps in some education groups. Differences in life expectancy and general health between blacks and whites followed similar trends. Black life expectancy narrowed until about twenty years ago, and then remained constant at about 6 years below that of whites. This means that life expectancy of blacks is comparable to mortality rates in the much poorer countries of Paraguay and Mexico. A large fraction of the racial difference in average length of life is due to differences in the incidence of cardiovascular diseases, life styles, and violent deaths. The economic literature on the statistical value of life shows that a typical white values a life year at over $120,000. A comparable analysis for typical American blacks would suggest that they value an additional life year at less than the figure for whites because black earnings are lower. If the value by blacks is about $100,000, the discounted value of the six-year lower life expectancy for blacks would be worth about $400,000. Since the discounted value over a lifetime of the racial gap in average annual earnings would be worth no more than $300,000, this calculation suggests that the racial gap in life expectancy is as important as the still considerable earnings shortfall of blacks. Why did the progress of blacks stop well short of achieving full equality with whites, and is the slowdown during the past 20 years in black progress only temporary, or is it an indication of what the racial situation will be during the next few decades? The sharp slowdown is surprising mainly because institutionalized and personal discrimination against African-Americans has continued to fall into this century. Probably the most important offset to the decline in discrimination is the rapid growth since the 1960's in the fraction of black children raised in households with only one or no parents-these households also grew among whites, but at a much slower pace. Moreover, white single parent households mainly arise from a divorce between parents who had their children while married (or while living together), whereas never-married and quite young mothers raise many black children. In addition, there is social pressure on young blacks growing up in segregated neighborhoods to engage in crime, including selling drugs, and to not "act white", where acting white sometime is taken to mean studying hard and investing in one's human capital. These pressures act more on black boys rather than girls, which help explain why the achievements of black women are much closer to those of white women than is the gap between black and white men. It is too early to tell whether these and other forces that have prevented blacks from achieving full parity with whites are temporary or more long lasting. A disturbing fact is that growing up in families that invest less in their children casts a long shadow since children brought up in these families tend also to invest less in their children. This process gets to be repeated to some degree over subsequent generations. Yet it may be possible to overcome to a considerable degree this intergenerational transmission of low status. The most promising approaches in my opinion involve self-help programs that encourage better choices in black communities, the legalization of drugs, personalized medicine that recognizes differences in vulnerabilities to disease between blacks and whites, head start type school programs, and school vouchers and charter schools that widen school choice and stimulate education innovations, On the whole, I am optimistic that some of these changes will be made, and hence that the convergence between blacks and whites will resume after the hiatus during the past 20 years, although it will probably be many decades before blacks achieve anything close to full parity with whites.

The Black-White Income Differential--Posner's Comment

According to Census statistics, average black family income was only 51 percent of average white family income in 1947, but rose to 56 percent in 1964, the year that the federal law forbidding employment discrimination (Title VII of the Civil Rights Act of 1964) was enacted. The gap continued to narrow for some years, as Becker notes; by 1990, average black family income had risen to 63 percent of average white family income. But it has not risen (it may have fallen slightly) since then. Notice that the annual rate at which the gap shrank was much faster before the modern civil rights era ushered in by Title VII--6 percent in 17 years, versus 7 percent in the next 36 years. It is nowhere written that all ethnic groups shall have the same average income, but the white population is itself an amalgam of ethnic groups; and it is not as if blacks are newcomers to America, who would be expected to lag the average income of the settled population. Quite apart from the black "stars" whom Becker mentions, there is a large and thriving black middle- and upper-middle class. The income gap, and the related gaps in longevity, law-abidingness, education, and family stability, are due to the disproportionate incidence of social disorder among blacks, creating a large black "underclass" that drags black average-income statistics down. There was very little civil rights law before Title VII; nevertheless the black-white income differential narrowed more rapidly in that