The events of September 11, 2001 have led, among many other things,
to the revival of an old debate about the relationship between
poverty and political extremism. To get at the root of apocalyptic
terrorism, many new initiatives to reduce global poverty have been
proposed. British International Development Secretary Clare Short
advocates a massive international effort to stop poor countries from
becoming breeding grounds for terrorism: "The conditions which bred
their bitterness and hatred", she has said, "are linked to poverty
and injustice." Britain's Chancellor of the Exchequer, Gordon Brown,
has called for a fifty-year Marshall Plan that would disperse aid in
exchange for an end to bad government, in his words, for "the
developing countries pursuing corruption-free policies for stability,
opening up trade and encouraging private investment." Some advocates
call for spending targets to be directly linked to the GDP of donor
nations, overlooking selectivity or effectiveness. Even more
ambitious proposals call for an international tax to limit the
adverse consequences of globalization by financing global public
goods.

The presumed connection between poverty and terrorism raises general
fears of global class warfare--but are these fears justified? The
argument that poverty directly causes terrorism is simple-minded. A
more compelling argument can be made that rising levels of material
well-being, and with them expectations of social and personal
empowerment, actually fuel political extremism and violence. It
cannot be refuted, however, that most of the countries incubating
terrorism tend to be on the low end of the per capita income
spectrum, nor that "poverty and injustice"--and these are not the
same--fuel resentment of the powerful and wealthy to the extent that
terrorism can attract admiration and even peripheral support. And it
is certainly true that bad government is a problem so serious that no
effort at poverty alleviation can succeed without facing it.

Whatever the motives or the depth of understanding underlying them,
new efforts are afoot to tackle global poverty through greater
funding of economic aid programs, and the Bush Administration is
contributing to those efforts. Its proposals on development, if
carried out, could represent a significant milestone in the way aid
is allocated in the future. The President affirmed that "Poverty
doesn't cause terrorism. Being poor doesn't make you a murderer. Most
of the plotters of September 11th were raised in comfort." But he
added that "persistent poverty and oppression can lead to
hopelessness and despair, and when governments fail to meet the most
basic needs of their people, these failed states can become havens
for terror." Thus, the administration has agreed to raise the level
of U.S. foreign economic assistance considerably, but, at the same
time, insists on conditions that can produce some reasonable
assurance that the money will not be wasted. In a sense, the
administration has adopted a kind of market-based aid program kindred
in spirit to that of Gordon Brown's. It proposes, in essence, a
contract with the governments of the world's poorest countries: Cease
the practices that keep the vast majority of your people poor, and we
will help you; persist in those practices and we will refuse to
subsidize them--and you.

In general, this approach makes good sense because economic aid
cannot truly succeed unless it prompts sound institutional reform. It
is the irony of foreign aid that it is used best by the countries
that need it least. But it does not solve the problem that many
countries are poor because their governments resist such reform, and
that such resistance will ultimately generate both humanitarian
crises and havens for terrorism. How do we deal with cases in which
conditions attached to U.S. economic aid are rejected or ultimately
distorted?

This essay has two purposes. The first is to lay out some empirical evidence about the relationship between economic aid and systems of governance. The second is to address the problem of how the U.S. government should deal with difficult partners. On the first task there is a wealth of data that suggests some surprising connections between the length of political tenure, the nature of governance, and the role of aid money. That data and those connections, in turn, can help us to think through the second task.

Political Tenure and Economic Development

Our standard approach to economic policy reform typically assumes
that leaders are rewarded politically if they help their nation
improve its economic performance. In reality, politicians succeed by
helping their constituents and, in the vast majority of poor
countries, politically significant constituencies are not
representative of the whole population. It makes perfect political
sense for autocrats in poor countries to enrich the clique of
supporters around them, even if it means keeping the majority of the
population poor. In short, in such situations, political rationality
and economic rationality are not in alignment. Under such conditions,
external aid money will not help to alleviate poverty; it is more
likely to buttress further what is already an economically
dysfunctional arrangement.

The economic rationality of autocracy underlies a good deal of what
we have learned over the past half century in the experience of
giving, monitoring and evaluating foreign aid programs. Two lessons
stand out.

