Can aid work? Thinking about development strategy.

In the 1990s, Canadian domestic concerns--deficits and rising debt, Quebec secession, reform of social programs--squeezed out concern with the world beyond our borders. Domestic concerns will always be at the cote of political debates, but in this decade international concerns, including development policy, have assumed new importance. Why is the typical French citizen four times richer than the typical Algerian across the Mediterranean, and does that have anything to do with the rise of politicized Islam and riots in French cities? Why is the typical American four times richer than the typical Mexican on the other side of the Rio Grande, and will NAFTA help close the gap? Why are Canadians four times richer than citizens of the Dominican Republic; in turn, why are those living in the eastern two-thirds of Hispaniola four times more prosperous than Haitians inhabiting the western third?

While there are self-interested reasons for Canadians and other members of the OECD club of wealthy nations to understand the dynamics of development and try to do something about it, humanitarian considerations inevitably loom large in this discussion. Such four-to-one ratios cast a moral shadow over our Canadian prosperity. Had he or she been born in, say, Port-au-Prince, with access to the education and health services of ordinary Haitians and subject to the same level of misgovernment, would a Canadian be any more productive than the Haitian? Probably not. If our prosperity has much to do with the good fortune of being born in the right place, surely we have some obligation to share that good fortune and undertake development aid.

One aspect of this obligation is Canadian engagement in the UN's Millennium Development Project. In 2000, a summit of world political leaders adopted eight "millennium development goals," goals that in turn entail a series of targets to be realized by 2015. (See Box 1.) In a major 2005 report, the UN projected that, as a region, sub-Saharan Africa will meet none of the targets. The dire prospects for the region have become a preoccupation of international politics, sufficient to have dominated the 2005 G-8 gathering in Scotland. Along with the other participants, Canada has agreed to a much-enhanced engagement in overseas development.

Box 1: Millennium Development Coals

In 2000, world leaders adopted the UN Millennium Project that contained
a number of development goals to be realized by the year 2015. Listed
here are the eight goals and a selection of accompanying targets:

* Goal 1: Eradicate extreme poverty and hunger. Targets 1 & 2: Halve,
  between 1990 and 2015, the proportion of people whose income is less
  than $1 a day and who surfer from hunger.

* Goal 2: Achieve universal primary education. Target 3: Ensure that,
  by 2015, children everywhere, boys and girls alike, will be able to
  complete a full course of primary schooling.

* Goal 3: Promote gender equality and empower women. Target 4:
  Eliminate gender disparity in primary and secondary education,
  preferably by 2005, and in all levels of education no later than
  2015.

* Goal 4: Reduce child mortality. Target 5: Reduce by two-thirds,
  between 1990 and 2015, the under-five mortality rate.

* Goal 5: Improve maternal health. Target 6: Reduce by three-quarters,
  between 1990 and 2015, the maternal mortality rate.

* Goal 6: Combat HIV/AIDS, malaria, and other diseases. Targets 7 & 8:
  Have halted by 2015 and begun to reverse the spread of HIV/AIDS, and
  the incidence of malaria and other major diseases.

* Goal 7: Ensure environmental sustainability. Targets 9 & 10:
  Integrate the principles of sustainable development into country
  policies and programs and reverse the loss of environmental
  resources. Halve, by 2015, the proportion of people without
  sustainable access to safe drinking water and basic sanitation.

* Goal 8: Develop a global partnership for development. Targets 12 &
  13: Develop further an open, rule-based, predictable,
  nondiscriminatory trading and financial system (includes a commitment
  to good governance, development, and poverty reduction--both
  nationally and internationally). Address the special needs of the
  Least Developed Countries (LDCs): tariff- and quota-free access for
  LDCs' exports, enhanced program of debt relief for heavily indebted
  poor countries and cancellation of official bilateral debt, and more
  generous official development assistance for countries committed to
  poverty reduction.

What has transpired since 2000? To quote the major 2005 report issued
by the UN Millennium Development Project (2005, 18-22):

* North Africa: Moving in the right direction on every indicator, North
  Africa needs to accelerate progress to achieve the Goals. It has seen
  modest economic growth since 1990 and is on track to reach the target
  of halving the poverty headcount rate....

* Sub-Saharan Africa: The region is off-track to meet every Millennium
  Development Goal. If has the highest rate of undernourishment, with
  one-third of the population below minimum level of dietary energy
  consumption. Sub-Saharan Africa has the lowest primary enrolment
  rates of all regions. Despite recent progress, gender disparity at
  the primary level is ... the lowest of all regions. The HIV/AIDS
  crisis is devastating much of the continent.... The region also has
  the highest TB incidence in the world and the highest maternal and
  child mortality ratios....

* East Asia: The region has seen rapid falls in income poverty and in
  hunger, and improvements have also been recorded in gender equality,
  education, and child survival.... Nearly 200 million people in East
  Asia live in slums, lacking access to secure housing and essential
  services....

* Southeast Asia: The region is on track to meet the Goals for income
  poverty, hunger, child mortality, and gender equality.... obstacles
  to achieving the Goals include rising numbers of people infected
  with HIV and TB, high maternal mortality, rapid deforestation and
  destruction of coast and marine environments....

* South Asia: The region has made some major strides in overall poverty
  reduction thanks largely to rapid economic growth in India ... [but]
  South Asia is still home to more poor people than any other region
  and remains off track for meeting many of the Goals. Primary
  enrolment and gender equality indicators are lagging. Child health
  is improving but hot quickly enough to meet the targets, while
  maternal mortality rates remain high, and, without urgent action,
  HIV is poised to spread....