benighted era than it has since. It is possible that antidiscrimination laws do not benefit their intended beneficiaries, because they give the beneficiaries a sense of entitlement and victimhood, foster tokenism, increase employers' costs, cast a shadow over the real achievements of outstanding members of the "benefited" group, create an unhealthy preoccupation with racial and ethnic identity, and cause white backlash. It is also possible that the sexual revolution of the 1960s promoted the break-up of the black family--of the white too, but the whites were in a better position to adapt. To the extent that the "Great Society" programs of the 1960s and the social disorder of the same period are correlated phenomena, together constituting a lurch to the Left, the net effect on black progress may have been negative. Probably the focus of reform should not be on the black-white income gap as such but on the social pathologies that are responsible (at least in part) for it. The best approach might simply be to remove obstacles to labor mobility and to competition more generally; Becker mentions school vouchers and charter schools. In addition, reducing or eliminating the minimum wage would expand employment opportunities for blacks. Measures can also be taken to reduce the out-of-wedlock birth rate of blacks; in this regard the Administration's effort to stress abstinence, rather than contraception, as a means of limiting teenage pregnancy is misguided. But there seems to be little political pressure for such reforms. The costs of the social disorders that afflict poor blacks are incurred mainly by poor blacks themselves, and poor blacks do not vote very much. Moreover, blacks support the Democratic Party so overwhelmingly that Democrat politicians have little incentive to expend their necessarily limited political capital on policies that might benefit blacks at the expense of groups that are in play between the two parties, such as public school teachers. A step in the right direction might be to allow (as many states already do) felons who have completed their sentence to vote.

The Candidates' Health Care Reform Plans--Posner

Virtually all the presidential candidates have proposed plans for reforming health care in the United States. All the plans would require federal legislation, although many include measures that the executive branch of the federal government could implement without new legislation. To evaluate proposed solutions, one must know what the problem is. Different candidates perceive the problem differently, but there is general agreement that health care in the United States costs too much--it accounts for more than 16 percent of GNP, compared to less than 11 percent in France, which the World Health Organization ranks first in the world for the quality of its health system; the WHO ranks the United States 37th. Now that is one of those multi-factor rankings that can be criticized for arbitrariness. However, if one confines one's attention to just one of the criteria, "disability-adjusted life expectancy," the United States still does not do very well. It ranks 24. (France is 3; Japan is 1.) There also is general agreement that too many people in the United States lack health insurance, whether public or private, and that this is either an economic problem or an ethical problem, or both. More than 45 million persons under the age of 65 lack insurance (few older persons do, because of Medicare, though Medicare coverage is incomplete and elderly people who can afford to buy medi-gap insurance usually do so), about 90 percent of whom are citizens or lawful residents. The uninsured are disproportionately poor and lower-middle-class (and therefore disproportionately black and Hispanic), though many poor children are covered by Medicaid or by SCHIP (State Children's Health Insurance Program). Contrary to popular impression, Medicaid is intended primarily for poor families with children; it does not cover the poor as such. Also, Medicaid reimbursement to health-care providers is chintzy, unlike Medicare reimbursement, and the quality of service is as a result poor. Most (70 percent) of the uninsured are in families with at least one full-time worker. Most are young: The age breakdown is children: 20 percent; ages 19–44, 56 percent; 44–64, 23 percent. The health of the uninsured is on average significantly worse than that of insured persons of the same age. As one would expect, the uninsured consume less health care than the insured--only about $1,000, on average, a year, though this is partly because elderly persons, who consume the most health care on average, are covered by Medicare, and more broadly because of the relative youth of the uninsured. The care they do not pay for--the uncompensated care--is provided to them as charity, for example by hospital emergency rooms, which swallow much of the cost, though some is reimbursed by various government programs. In part because they consume less health care, in particular less emergency health care, the uninsured have as I have mentioned poorer health and greater mortality than the insured, though I do not know how large a part; low income, and the style of living that goes with low income, may explain more of the difference in health and longevity between the insured and the uninsured than the