First, the amount of aid a poor country receives does not directly
correlate with how well it does. Getting external resources into a
country, whether by public or private investment, is not the main
factor. Second, conditionality--the linking of aid money to
macroeconomic reform--does not work as advertised. Aid to low-income
African states has been significant, especially since the end of the
Cold War, and nearly all of it has been conditional on macroeconomic
policy reform. Excluding South Africa and Nigeria, the average
African country received the equivalent of 12.3 percent of its GDP in
Official Development Assistance, a sustained 5 percent increase in
real terms between 1970 and 1995, before declining. This
unprecedented international transfer far surpassed the Marshall Plan,
which at its peak accounted for only 2.5 percent of GDP in France and
Germany. If conditionality had worked, Africa would be the world's
stellar performer. Instead of helping the reform process, however,
substantial aid increases have allowed governments to avoid reforms
that might have led to growth and peaceful political change. Aid
allocation decisions often reward countries that least follow donor
dictates, and it is all too easy to find examples of donor programs
that have undermined the very objectives that donors wished to
promote.

Put differently, we have the economics of foreign aid right, but we
have got the politics wrong. There is a strong consensus about what
sound economic policies are at various stages of economic
development, and professionals in this field recognize in turn that
external aid works best in a good economic policy environment. There
is no consensus, however, about the delivery of results in poor
policy environments. For example, how do we ensure that programs
targeted at reducing infant mortality and making certain that
children learn to read achieve their objectives when the "right
policies" are missing in many of the poorest countries?

Many economists argue that better dissemination of their own ideas is
the solution. In this spirit, Joseph Stiglitz, former chief economist
of the World Bank and Nobel laureate, started an international
network "to ensure in one place a comprehensive catalogue of what is
at issue, and what is the theory and evidence that underlies
positions, so that those in the developing world are in a better
position to make decisions for themselves." But with abundant
technical assistance being provided by international agencies, the
supply of good advice often exceeds the demand. Ignorance of sound
economic policies is not really the issue.

Stiglitz does note, almost in passing however, that poor countries
are poor because their leaders have only half-heartedly implemented
sensible economic ideas. The real question, then, and one that
economists alone are powerless to answer, is why that is so. Why have
many governments rejected good economic advice that they have been
paid to receive? The answer may be found in the economic logic of
autocracy.

Just as we naturally consider successful those leaders who foster
economic growth and prosperity for their citizens, we expect that
leaders who produce famine, poverty and misery will earn a rapid
retirement. But the data show that leaders who produce poverty and
misery through the systematic corruption that is characteristic of
autocracy keep their jobs much longer than do those who enrich their
countries. Indeed, the eight countries consistently rated the most
corrupt in the world--Congo, Iraq, Myanmar, Sudan, Indonesia, Syria,
Pakistan and Burundi--are those in which political leadership has
been most secure, measured by the longevity of its tenure. (Only
countries that have experienced a complete breakdown in social order
can rival an entrenched autocracy in generating extreme levels of
corruption.)

With rare exception, only autocrats--leaders who are unresponsive to
the popular will and who exercise power unchecked either by law or
other institutions--hold on to power for a long time. Over the past
century, the only leaders who have remained in office for forty years
or more have been autocrats. By contrast, nearly half of all
democratic leaders--leaders who hold power at the pleasure of the
voters or an elected legislature--are out of office within about one
year of coming to power. Such a short tenure is true of only about
one-third of autocrats, a remarkable difference in survivability.
Virtually no democrats--but one-quarter of autocrats--stay in office
for more than eight years, even though few democratic leaders are
subject to term limits.

To elaborate the point, one can divide the leadership structures of
poorer countries into two groups: those who depend on a small group
of backers, which may be called exclusive regimes, and those who rely
on a relatively broad coalition of support, which may be called
inclusive regimes. Exclusive regimes tend toward narrow autocracy and
oligopoly; inclusive regimes tend toward what we normally think of as
democracy. But there are many poorer countries short of reaching
mature and genuine democracy that nonetheless exhibit inclusive
characteristics--among these a decade or so ago were, for example,
the East Asian tigers and Chile. In other words, there is a spectrum
of governance structures, and the more inclusive side of it contains
democratizing regimes as well as fully democratic ones. One may
measure how broadly based a regime is by taking into consideration
such factors as constraints on executive authority and the openness
and degree of political competition. A comparison of political
survival rates between these two groups, based on economic
performance, tells a depressing but important story.

Leaders who depend on a broadly inclusive coalition do better at
staying in office only if they manage to promote exceptionally high
growth rates. They do worse if, instead of growth, they promote
rent-seeking opportunities of the sort that typify countries with
vibrant black markets, often controlled by the friends and allies of
the leadership. Inclusive leaders who promote growth stay in office,
on average, 15 percent longer than those who do not.