* West Asia: This region ... is off-track for a majority of the Goals.
  Both income poverty and hunger are increasing, and progress toward
  gender equality has been slow....

* Latin America and the Caribbean: The region has experienced little
  economic growth since 1990, yielding stagnant poverty headcounts and
  persistently high inequality.... Accelerating deforestation remains
  a major issue throughout the region....

* CIS countries in Europe: In the 1990s ... the CIS countries had their
  economies collapse, with significant increases in poverty and hunger.
  Most are now making progress, but a few have not yet achieved
  pre-reform levels of per capita income....

* CIS countries in Central Asia: Central Asian countries experienced
  an increase in poverty levels after the collapse of the Soviet
  economy.... Since 1990 the poverty headcount rates have risen
  significantly in several countries, as have undernourishment and
  child mortality rates....

A symbol of this shift in Canadian political priorities was Ottawa's publication in 2005 of a four-volume International Policy Statement. (See Box 2 for a precis of the volume on development strategy.) Canada has no legacy of imperialism to overcome; it has no announced enemies, and occupies a prosperous corner of the globe. Hence, the immediate benefits to Canadians from a more prosperous world beyond North America are small. Nonetheless, there are self-interested reasons to engage with the fate of developing nations. Two such are international aspects of terrorism and the spread of new diseases. The authors of the Statement expressed Canada's self-interest as follows:

   Today war between major powers seems more remote than ever. The
   threats we face now are different--from non-state actors such as
   terrorists to new infectious diseases. Increasingly, they are
   beyond the control of any one country. Where states pose threats
   today--for example, those developing nuclear weapons--they often
   operate beyond the constraints of international law. We need
   new strategies to influence their behaviour. In addition, many
   countries today threaten us--and their own citizens--not because
   of their strength, but because of their fragility. Their collapse
   creates humanitarian tragedy and poses wider security threats,
   including to Canadians. These challenges, which increasingly
   transcend national boundaries, demand more than ever that
   governments work multilaterally to find durable solutions.
   (Canada 2005b, 9.)

Box 2: The International Policy Statement

The International Policy Statement contains the commitment that annual
official development assistance (ODA) reach $5 billion by 2010, double
the level in 2001. In 2004/05, Ottawa budgeted $3.7 billion to
ODA--including humanitarian relief associated with the December 2004
tsunami. It also repeats a commitment that, at some undefined point in
the future, ODA will reach 0.7 percent of GDP. (The 2004/05 expenditure
is about 0.3 percent of 2004 GDP. Given 5 percent nominal GDP growth,
the $5 billion commitment implies a similar ODA/GDP ratio of 0.3
percent in 2010.)

CIDA has promised that, by 2010, at least two-thirds of direct
country-to-country assistance will be focused on 25 developing
countries. The following 25 countries have been selected:

* Africa (14): Benin, Burkina Faso, Cameroon, Ethiopia, Ghana, Kenya,
  Malawi, Mali, Mozambique, Niger, Rwanda, Senegal, Tanzania, Zambia.

* Americas (4): Bolivia, Guyana, Honduras, Nicaragua.

* Asia (6): Bangladesh, Cambodia, Indonesia, Pakistan, Sri Lanka,
  Vietnam.

* Europe (1): Ukraine (Canada 2005c).

The statement promises to "strengthen coherence among the Government's
aid and non-aid policies and actions" (Canada 2005a, 8). This, in turn,
becomes a commitment to "target and concentrate programming in rive
sectors directly related to achieving the MDGs" (Canada 2005a, 11).
Promotion of gender equality will be a "crosscutting theme." The
statement defines the five sectors as follows:

* Sector 1. Promoting Good Governance

Particularly through Canada Corps [see below], and through other
programming channels, Canada will assist countries to build the
conditions for secure, equitable development by promoting good
governance, focusing Canadian efforts on democratization, human
rights, the rule of law, public-sector capacity building, and
conflict prevention. (p.12)

* Sector 2. Improving Health Outcomes

Canada will assist countries to improve health outcomes, particularly
among the poorest, through a focus on: preventing and controlling
high-burden, communicable, poverty-linked diseases (especial]y
HIV/AIDS); strengthening the capacity of health systems; improving
infant and child health; strengthening sexual and reproductive
health; and improving food security. (p.14)

* Sector 3. Strengthening Basic Education

Canada will assist countries to accelerate progress toward ensuring
that every girl and boy is able to access and complete free,
compulsory, quality basic education. (p.15)

* Sector 4. Supporting Private Sector Development

Consistent with the recommendations of the UN Commission on the Private
Sector and Development, Canada will support developing countries' own
efforts to strengthen their private sector as the engine of economic
growth through creating an enabling environment, including through
"smarter" regulation; promoting entrepreneurship; and supporting
connection to markets (economic integration). (p.16)

* Sector 5. Advancing Environmental Sustainability

Environmental sustainability will Re a priority in Canada's development
cooperation. It will also be systematically integrated into decision
making across all programming. Canada will assist countries to create,
maintain and enhance environmental sustainability, particularly in
relation to climate change, land degradation, freshwater and
sanitation, and urbanization. Canada will also work to strengthen
global environmental agreements, capacity development and multilateral
funding institutions. (p.18)

In addition to bilateral aid, the statement promises Canada's continued
commitment to multilateral cooperation. Currently Canada delivers over
40 percent of its aid via multilateral organizations (Canada 2005a,
27). A final component of the statement worth noting is creation of
the Canada Corps as "a new vehicle to strengthen Canada's contribution
to human rights, democracy and good governance internationally. Canada
Corps will develop collaborative partnerships across government, NGOs,
the private sector, and with Canadian citizens to bring greater
engagement, expertise, coherence and recognition to Canadian governance
interventions abroad" (Canada 2005a, 28).