lesser demand for health care by the uninsured. A further complication is that since premiums for employees' health insurance plans are deductible from corporate income tax and heavy medical expenses are deductible from individual income tax, the health care of group-insured persons (and most health insurance is employee group health insurance), and of persons with high incomes (and therefore high deductibles from income tax), is subsidized. The goals of reducing the costs of health care (at least without reducing quality or producing political outrage) and increasing health-insurance coverage are in conflict, but the candidates' plans strive somehow to achieve both goals. Some of the proposals for reducing aggregate costs are either fluff, like reining in jury awards in medical malpractice cases (those awards are a tiny fraction of total health costs, and already are being reined in by judges and by tort-reform measures adopted by state legislatures), or measures that the market is in process of implementing, such as the digitization of medical records. Other economizing proposals have hidden negative implications for quality--such as placing price controls on prescription drugs, reducing the protection that the patent laws provide against competition by generic (nonpatented) substitutes, and permitting the reimportation of drugs from countries that have price controls on drugs. Reducing property rights in medical innovations is likely to reduce the rate of those innovations and hence, in the long run, health and longevity, and those costs have to be traded off against benefits in lower prices for existing drugs. Some measures defended as economizing because they would simplify the administration of health insurance would generate offsetting costs, such as forbidding "discrimination" against persons with preexisting health conditions. Which brings me to the essential point in evaluating the candidates' health care reform proposals: significantly expanding health insurance coverage is bound to be very costly, whether the role of government in bringing about the expansion of coverage is large, as in the case of the Democratic candidates' proposals, or small, in the case of the Republicans' proposals, which generally are limited to increasing the tax subsidies for the purchase of private health insurance. Although some of the uninsured are healthy risk takers, most would have difficulty affording health insurance, and, as a practical matter, would require a subsidy of some sort. The subsidy itself would just be a transfer, financed presumably by a tax increase; the social cost (that is, the consumption of scarce resources by the program) would be the cost of administering the subsidy program and the misallocative effects that a tax increase would create. The larger social cost would be the additional health care resulting from the expansion of coverage. Insured people use more medical care because the possession of insurance lowers the marginal cost of that care to them. And because the uninsured are on average less rather than more healthy than the insured, forcing them to buy insurance would not lower insurance rates to others. The average annual cost of employee group health insurance for a family of four is $12,000. Supposing there are 10 million families without health insurance, and that two-thirds could not afford such insurance, it might well cost more than $80 billion a year to buy it for them. This would be more than 3 percent of the federal budget. That is not an unthinkable amount, but the political opposition would be great, because the majority of the population--the people who have public or private health insurance already--would not benefit from it. Might there be a compensating offset because with greater medical care the people who now are uninsured would be healthier and live longer, and thus cost less in subsidized medical care in the long run? Not necessarily, since the longer a person lives, the greater his average medical expenses because average annual such expenses grow with age. Living a healthier and longer life is of course a benefit to a person; my point is only that it need not reduce his average annual health costs. The way to economize on expenditures on health care, though it is utterly infeasible politically, would be to eliminate the tax subsidies for health insurance and health care and institute a means test for Medicare, and at the same time to limit medical services. Then both the demand for and the supply of those services would be reduced, and the percentage of GNP that goes for health care would drop. But the principal result might be to reallocate consumption spending to goods and services that most people value less at the margin than they do health care. Moreover, there is an economic argument for some level of tax subsidies for health insurance premiums or health care. Medical care increases human capital, and is thus an investment, and investment expenditures need not be (probably should not be) taxed as long as the revenues generated by them are. Medical treatment that extends life or enables a person to work increases the person's income, which is taxable. Maybe a little patchwork here and there is the most that is both economically desirable and politically feasible by way of reform of American health care.