The reverse is true for those who rule at the pleasure of a small,
exclusive group. Exclusive leaders who rely on black-market
corruption have a better chance of staying in power than those who
engender high rates of growth, staying in office, on average, 25
percent longer. Indeed, at all periods during their tenure in office,
these leaders do much better at retaining their jobs if they promote
black marketeering, corruption and cronyism--distorting the
economy--than if they promote economic policies that lead to growth
and prosperity (see table). Why does this perverse outcome occur? As
suggested above, leaders who would keep their jobs must produce what
their supporters want; when those supporters are unrepresentative of
the country, autocrats will not pursue policies that encourage the
creation of healthy, educated, prosperous citizens.

Autocrats not only retain power by maintaining the loyalty of a
relatively small group of supporters--which usually include those who
control the military, the civil service, the communications and
information infrastructure, as well as key economic levers--but they
also have an interest in keeping that core group as small as
possible. In a poor country, an autocrat faces personal political
risks if he implements policies that dissipate resources away from
the few upon whom he relies to those who have little say in ensuring
his political survival. It is therefore politically irrational to
implement transparent economic policies aimed at protecting and
promoting property rights, rule of law, a broadly educated
population, low taxes and free trade, if they enable challenges to
the incumbent. It is not in an autocrat's interest that people have
ways to enrich themselves that he does not control.

This is why autocrats face their highest risk of being deposed in
their first year in office; they have not yet identified their most
loyal backers and have not yet fully secured their ability to
transfer benefits to them. With time and experience, they get better
at identifying those on whose support they really rely. They discover
that excluding "the many" from sharing in the wealth of the country
is the best way to reward a small clique of supporters.

Even worse, such systems tend to perpetuate themselves, which is why
a reform-minded leader who manages to come to power in such a system
faces excruciating dilemmas. If he is committed to promoting growth
and prosperity, he may find that pursuit of such goals can ensure the
loss of office. Elections give a mandate to a democratic leader, but
in pursuing reform, even a democratic leader may be forced to
implement policies that injure the interests of the constituents who
brought him to power. Any politician unable to satisfy his core
constituents faces a risk of their defection to another politician
who shows more promise of improving their lot. After listening to a
long list of measures designed to improve his country's economy,
President Rafael Caldera of Venezuela told Nobel laureate Douglass
North, "If I were to do the things you are recommending, I would not
survive in office long enough to enjoy the benefits."

Foreign Aid and Autocratic Longevity

How does the provision of foreign aid fit into the logic, and
economic implications, of autocracy? In those many cases in which aid
has not led to economic growth, what has it accomplished?

External aid often promotes longevity in office for autocratic
leaders who are otherwise at risk of being deposed; it simply makes
it easier for them to patronize their core group of supporters. In
such cases, aid not only fails to promote economic growth, but it
also diminishes the odds that the political system will evolve in a
more inclusive, democratic and growth-oriented direction.

This may seem too large a claim to some observers. After all,
external aid generally comprises only a small component of a nation's
total economy. Since 1975, for instance, international aid has
averaged only about $7 or $8 per citizen. Such numbers imply that
foreign assistance is not significant enough to reshape economic
prospects and barely enough to provide relief to the world's poorest
people. This assumption, however, misses the fundamental benefit that
aid provides to autocratic leaders, and again, the data illustrate
it. Autocrats in countries with below-average growth rates who do not
get aid have a 25 percent chance of staying in office for five years.
If they receive economic assistance, that survival time rises to
seven years, a 40 percent increase. A few dollars of aid per capita
is small in terms of any impact on the national economy, but it is
huge with respect to helping autocrats enrich their small coterie of
supporters.

On average, every dollar of per capita foreign aid improves an
incumbent autocrat's chance of surviving in office another year by
about 4 percent (even after taking into account the independent
effects on political survival exerted by such factors as the
country's economic growth rate, black market exchange rate premium,
national debt, and its geographic situation). Since the average
autocracy gets about $8 per capita in aid, foreign assistance may
boost the survival prospects of poorly performing leaders by 30
percent or more.

The data also suggest that giving assistance to leaders of exclusive,
autocratic regimes is especially critical early in their tenure, when
they do not yet have a solid hold on office. Aid frees up money for
the incumbent leadership to buy support from a relatively larger
number of backers without having to generate the wealth to do so
through policy choices. Even small amounts of aid provide major
opportunities for novice autocrats to outbid political rivals and
hence deter defections to them. Mobutu Sese Seko, for instance, used
foreign aid after his successful 1965 coup (as well as the
nationalization of assets) to buy the loyalty of the Zairian military
and the country's economic elite, thereby securing his hold on power.
(His eventual overthrow, in the late 1990s, was directly tied to his
growing inability to meet the military's payroll.)