Without ruffling diplomatic feathers by citing particular countries, the statement refers to threats "from a large number of weak, ineffectively governed states" (Canada 2005b, 9). The most obvious "non-state actors" are radical Islamic groups. Such groups may espouse seemingly irrational goals--a rejection of the modern and return to a religious-inspired feudal ideal--but there are rational reasons for the rise of politicized Islam. (1) An understanding begins with the failure of Arab states to realize reasonable economic and social progress over the last half century. Part of any strategy to counter their influence is to foster economic development in the Muslim world. Presumably, this rationale underlies the fact that Afghanistan is not only the location of significant Canadian military activity but is currently the largest recipient of Canadian bilateral aid (Canada 2006).

The above passage also refers to "new infectious diseases." There is a grisly parallel between terrorism and, on the other hand, AIDS, new diseases such as ebola, and virulent mutations of the pathogens causing tuberculosis, malaria, and other well-known infectious diseases. Deaths attributed to the former have forced us to take seriously the politics of the Muslim world; deaths caused by the latter are doing the same for the politics of sub-Saharan Africa. Contending with these threats to world health requires policies to grapple with stagnant and impoverished economies (which encourage large-scale migration and hence rapid spread of disease) and poorly governed states (which are unable to undertake effective public health programming).

In summary, this Commentary proceeds as follows. It offers a synopsis of major themes in current academic development policy analysis. There follows a cautionary section on the limited ability of development aid to achieve results, followed by discussion of four general policy recommendations. By way of conclusion, I offer a sketch of the relevant and interrelated variables that affect aid outcomes. The International Policy Statement proposes that Canada concentrate its bilateral aid on 25 countries. Appendix A surveys the economic growth, GDP, education and life expectancy outcomes among those 25 targeted countries.

Why are the poor poor?

If Canadian development aid is to be more than a humanitarian exercise, it requires an understanding, however imperfect, of the underlying dynamic of economic development.

The "developing world" comprises regions and countries of immense diversity. It includes cultures that crossed the threshold from hunting/gathering to settled agriculture many millennia ago, and others that have yet to do so. It includes countries with long-established formal state hierarchies, and others in which tribal loyalties trump loyalty to a state formed within living memory.

It is also worth underlining the diversity of recent economic performance, and consequent diversity in poverty trends. A straightforward means to do this is to consider changes in the numbers and proportions living below two frequently used poverty lines: those living in extreme poverty on less than $1 a day, and those living on less than $2 a day. There are two hugely important success stories over the last quarter century: China and, to a lesser extent, India. In the 1990s, under both measures, the number of poor in East Asia declined by more than 200 million. (See Table 1.) In South Asia, the number with less than $1 a day fell, but the number falling below $2 a day did not. Based on either poverty line, poverty declined in both regions as a share of population. (2) At the other extreme, several sub-Saharan countries collapsed into chronic civil war, and regional wars centred on the Democratic Republic of Congo have caused several million casualties. Poverty in sub-Saharan Africa has unambiguously worsened in recent years--as measured by those with less than either $1 or $2 a day, both in absolute numbers and as a proportion of the regional population.

Since the demise of European colonial empires following World War II, academics and political leaders have been searching for the optimal set of policies whereby developing nations can transform poverty into prosperity. Initially, development agencies recommended policy on the assumption that there exists a ratio, varying according to level of development, relating any country's stock of fixed capital to annual GDP. Policy recommendations called on aid agencies to fill a "savings gap"--to provide grants and soft loans enabling investment in infrastructure and capital-intensive industries at levels beyond the constraint imposed by the country's domestic savings. Freed from colonial tutelage, newly sovereign states would, it was assumed, make good use of the aid. In the three decades following WWII, much investment in the developing world was undertaken directly by government or by state-owned enterprises. (3)

By the 1980s, conventional wisdom among development analysts had swung from left to right. Pessimism reigned about the potential of the typical government in the developing world to undertake efficient resource-allocation decisions. The state should establish an independent judiciary able to enforce commercial contracts, maintain international trade unhindered by tariffs and non-tariff barriers, and assure macroeconomic stability by balancing the budget (at low levels of taxation) and by controlling the money supply. Having accomplished these tasks, private entrepreneurs would do the rest.

The minimum state strategy generated, however, no clear-cut cases of sustained economic success beyond Chile and Hong Kong. By the 1990s, conventional wisdom among analysts and major donor agencies shifted again. Here, at the risk of serious oversimplification, I summarize three dominant themes of current thinking on development policy.

"Social capital" and "trust"

An old joke from the Soviet era has one comrade asking the other, "what is the definition of socialism?" The second replies, "an economic system in which the state pretends to pay us and we pretend to work." The underlying insight is that the absence of trust became, under communist rule, a generalized non-cooperative strategy pursued by the great majority of the population.

Following the collapse of the Berlin Wall, development analysts initially assumed that workers would work and the state--or more likely private employers--would pay decent wages. Market relations would quickly blossom and government performance would improve. Instead, the economies of most countries of the former Soviet empire experienced painful economic contraction and serious disruption in quality of publicly delivered social services. Political corruption erupted as a major problem. Yugoslavia disintegrated in civil war.