Potential Reforms of American Health Care-Becker

As Posner indicates, American health care generally gets poor grades in international comparisons of health care systems. Although major reforms are needed in the American approach, international comparisons underrate American health care. This is partly because these comparisons give insufficient weight to the fact that most of the new drugs to treat major diseases originated in the US, along with many of the new surgical procedures, and insights about the importance of lifestyles in good health. This helps explain why many Canadians and those from other countries come to the US to treat serious diseases rather than visa versa. The US is also much more generous than other countries, such as Great Britain and France, in making expensive surgeries and drugs available to older persons through Medicare and private insurance. This too significantly raises the cost of health care. Moreover, the American health system is decentralized and "messy", and many health evaluators prefer a single payer (i.e., government) centralized approach to health care as opposed to any market-based approach. This is not to deny that the American health care system has serious defects. If I were running for president, and allowed only four reforms, I would emphasize the following (assuming I do not worry about getting enough votes to be elected!): 1) Eliminate the link between employment and the tax advantage of private health insurance. Since much of the spending on health are investments in human capital, there is good reason to exempt these expenditures, along with other investments, from income taxes. However, this employment link is inequitable because it does not provide the same tax advantages to families without employment-based insurance. It also encourages expensive employer health plans that have significant consumption components since the government picks up much of the cost of such coverage. President Bush has proposed a reasonable alternative; give every family a flat $15,000 standard deduction (and half that amount for individuals), whether or not their health insurance is obtained through their employer. They would still get this deduction if they spend less on their insurance, so they have incentives to economize on their health care (but by my reform number 4, everyone would have to take out catastrophic coverage). Consumers would have to pay for any coverage in excess of $15,000, so they would only choose such coverage if they were willing to spend their own money, not taxpayers. 2) Encourage the spread of Health Savings Accounts (see my discussion on Feb. 5, 2006) that encourage consumers to economize on unnecessary medical expenditures. Present law allows tax-free contributions to these Accounts of up to about $2700 for individuals and double that amount to $5450 for families, as long as these contributions are not greater than the deductibles on their health insurance. Contributions to HSAs that are not spent in any year can be carried over to future years without any tax liabilities, and even into retirement income. So HSAs are an efficient way to save as well as to spend on non-catastrophic medical care. Health Savings Accounts have spread since they were introduced several years ago, but might need greater encouragement, such as higher limits. 3) Medicare spending amounts to about $350 billion a year, it constitutes about 12 percent of federal spending, and it is one of the most rapidly growing entitlements. It is projected to continue to grow as a fraction of GDP from its present 2.7 percent level to over 11 percent in 2080. The source of the growth is the continued aging of the population, and the increased per capita medical spending on older person as new medical technologies and drugs are developed. Projections made by Medicare Actuaries indicate that the Medicare HI Trust Fund will be exhausted by the year 2018-only a decade away. Reform of Medicare is probably among the most challenging not only because of the elderly's political clout, but also because Americans have come to expect access to expensive medical treatments as they age. Still, the prescription drug coverage introduced into Medicare in 2003 was an important step in the right direction, despite the flaws in the program (see my discussion on February 3, 2005). Drugs are not only increasingly available to fight many diseases of old age, but drugs, once developed, are relatively cheap to extend to large numbers of users. Even when drugs provide only small benefits as they are extended to groups that can benefit less from the drugs, the costs are far less than would be required to provide expensive surgeries or hospitalizations to older persons with few years of life remaining. This is why I would greatly increase the generosity of Medicare drug coverage, and compensate for the additional expense by cutting down on allowances for lengthy hospital stays, and raising other co-pays. 4) I do not believe the problem of the uninsured in the US is as serious as usually claimed since most of those without health insurance are young and do not have major medical expenses. When they do, they can use emergency room service at major hospitals, although studies show that they do not even use emergency room care more often than others. Still, it may be desirable to require that everyone must contract for private catastrophic health care since the uninsured tend to use taxpayer and philanthropic funded medical care facilities to pay for the costs of any major illnesses. Medicaid should be extended to cover anyone who cannot afford such catastrophic insurance. Compulsory coverage would integrate the 45 million or so uninsured Americans into an overall health care system while still preserving the desirable decentralized private system of health care.