Aid is most likely to be effective if it is given as a benefit to
those who have demonstrated that they can use resources to improve
economic performance. However, assistance given in response to
expressed intentions, lacking a proven track record of effectiveness,
is likely to lead to perverse incentives. This creates the central
paradox of foreign aid: under many of the least politically inclusive
systems, good policy is bad politics, and bad policy can be good
politics. In those more inclusive systems, in which good policy is
also good politics, leaders face greater obstacles to maintaining
incumbency because it is not enough for them to distribute a finite
number of private goods to a small oligarchy to win support. They can
only remain in office if they generate and sustain policies that
deliver public goods (schools, health care, a safe environment) to
the general populace.

The policy implications of the dichotomy between exclusive and
inclusive systems of governance go beyond the matter of leadership
tenure. When the political system is dominated by a small coalition
of cronies, relatives or military officers, citizens do markedly
worse on public welfare or humanitarian indices than do their
counterparts who live under more inclusive systems. Across the board,
the data show that more inclusive systems generally do a better job
at producing safe drinking water, expanding public education,
offering access to medical care, encouraging free trade, avoiding
corruption and black marketeering, attracting investors and so forth.
Moreover, even quite poor inclusive societies usually offer more of
these advantages than do autocratic countries lucky enough to possess
some important and valuable resource. The reasons are not difficult
to fathom: inclusive governance promotes greater government spending
on social policy because, in such systems, the longevity of political
leaders is directly tied to the welfare of the majority. El Salvador
and Jamaica are two excellent examples of relatively poor but
inclusive societies with above-average social welfare (as
demonstrated by their low infant mortality rates and high-quality
drinking water). By contrast, during their non-democratic years,
Mexico and Brazil had above-average income levels, but performed
poorly on these social indicators.

Consider, too, the impact on life expectancy at birth of the
coalition size that a ruler uses to govern. Being born in a polity
that scores highest on the inclusiveness index adds nearly fourteen
years to life expectancy, whereas an order-of-magnitude increase in
per capita income adds only five years. Both are significant, but the
impact of coalition size can have dramatic results, both by
increasing life expectancy (because of government spending
priorities) and by creating a more open economic system, since a
competitive political system is a significant contributor to economic
growth. By way of illustration, per capita income in Brazil in 1972,
eight years after the military coup, was $2,907. In the same year in
Jamaica, a functioning if narrowly-based parliamentary democracy, per
capita income was about the same, $3,099. According to the "coalition
size" index of inclusiveness, however, Jamaica scored 1.0 while
Brazil scored 0.25. It is therefore not unusual--though it may be
surprising to some--to find that whereas life expectancy in Brazil in
1972 was 59.8 years, in Jamaica it was 68.6 years, nearly a decade
longer. Similar evidence can be marshaled using other measures of
social welfare: regarding the equality of educational opportunity for
women and men, differences in infant mortality rates, and so on. What
is more, the data also suggest that when a government switches,
whether by choice or the compulsion of circumstances, from being
exclusive to being inclusive in nature, economic growth shows a
marked improvement over the next three to five years.

In designing assistance programs, it is crucial to recognize that the
social welfare improvements sought by donors are best promoted by
encouraging leaders to adopt more inclusive political institutions.
Those who fail to do so signal their unwillingness to put their hold
on office at risk, even when doing so stands to improve the
well-being of their citizens. If assistance programs are not designed
to address the political incentives of leaders, further aid to
autocratic regimes will continue to present the sorry record of the
past: assistance without proper political incentives and inducements
will not bring political civility and prosperity. Instead, donor
assistance will continue to substitute for governmental efforts to
make citizens better off, and may even help pay to keep failed
leaders in office.

Engaging Difficult Partners

Recognizing the political causes of poor economic performance is
critical, but it does not relieve us of difficult choices. It is
correct to say that aid will do more good in consolidating progress
in countries already pointed in the right direction than it will in
those countries still burdened by bad government. But that does not
mean we can walk away from the people who live in those unfortunate
lands where misery, resentment and the potential incubation of
terrorism is likely to be most acute. So what to do? We suggest seven
basic guidelines.