This post-1989 performance gave spur to analysis stressing that successful market competition and democratic electoral politics require important elements of mutual trust, a willingness to obey informal conventions, and so on. Without these elements of cooperation, economic development falters. The theme figured prominently in, for example, the 1997 Annual Development Report of the World Bank (1997a). Among the important studies elaborating on this theme was that of Robert Putnam and colleagues (Putnam et al. 1993) on regional differences in quality of Italian government and civic institutions. They explained the prosperity of northern Italy in terms of extensive voluntary participation in networks of civic activities, which in turn nurtured high "social capital." Conversely, the poverty of southern Italy could be explained in terms of low "social capital" that veered into political corruption. In another frequently cited study, Francis Fukuyama (1995) argued that economic prosperity in a market economy requires a high level of informal contracting among individuals and firms, and that such contracting is possible only in a context of mutual trust. In the absence of such trust, complex economic contracting is constrained to agents bound by family or other narrow loyalties.

"Market failures" in private health/education markets

A robust generalization about economic development is that no country realized sustained economic growth in the 20th century without achieving near-universal literacy. To be functionally literate as adults requires that children spend a minimum of rive or six years at school (Bruns et al. 2003, 29ff). And, while the elite in all societies usually invest in educating their children, there are rational reasons to explain why the majority in developing countries, in the absence of a decent public school system, do not.

One reason is the short time horizon that poverty imposes on parents. The benefits to a family of educating children may be realized only a decade or more in the future, once children have become adults. The direct costs to parents of educating their children--that is, costs in addition to taxes paid--include out-of-pocket expenses (school fees, transport of children to school, and so on) plus the value of foregone work (usually unpaid domestic work such as care of pre-school children). For poor families, these costs loom large; the future benefits may be heavily discounted.

Another reason pertinent to patriarchal cultures is that women may be barred from many productive jobs. In such circumstances, the expected material benefits from formal education of girls are low. Furthermore, it is nearly impossible for village parents to measure the quality of locally available schools. This combination of factors results in a weak demand by the majority of parents on behalf of good quality K-12 schooling. Similar arguments arise in assessing the demand for primary health care.

Appreciation of these problems lies behind a second trend of the 1990s: a renewed emphasis that, for reasons of both equity and efficiency, developing countries should invest heavily in delivery of high-quality primary/secondary education--for girls as much as for boys--and in basic ambulatory health care. (4) On the other hand, these are complex services. Delivering education and health services of reasonable quality requires governments of reasonable competence. This brings us to the third overriding theme of contemporary development analysis.

Good government

By the 1980s, the majority of development analysts agreed on the importance of enabling private entrepreneurs and competitive markets. For private markets to function, the state must assure that the rule of law more-or-less prevails. Complex long-term fixed investments will not take place without a reasonable expectation that the state will enforce contract provisions. On that there is consensus. Relative to the minimum state agenda of the 1980s, however, the current consensus expects the state to perform significant additional tasks.

If domestic and international competitors provide the incentive to realize efficiency in the market, what comparable forces operate in the public domain to realize reasonably "good government"? The blunt answer to this question is, there are no comparably powerful forces. Hence, a dilemma. Successful development in any country requires the host government to be sufficiently powerful to tax and use the revenue to maintain apolitical courts, reasonable quality schools and primary healthcare facilities. But, a state powerful enough to do these things is an inducement for "state capture" by coordinated groups wishing to divert resources. (See Box 3 on the dire problems of school governance in Bangladesh.) This diversion takes many forms: politicians over-hiring in state-owned enterprises, thereby creating dependent clienteles; officials enriching themselves by the sale of discretionary licences and permits; meter readers extracting bribes from customers under threat of power shutoff; and customs and excise officers exploiting their discretion over flow of imports and exports.

Box 3: School Governance in Bangladesh

In terms of public administration, Bangladesh is arguably the most
centralized of the world's populous countries. (Small in area, it has
a population of 150 million.) Elected local governments (union
parishads) have very limited taxing authority, and no formal authority
over local schools. The administration of government schools, including
the hiring of teachers, relies on administrators appointed, directly or
indirectly, by the central government in Dhaka or by the relevant
Member of Parliament.

The results of centralized administration are hot impressive. The
Campaign for Popular Education (CAMPE), an organization that evaluates
Bangladesh education outcomes, summarized its most recent findings as
follows:

* One out of five children does not enrol in primary schools.

* One out of three of those enroled drops out before completing
  primary education.

* One out of three who complete five years of primary schooling still
  remains non-literate or semi-literate.

* Therefore, the large majority of children, mostly poor and
  disadvantaged in other ways, grow up without basic skills and
  preparation for life (Ahmed & Nath 2005, 60).

There has been improvement along some dimensions. The enrolment rate
of 6- to 10-year-olds has increased over the last decade, and the
gender difference among primary school students has been eliminated.
However, only about 35 percent of children are currently achieving
reasonable literacy. Uncertainty surrounds education data, but there
appears to be no or little intergenerational improvement in the
national literacy rate.