Does Economic Development Reduce Terrorism? Becker

Face-to-face interviews of an apparently random sample of the Pakistani population were conducted in August 2007 for Terror Free Tomorrow, a non-partisan Washington policy organization (www. TerrorFreeTomorrow.org). Those interviewed were asked questions about Al Qaeda and other issues facing Pakistan. The results indicate that more than a third of Pakistanis have a favorable view of Al Qaeda, the Taliban, and bin Laden, and that President Musharraf is the least popular political leader in Pakistan. Respondents also have a decidedly unfavorable view of the US-led war on terror, for they believe that its real purpose is to kill Muslims, break Muslim countries, and achieve other related goals. There are many causes of such attitudes, but I want to explore the effects of economic development on the degree of support for terrorism. Many surveys of populations in poor nations give a distorted picture about attitudes in these countries toward controversial issues because they are confined to urban areas that are safer and more easily accessible, and where inhabitants tend to be more educated and better off economically. By contrast, this survey of Pakistani opinions seems to be a reasonably representative sample of about 1,000 Pakistanis age 18 or older in urban and rural areas in all four provinces of Pakistan. The vast majority of these respondents are married Sunni Muslims who live in towns and villages, and have 10 or less years of schooling. A little less than half are women. Unfortunately, the results so far published from this survey do not separate answers by years of schooling, income, urban-rural location, gender, or other useful personal characteristics. Pakistan is a very poor nation that is low on international rankings of both per capita income and the extent of economic and political freedoms. According to the World Development Report of the World Bank, Pakistan's purchasing-power-adjusted real per capita income is considerably below India's, and is less than one half of China's. Evidence from changes in other countries that have developed indicates that if Pakistan experienced a prolonged period of rapid economic growth, behavior and attitudes on many issues would change radically, regardless of the fact that it is a Muslim nation in Asia. Consider what happens to the family in response to economic development. The family organization and structure that are the foundation of traditional societies evolved over hundreds, indeed thousands, of years. Families are by no means the same in different cultures, but in all poorer nations, birth rates are high, and the extended family is usually close. Yet regardless of culture, birth rates greatly decline, and extended families evolve into much greater reliance on the nuclear family, in every country that has experienced sizable economic development. Examples of sharp declines in family size include the Chinese cultures of Taiwan, Hong Kong, and China (although birth rates in China were partially forced down by government pressures on families to have only one child). Big declines in fertility also occurred in India, Turkey, and Malaysia ( Malaysian birth rates are still relatively high), Malaysia and Turkey being the main Muslim countries that experienced sizable economic development without having large resources of oil or natural gas. What happened in Malaysia suggests that poor Muslim countries, like Pakistan, or Morocco, or Egypt, would also have rapid falls in birth rates if they managed to have serious economic development. The power of economic development is also shown by the well-established finding that countries become more democratic when their economies undergo significant development. This finding is illustrated by Taiwan, South Korea, and Chile, all countries that started growing rapidly under non-democratic governments, and evolved into vibrant democracies. China has had significant expansion of civil and economic freedoms since it started developing rapidly in 1980 (about the time when I first visited there, and I was impressed by how restrictive conditions were). I believe China will open up further, and will attain greater political freedom if it continues to grow rapidly. Similar changes toward greater economic, political, and social freedom will take place in Pakistan, Egypt, and other Muslim countries if they too take off economically. Terrorist groups rely on populations that are sympathetic to their cause to hide and protect their members. They also recruit disaffected youth in significant numbers who are willing to commit suicide to destroy enemies. Just as economic progress greatly affects family structure and the amount of freedom available, it also sharply reduces the willingness of people to hide or otherwise protect terrorists because they have more to lose if they are caught. Although leaders of terrorist organizations usually come from more educated classes, these organizations rely on numerous foot soldiers to do a lot of the dirty work. They are generally recruited from younger and less educated groups. It becomes much harder to recruit many of these soldiers when good jobs are available, especially if these recruits are asked to commit suicide. To be sure, Al Qaeda and other radical violent groups have attracted members from the richest nations: Great Britain, France, Germany, and even the United States. Certainly in the US and Great Britain, Muslims have been rather well integrated into their economies, and both countries provide very good opportunities for advancement to younger Muslims. For this reason, in both countries, and even in France and Germany, only tiny numbers of their Muslim populations have been recruited to active participation in radical causes. If better opportunities reduce the attractiveness of suicidal terrorism, how does one explain that all the participants in the 9/11/01 suicide attacks were college-educated Muslims, and generally they were in their late twenties? Posner and I show in a paper on suicide why educated terrorists with good economic opportunities would be unwilling to engage in run of the mill terrorism or ordinary suicide attacks because the cost to them would be too great. Such types can only be attracted to terrorist organizations by influential leadership roles, or by dramatic and exceptional missions, as the 9/11 terrorist mission. That is why the education-age backgrounds of the 9/11 terrorists are the exceptions, not the rule, for the profiles of suicide terrorists. A strong counterexample comes from the backgrounds of suicide bombers during the first Intifada against Israel: they were mainly young and unmarried (the mean age of male bombers was 20), and few had a college education.