First, donors must take greater responsibility for outcomes. Whose
fault is it when aid is ineffective because external efforts are not
well targeted, well coordinated and rigorous in measuring results? If
money is dispensed without adequate assessment of the potential for
mismanagement? If there is inadequate information about what the aid
is designed to accomplish? Donor countries need to institutionalize a
more credible monitoring system with third-party safeguards that
unambiguously separate disbursement from assessment and monitoring.
The international financial institutions (IFIs) should introduce an
"inspector-general function." A permanent quality control staff,
composed of individuals unaffiliated with any IFI, that sits in on
important decisions and is enhanced with the participation of the
intended beneficiaries, is one way to go.

Second, end the debt trap. Loans to already heavily indebted
countries reduce their opportunities for economic growth. Loans have
been the preferred method of assisting developing countries because
donors believed that loans would foster greater responsibility for,
and ownership of, outcomes. However, the assumption that poor
countries are less likely to squander funds that they must someday
repay has proved wrong. Most non-democratic leaderships have been
more than happy to consume the benefits of the loans--enriching
themselves and their supporters--while passing the future burden of
debt service and repayment onto the countryat large. This is just the
economic logic of autocracy at work across time rather than across
the population at any given moment. The resultant "debt overhang",
not surprisingly, diminishes the appetite of investors for developing
market investments. Even if the society offers otherwise attractive
investment opportunities, the debt burden reduces the means to
capitalize on them. With inadequate cash even to pay current
obligations, little is available to pursue new investment.

The crushing weight of debt is so debilitating that the idea of debt
forgiveness has gained widespread popularity. Debt forgiveness,
however, does not address the central problem of perverse incentives.
Excessive debts exist because the political incentive structure
protects leaders from the consequences of poor economic results.
There is a strong correlation between the magnitude of "debt
overhang" and the degree to which a small coterie of insiders control
government. Inclusive regimes, by contrast, have a much smaller
debt-to-revenue ratio. Statistics collected on 46 countries
concerning the 1946-93 period demonstrate this fact, even after
controlling for other variables such as the impact of previous
inflation and growth rates on indebtedness. Thus, debt forgiveness
must not be misused to wipe the slate clean, only to start a new
cycle of mismanagement under a new clique of autocratic rulers.

Third, use grants to work around governments in order to stay engaged
with difficult partners. It is callous to withhold all aid while
awaiting the establishment of sound political and economic
frameworks in impoverished autocracies. Many of the poorest countries
have governments and institutions that either do not support
development or cannot use development assistance effectively. In such
circumstances, grants may be in order.

Grants are easier to monitor than direct budgetary support because
they are given for a particular project with narrowly defined goals
and, most importantly, they need not be made directly to governments
but can be used as a tool to build civil society. Directly
administered, grants represent just the kind of challenge needed to
overcome the entry barriers set up by governments that survive as
incompetent monopolists. A donor can put the project out to bid,
select the contractor, and then pay for an independent performance
assessment when the project is completed. If properly designed,
grants also enable donors to empower people through building
institutions outside of government. Whereas governments are unlikely
to borrow to build capacity in society, a grant facility will allow
IFIs to contribute to such capacity. The programs supported by grants
can also be used as an incubator of leaders who can help focus
demands for change and give those neglected by their government a
voice. By empowering groups outside the civil service--which is often
racked by patronage and corruption--to administer grants, donors
build moral space that allows citizens to challenge the credibility
of abusive governments. Concrete examples of successful project
completion can lead populations to demand more from their
governments. Unlike loans, grants deprive the government of the
opportunity to channel the money to cronies rather than to those most
competent to satisfy the grantor's objectives.

Fourth, stop trying to address economic questions in isolation from
political considerations. Most causes of poverty in the world can be
linked to leaks in the ship of state. That is why developing tools
for broad stakeholder alliances (such as the World Bank's new poverty
reduction facilities) is so important; otherwise, donors risk
unwittingly collaborating with leaders who have created the
conditions for economic collapse in the first place. Linking the role
of aid in humanitarian assistance to its role in policy reform is
absolutely essential. This requires coordinating two objectives that
require different mechanisms (loans and grants), different partners
(the state and civil society), and different delivery systems (for
example, the multilateral development banks rather than organizations
such as the World Health Organization). Each grant and each project
must serve as a building block toward the creation of a more
inclusive regime.