In a 2000 survey of governance in the Bangladesh education sector,
Transparency International noted "major irregularities." Students were
required to make unauthorized payments for admission into the school
(such as fees to sit exams). Distribution of food via the Food for
Education Programme was influenced by bribes and intervention of
locally influential people. (This program has now been replaced with
a cash payment intended for low-income families. It, too, is subject
to serious administrative problems.) Focus groups reported arbitrary
behaviour by many regional education officers, including bribe-taking
(cited in Ahmed & Nath 2005, 118-119). Low quality of in-school
teaching and lack of adequate school facilities have induced wide
resort to private tutoring. In a detailed survey of schools in 10
representative regions across the country, CAMPE round that 43 percent
of children were using private tutors. Often, the tutors are the
students' own teachers who thereby supplement very low salaries.
CAMPE lamented the extent of "political control" at the local
level--entailing "cronyism of head teachers" and the direct
intervention of the local Member of Parliament (Ahmed & Nath
2005, xxxii, xxxvi). In assessing school management, CAMPE concluded:

   Governance problems, including corruption and mismanagement related
   to all aspects of education management, seriously undermined
   management of primary education for quality and equity. The problems
   concerned recruitment and posting of teachers, construction and
   repair of school buildings, charging unofficial fees and the common
   practice of private tutoring, "renting out" teaching jobs,
   administration of stipends, distribution of textbooks, and cheating
   in examinations. (Ahmed & Nath 2005, 137.)

CAMPE has called for major pilot projects that decentralize authority
over education decisions to regional authorities below the national
government. It has called for "a concerted effort ... to implement
competency-based curriculum, classroom work, and learning assessment."
A meaningful voice for parents and teachers is needed. Hence, the call
for creation of parent-teacher associations, for "periodic sharing of
information and plans, and monitoring of progress," and "transparent
and public selection/election of school managing committee and
[regional] education committee members" (Ahmed & Nath 2005, 140,142).
In conclusion, CAMPE warns:

   Political interference and undue involvement of politicians,
   institutionalized by government regulations about [school] managing
   committees, have been identified as a major contribution to
   corruption, mismanagement, waste and obstacle to good management
   practices in general. (Ahmed & Nath 2005, 142.)

The threat of electoral defeat may bring about reasonably effective government. But democracy is not a panacea. There are many means whereby politicians can pervert electoral constraints. In poor countries, there is an understandable premium among the majority on here-and-now benefits, and politicians often win elections by diversion of public funds to inefficient short-term redistribution. A relevant example is provision of electricity from state-owned power companies to targeted groups of voters at rates far below cost. Such tactics win elections, but the result is diversion of public revenues to subsidize financially crippled state-owned enterprises that supply an inferior product subject to voltage variation and random daily "load shedding." (5) In other instances, political parties use aggressively non-cooperative strategies against one another. The result is inefficient political equilibria characteristic of prisoners' dilemma games.

On occasion, "benign dictators" have served to assure acceptable government performance. The reforms initiated by Deng Xiao Ping, who consolidated power following Mao's death, were a catalyst for China's phenomenal economic progress in the last three decades. As measured by literacy and life expectancy, or by average per capita income, the authoritarian regime in China has, since the death of Mao, outperformed the democratically elected governments of India. Neither the western nor eastern portions of what was, pre-1971, the state of Pakistan has prospered--both are poorer than India. But the military dictatorship of Pakistan is slightly more prosperous than the notionally democratic state of Bangladesh. (6) However, dictatorships are often not benign. At their worst, they simultaneously abuse basic civil liberties and degenerate into inefficient kleptocracies, such as the military regime in Myanmar.

How aggressive should be donor interventions in the domestic policies of aid recipients? The answer is, it depends. This answer is deeply unsatisfactory to anyone seeking certainty. Some countries are well governed and provision of aid generates valuable benefits. On the other hand, the quality of many governments in sub-Saharan Africa is so dubious and the severity of the AIDS epidemic so great, that to meet the millennium development goals probably requires major breaches of national sovereignty. Some multilateral agency operating under UN sanction could, for example, assume responsibility for regional health programming. Such intervention is a necessary, hardly sufficient, condition for meeting the goals--and, to state the obvious, is highly unlikely in an age of postcolonial nationalism.

Following World War II, European overseas empires collapsed--a well-deserved collapse in most instances. Subsequently, the principle of national sovereignty has reigned supreme. So long as aid policy is cast in terms of quantity of aid, the underlying rationale is redistribution. An emphasis on quality of aid implies a legitimate if circumscribed role for donor countries in the domestic politics of aid recipients. The reforms proposed by multilateral agencies, such as the World Bank and International Monetary Fund, are usually--not always, but usually--based on reasonable, efficiency-enhancing policy ideas. The proposed reforms inevitably entail lost income and power among particular interest groups privileged by the status quo. Predictably, the losers object, frequently condemning multilateral agencies as instruments of neocolonialism.

Among prominent Canadians to have acknowledged clearly the dilemma of governance and development is Michael Ignatieff. He has written about an "ongoing crisis of order" in many post-colonial developing nations:

   For thirty years after decolonisation this crisis has grown upon
   us, and all the rich Western world has done is to pretend if is hot
   occurring. First, we believed in the theology of development, only
   to see development founder on corruption and the incapacity of weak
   state structures to develop honest government and equitable
   programmes for growth. Then we told ourselves globalisation
   itself--capitalism's sheer voracious dynamism--would bring
   prosperity and order in its wake. But markets alone cannot create
   order: markets require order if they are to function efficiently,
   and the only reliable provider of order--law, procedure, safety and
   security--is the state. A globalised economy cannot function without
   this structure of authority and coercive power, and where it breaks
   down markets break down ... large corporations will not patrol the
   street corners. They will not provide the schools, roads and
   hospitals that distinguish society from the jungle. (Ignatieff 2003,
   124.)