Terrorism and Economic Development--Posner's Comment

I agree with Becker's analysis as far as it goes, but I question whether the amount of terrorism is highly sensitive to economic development, to which the "demographic transition"--the well-documented tendency of birth rates (also death rates) to decline sharply when a nation reaches a threshold level of economic development--contributes. When birth and death rates decline, the average age of the population rises, which is a stabilizing force, the number of young men declines, and the economic opportunities of the young are greater because there are fewer young. So the number of potential foot soldiers for terrorism is diminished, as it is by anything that raises the opportunity costs of prospective terrorist recruits. But how important are those opportunity costs to the amount of terrorism? It is helpful to think of terrorism as of other goods and services in demand and supply terms. There is a demand for terrorism, and a supply of terrorism, and the intersection of demand and supply gives the amount of terrorism. Terrorism is a political phenomenon, and the demand is driven mainly by political grievances, real or imagined. Often the grievances are related to foreign occupation. France in Algeria; the British in Palestine; now the Israelis in the West Bank; the United States in Iraq (and earlier in the Philippines)--though in the case of Islamic terrorism, the major factor seems to be the Western "presence" in the Middle East, rather than foreign occupation; even Israel's occupation of the West Bank seems a subsidiary factor. And the Baader-Meinhoff gang in West Germany, the Red Brigades in Italy, and Aun Shirikyo in Japan are examples of terrorist groups unrelated to foreign occupation. But it is the existence of grievance that is key, and often--probably typically--the grievance is political rather than economic. If demand for terrorism is grievance-driven, then one can expect the supply of terrorists to come mainly from the intelligentsia, for the members of the intelligentsia are more likely than ordinary people to be moved by ideas, resentments, and political ambitions rather than by material concerns. They have the leisure and the education to think big thoughts, like overthrowing a government, which rarely brings material improvements. Nor is it the case that the intelligentsia supplies merely the leaders, who then send their simple-minded followers to destruction. The leaders are at risk themselves; more important, the perpetrators of the actual terrorist attacks tend to be middle class (though the second intifada, mentioned by Becker, may be an exception). From a labor-market standpoint, there are two important tradeoffs in recruiting a supply of terrorists: quality-quantity, and capital-labor, and they are related. Because terrorists tend to be few in number if only because of the need for concealment, and to be operating in a hostile environment, the recruitment of a large number of poorly trained and motivated cannon fodder is unlikely to be optimal; they are likely to give the game away. Moreover, the most effective terrorism requires some technical sophistication (such as piloting an airplane), and this is a further reason for terrorist leaders to recruit high-quality personnel. The relation of economic development in general or the demographic transition in particular to terrorism is likely to be extremely indirect, and is probably small. If one looks at a list of 195 countries ranked by birth rate, see http://en.wikipedia.org/wiki/List_of_countries_by_birth_rate, one discovers that of the 25 nations with the highest birth rates, all but one (Afghanistan) are in Africa, and Africa has not proved to be a major source of terrorists relative to its vast population. Pakistan has the world’s 57th highest birth rate--27.2 per thousand. This is high--replacement is 21; the U.S. birth rate 14; Germany’s 8.2--and Pakistan is often used as an illustration of a nation that has not made the demographic transition yet. Saudi Arabia, that cradle of Islamic terrorism, has a lower birth rate--24.2--though it is still high. On the other hand, Saudi Arabia is a relatively wealthy country by international standards; its per capita income is similar to that of Poland and Chile. Algeria, with a birth rate (20.8) considerably below Saudia Arabia's, has a severe terrorism problem. Jordan has a substantially higher birth rate than Algeria (in fact it is only slightly lower than Pakistan's), but is not a hotbed of terrorism. All this said, there is some negative correlation between birth rates and terrorism in Muslim countries, but it is weak, and probably swamped by other factors. The major factor in Islamic terrorism may have nothing directly to do with economic development or the factors that influence it; it may simply be the influence of extremist Islamic religious beliefs in particular Muslim nations and communities.