Fifth, aid disbursement techniques must prevent backsliding toward
exclusive governance. The tendency for an incumbent to shift policies
toward autocracy will be enhanced if unrestricted aid is given before
a country has a track record of good governance. Aid given before
policy results are firmly established allows countries that only
promise reforms to revert to bad policies in order to be eligible for
fresh aid. Countries with bad policies can then perversely obtain
more aid than countries with socially productive policies. Aid could
be used more effectively if donors find ways to reward non-lending
activities, such as designing a change strategy for a country that is
not eligible for assistance because of failed governance, or helping
to create a domestic constituency that favors reform.

Sixth, encourage trade liberalization as an essential component of
global poverty reduction. Full trade liberalization has the potential
to lift at least 300 million people out of poverty by 2015 and can
create domestic coalitions for reform and openness. Mexico's trade
with the United States has more than tripled since the passage of
NAFTA and, as a result, per capita income has begun to rise. In turn,
economic growth has helped to consolidate both democratic and
economic reforms. By creating jobs, it has thinned patronage networks
and eliminated the traditional rents from controlled markets that
once kept corrupt leaders in power. A new and independent middle
class is emerging to demand genuine accountability from Mexican
leaders.

The World Bank notes that the average poor person confronts trade barriers roughly twice as high as those confronting the typical worker in an OECD country. In general, tariffs in advanced countries on imports from developing countries are four times higher than those on imports from other advanced countries. Exceptional protection is imposed against imports from many developing countries of many farm products (particularly in the EU and Japan), of textiles and clothing (notably in the United States and Canada) and of footwear. According to the World Bank, a new trade round coupled with market reforms could boost global income by $2.8 trillion over the next 15 years.7 If sub-Saharan Africa, for example, had maintained the share of world trade it had in 1980, annual exports would be $191 billion, more than double what they are today and almost four times the amount of the most ambitious calls for increased foreign aid.

Finally, remain mindful of the inherent limitations of aid and focus on the maintenance and expansion of private sector flows. No matter how successful the United States and other donor counties are at increasing state aid, it can never substitute for a healthy and conducive environment for private sector investment and capital formation. The hallmark of a good aid policy is one that strives to build a transmission belt from public to private investment flows.

THERE ARE good reasons to want to alleviate poverty, especially in the wake of September 11. But poverty is not exclusively an economic problem; it is a problem of political economy. Until this fact is squarely faced, no amount of aid will come remotely close to solving the problems that breed the hopelessness and despair the President wants to alleviate. We realty do have a passing fair idea of what policies can overcome poverty, but most efforts to educate leaders and bureaucrats about good economic practices flounder because we have too often ignored the reasons that leaders fall to do what is right for their country. To be frank, much U.S. economic aid during the Cold War was motivated by a desire to keep certain strategically situated autocrats on our side during that conflict. How well or how poorly they used our money was a secondary concern. This is no longer the case. Our aims remain political, as well they should, but now those political aims can be served only by getting positive results from the aid effort.

The Bush Administration has taken a major step forward in establishing new criteria for foreign aid, by explicitly linking assistance to good governance. The President has wisely avoided facile and misleading depictions of the sources of global poverty and terrorism. But many difficult questions remain. How can the needs of people living in poor countries be served without progress toward democratization, especially when they lack the institutions needed to generate growth or to interact with the global system? If we agree that, until the fundamentals are in place, development assistance will reach only a fraction of its potential effectiveness, does this imply that we should ignore poorly performing nations? How do we distinguish between "states that are poor because they perform poorly" and "states that perform poorly because they are poor"?

With few options for successful assistance, the challenges posed by poorly performing countries go well beyond the usual limits of what we have understood to constitute development policy. What makes the task ahead particularly daunting is that the ultimate solution to poverty comes down to nation-building. Yet, in vast parts of sub-Saharan Africa, Central Asia and elsewhere, states have atrophied, enabling them to be abused or captured by transnational terrorists, traffickers and smugglers. In a world where development is state-driven, what will happen to countries without minimally functioning states? What institutional alternatives should we be thinking about where an effectively functioning state is a distant reality? Overcoming the curse of bad government will require some particularly creative thinking, and much patience. Hard labors and long delays lay ahead, but a meaningful marker has been put down. At last we seem to have the definition of the problem down right.

Bruce Bueno de Mesquita is the Silver Professor of Politics at New York University and a Senior Fellow at Stanford's Hoover Institution. Hilton L. Root served in the current administration as U.S. Executive Director-Designate of the Asian Development Bank and as an advisor to the U.S. Treasury. He is currently a visiting fellow at the Economic Strategy Institute.