A group at the World Bank has pursued most rigorously the definition of governance, its measurement, and its relation to economic development. Its members have constructed indices of governance by aggregating results from a large number of surveys on comparative perceptions. Since 1996, they have compared quality of governance across countries based on six broad dimensions:

1. Voice and Accountability--measuring political, civil and human rights.

2. Political Instability and Violence--measuring the likelihood of violent threats to, or changes in, government, including terrorism.

3. Government Effectiveness--measuring the competence of the bureaucracy and the quality of public service delivery.

4. Regulatory Burden--measuring the incidence of market-unfriendly policies.

5. Rule of Law--measuring the quality of contract enforcement, the police, and the courts, as well as the likelihood of crime and violence.

6. Control of Corruption--measuring the exercise of public power for private gain, including both petty and grand corruption and state capture (Kaufmann et al. 2005, 4).

Their most recent indicators, for the year 2004, compare quality of governance across more than 200 countries. Figure 1 illustrates index values for one dimension, government effectiveness, among 209 countries, ranked from left to right in ascending order. (The six indices are normalized with mean 0, and standard deviation 1.) They measure the quality of governance of one country relative to others, not relative to any absolute benchmark. Such indices are obviously subject to a great deal of imprecision, both conceptual and statistical. On the assumption that the surveys utilized offer meaningful measures of the six governance dimensions, the precision of any country's index value rises with the number of available surveys and their consistency. The bar surrounding each country's index value is the relevant confidence interval. (7)

[FIGURE 1 OMITTED]

In Figure 1 the index values for the 25 CIDA-targeted countries are highlighted. As measured, government effectiveness among these countries ranges from a low of -0.96 (for Ethiopia) to a high of -0.13 (for Senegal). None of the 25 has an index value above zero. This implies that over one-half the world's governments are more effective than Senegal's; about one-sixth of the countries measured have governments less effective than Ethiopia's.

It may seem dubious to measure subjective perceptions of governance. The rationale for doing so is that objective measures of concepts such as the extent to which one country controls corruption better than another are inherently impossible to establish. Better to measure imperfectly that which is important than to measure with great precision that which is not.

What tan aid accomplish?

Amartya Sen, a Nobel laureate in economics, describes a division running through development policy analyses. "One view sees development as a 'tierce' process," he suggests, "with much 'blood, sweat and tears' ... This hard-knocks attitude contrasts with an alternative outlook that sees development as essentially a 'friendly' process ... as exemplified by such things as mutually beneficial exchanges ... or by the working of social safety nets, or of political liberties, or of social development" (Sen 1999, 35-36).

Sen places himself firmly on the "friendly" side of this division. Undeniably, most who engage in development policy, whether as politicians and officials in donor countries or as on-the-ground practitioners in host countries, do so with the intention of contributing to a "friendly" humanitarian exercise. But there is much to be said for the "fierce" view: acknowledging the tenacity of perverse dynamics at work on host governments and how tenuous are the links between aid offered and tangible benefits. In point form, here are four fundamental difficulties:

* Lack of local information: Donor agencies often possess an entirely inadequate understanding of political dynamics in the host country, and consequently propose inherently inefficient aid programs.

* Fungibility of donor aid: Donor aid may wind up displacing host-government spending on the targeted sectors. The donor may seek to improve, say, school completion rates. The host government may decide to rely on aid to finance schools, and divert own-source resources elsewhere.

* Provision of subsidy to rent-seeking elites: In the case of aid-receiving host countries suffering poor governance, where the host government has been effectively captured by inefficient rent-seeking elites, access to donor aid may be counterproductive. The aid may augment the ability of these elites to distort activities in both the public and private sectors.

* Legacy of the past: Citizens of the host country may balk at the policy goals of the donor agencies. Rightly or wrongly, donor aid may be interpreted as means whereby former--or future--imperial powers attempt to wield undue political influence.

In a much-cited study, Assessing Aid, the World Bank (1998) concluded that the effectiveness of aid in promoting economic growth depends crucially on the quality of governance of the host country. The study attempted to explain growth in per capita GNP in 56 developing countries over two decades, from the early 1970s to the early 1990s. The study included a number of standard variables, among them the ratio of aid to GDP; it also incorporated an "index of economic management." The key result was the interaction of quality of governance, as measured by this index, with the ratio of aid to GDP. In those countries where the index was high, incremental aid contributed to increased rates of growth of per capita GNP. Where the index was low, more aid actually retarded economic growth.

Figure 2a offers evidence on the role of governance in explaining economic growth among the CIDA-targeted countries. It plots average annual rate of growth in per capita GDP, from 1990 to 2003, for each of the 25 countries against the country's index of government effectiveness. There is an obvious positive correlation between index values and growth rates, and evidence that, for any given quality of governance, the African countries are faring less well than those in Asia or Americas.

[FIGURE 2 OMITTED]

A series of simple regressions suggests that the results obtained by the World Bank in Assessing Aid also apply to the 25 CIDA-targeted countries. (See Table 2.)

A useful intuition in assessing the role of governance is the image of cycles, either "virtuous" or "vicious." China is obviously not a democracy, and corruption remains a serious problem. However, China has experienced, since the 1970s, a virtuous cycle in which somewhat better governance has led to better economic performance that, in turn, has created a demand for even better government performance. Conversely, Ukraine illustrates the problem of a vicious cycle in which poor governance contributed to a decline in economic performance that exacerbated social tensions between Russian- and Ukrainian-speaking regions; tensions exploited by corrupt elites who preserved political power by tactical applications of patronage and coercion. The "orange revolution" of 2004 may mark a break in this cycle, although post-election conflict among the reform leaders indicates the difficulty of introducing new ways of conducting politics.

Let us allow that, for Ukraine, the disruptions occasioned by collapse of the Soviet empire, and for sub-Saharan African countries, the combination of regional wars and the AIDS epidemic rendered growth prospects particularly grim. Regression (1) attempts to explain annual per capita GDP growth between 1990 and 2003 among these 25 countries in terms of these events. Having made this allowance, what effect has official development aid (ODA measured relative to national GDP) had on economic growth? Regression (2) gives a positive coefficient to the aid variable, but it is small and statistically insignificant. On the other hand, quality of governance matters. The 2004 World Bank government effectiveness index generates a positive and statistically significant coefficient in regression (3). (8)

Regressions (4) and (5) allow for the interaction of governance and aid. Ranking countries in terms of the government effectiveness index permits a simple division between the "more effectively" and the "less effectively" governed. (See notes to Table 2 for allocation of the countries.) Among the former, aid has the desired effect: an increase of ODA/GDP by 10 percentage points implies an increase of 0.76 percentage point in average annual per capita GDP growth. Among the latter, increasing aid has a negative, but statistically insignificant, effect on per capita GDP. An alternate specification in regression (5) is that, among the "less effectively" governed, aid has no effect on economic growth. Under this specification, an increase of ODA/GDP by 10 percentage points among the "more effectively" governed increases average annual per capita GDP growth by 0.96 percentage point.

Figure 2b affords a visual summary. It plots average annual per capita GDP growth between 1990 and 2003 against generosity of aid. As illustrated, growth rates include a regional compensation for Ukraine and sub-Saharan countries. (9) Superimposed are the fitted regression lines from regression (4). (The line among the "less effectively" governed is dotted as reminder that the negative impact of aid is hot statistically significant.)

Countries experiencing poor governance often settle into a stable equilibrium of reciprocal mistrust and bribery. Parents adjust to absentee teachers by paying for private tutors (who may be absentee teachers); citizens mistrust the police; plaintiffs mistrust judges' impartiality; lenders mistrust borrowers' intent to repay loans; utility customers pay bribes to receive service; one businessman instinctively mistrusts the intentions of the other, and so on. To do otherwise is to invite others of take opportunistic advantage of your naivete and/or to be denied service. In such circumstances, poor governance becomes equivalent to low social trust. A poorly governed country may sustain its political equilibrium for generations--and may nonetheless realize modest economic growth. But living and working within such a society is debilitating for all concerned. Vast amounts of time and financial resources are dissipated in "working the system" as opposed to producing the goods and services of intrinsic value to the country's population.

Some analysts doubt that aid can contribute to economic growth, even when it takes place in the context of reasonably good governance. William Easterly (2003) finds essentially no positive statistical relation between aid and increased per capita incomes. Easterly's pessimism may be excessive but, as a generalization, CIDA has historically tilted very far to the "friendly" side of Sen's division. The recent International Policy Statement is a welcome, if modest, shift toward realism. Another acknowledgment of the "fierce" intractable problems of undertaking development aid is to be found in CIDA's journal (Journal of Development Policy and Practice). The author of the first issue's lead article is Daniel Kaufmann (2004), a major contributor to the World Bank's governance indicators. Ten years ago, CIDA would not have afforded prominence to his ideas. The editorial statement for the journal admitted as much:

   We inaugurate our ... journal with a politically sensitive and
   contentious subject of international dimensions: corruption.
   Banished from all development discourse and denied recognition
   by governments and policy makers world-wide until recently, today
   corruption is centre stage to the development discourse. Never
   before has the development community witnessed such a surge in
   public recognition or determination to confront corruption, or seen
   such substantial theorizing, or produced such a collection of
   alternative approaches, interpretations and typologies on the
   subject. (Shanker 2004, v.)

Recommendations

A group at Harvard have conducted an exhaustive survey of incidents of sustained spurts in economic growth in developing countries over the last halfcentury. They have insisted on the heterogeneity of policies prevailing in countries at the time of their growth spurt and are consequently reluctant to draw policy recommendations. Instead, they have retreated to a list of general "policy objectives." (See Box 4.)

Box 4: The importance of small changes in the right direction

A group at Harvard University (Hausmann et al. 2004, Rodrik 2004)
conducted an ambitious exercise, in which they sought all instances
of sustained acceleration in real GDP growth across a large sample of
developing countries over the second half of the 20th century.
Candidates for inclusion met the following three criteria: 1) the
average annual rate of growth in per capita GDP exceeded 3.5 percent
for at least 8 years; 2) the average growth rate over the 8 years
following the start date exceeded the average growth rate over the
preceding 8 years by at least 2 percentage points; 3) GDP at the end
of 8 years exceeded the country's previous peak GDP.

   They concluded their analysis with "two main surprises:"

   First, growth accelerations are a fair]y frequent occurrence. Of
   the 110 countries included in the sample, 60 have had at least one
   acceleration in the 35-year period between 1957 and 1992--a ratio
   of 55 percent ... This is a useful antidote to the pessimism that
   often pervades policy discussions on growth.

   Second, and not unrelated to the previous finding, most growth
   accelerations are not preceded or accompanied by major changes in
   economic policies ]such as financial market liberalization],
   institutional arrangements [adoption of new constitution],
   political circumstances [such as change of regime either away from
   or toward democracy*], or external conditions [such as favourable
   terms of trade shocks]. As we have shown, standard growth
   determinants have some statistical leverage over the timing of
   accelerations. But on the whole those determinants do a very poor
   job of predicting the turning points. It would appear that growth
   accelerations are caused predominantly by idiosyncratic, and often
   small-scale, changes. (Hausmann et al. 2004, 21-22.)

In making policy recommendations for developing countries, the group
retreated from advocacy of specific policies to advocacy of what they
labeled policy objectives:

   One such objective is the maintenance of macroeconomic stability.
   All high growth countries pursue "responsible" monetary and fiscal
   policies that prevent high inflation and the buildup of
   unsustainable debt levels. Another is the desire to integrate in
   the world economy. While China, Vietnam, India and most other
   high-growth economies have had high levels of protection against
   imports, and often have remained outside the formal rules of the
   world trade regime (i.e., the GATT/WTO system), they have also
   found ways to spur exports and attract direct foreign investment. A
   third objective is to provide investors with effective protection
   in terms of property rights and contract enforcement. Without such
   protection, firms and entrepreneurs do not have the incentive to
   accumulate capital and improve productivity. A fourth objective is
   to maintain a certain degree of social cohesion, solidarity, and
   political stability. Without social and political peace, the
   economy cannot perform adequately.

   One can perhaps list a few more of these higher-order
   principles--such as an appropriate environment for productive
   diversification and innovation, social insurance and safety nets,
   prudential regulation of financial intermediaries, appropriate
   management of the exchange rate and of the capital account, and
   so on. What is relevant and very important is that these can on]y
   be stated in terms of broad objectives. There is no unique mapping
   between these objectives and specific policy proposals.
   (Rodrik 2004, 3.)

Like many, Rodrik and colleagues draw lessons from China's development
following Mao's death. It is by far the most impressive development
success of the last quarter century. Deng Xiao Ping and his allies did
not undertake wholesale dismantling of low-productivity state-owned
firms. Nor did they introduce a private property regime based on
traditions of the common law. To attract foreign investment, they relied
extensively on private-public partnerships. In effect, they
superimposed progressively larger elements of market institutions on
the existing foundation of communist legal and political arrangements.
Their initial choice of agriculture for marketization was astute:
higher agricultural productivity in the 1980s generated widely
dispersed benefits, in the form of higher peasant incomes and
increased quality and variety of food across the country. The broad
distribution of these benefits established popular legitimacy for
further, more controversial market reforms that carne later.

* While both a change of regime toward democracy and toward a more
authoritarian regime had a significantly positive probability of
leading to accelerated growth, a shift toward an authoritarian regime
was three times more like]y than a shift toward democracy to be
associated with a growth spurt.

The recommendations below are also of a decidedly general nature. (In a forthcoming Commentary Danielle Goldfarb and Stephen Tapp compare development agencies across a sample of donor countries, and draw therefrom more precise recommendations for Canadian development aid policy.)

Recommendation One

Design bilateral aid programs with due attention to host-country governance. Continue to devote a large share of aid to multilateral agencies with a strong analytic capacity and a willingness to apply conditionality in the supply of aid to host governments.

In a recent study measuring the sensitivity of donors' aid distribution to the quality of host-country institutions over the years 1999-2002, Canada's score was mediocre. Its average ranking over the four years was 19 out of a total of 41 donor agencies (Dollar & Levin 2004, 18). Since there is an inverse correlation between a country's poverty and quality of government--the poorer the country the worse, in general, the quality of government institutions--a potential defence of Canada's aid strategy is that CIDA disproportionately targets very poor countries. But on a related index of sensitivity to host-country poverty, Canada's ranking was also mediocre: an average rank of 22 out of 41 over the four years.

On an overall index, incorporating sensitivity to both quality of government institutions and poverty of aid recipients, Canada's ranking over the four years was stable; the average rank was 20 out of 41. A number of bilateral agencies, notably Denmark, Sweden, and United Kingdom, significantly improved their performance over the four years. The agencies that scored highest on this overall index for 2002, the most recent year analyzed, were the World Bank's International Development Association (first rank), Denmark (second), the International Monetary Fund's Enhanced Structural Adjustment Facility (third), United Kingdom (fourth), and Sweden (fifth). (10)

At a minimum, results showing an inverse relation between aid and economic growth in poorly governed countries reinforce the theme that Canada design bilateral aid programs with due attention to the quality of governance in the host country. Inevitably, this implies a tension between Canada as donor and governments of aid-receiving countries where performance leaves much to be desired. Due attention may take the form of strategizing to minimize the potential of corruption on proposed projects. It may take the form of selecting projects whose primary purpose is not financing the delivery of a particular service but improving quality of governance. For example, donor projects that aim to restore decent management of the Bangladesh public school system might well yield better results than projects that simply increase donor funding for education. (Refer back to Box 3.) Due attention may be more tough-minded. It may entail tactical withholding of aid in an attempt to influence host-country behaviour, or an explicit refusal of developmental aid to countries with especially low governance rankings. (11) Even Jeffrey Sachs (2005, 168-169), an ardent advocate of increased donor aid, acknowledges that "poor countries have no guaranteed right ... to receive development assistance from the rich countries. They only have that right if they themselves carry through on their commitments to good governance."

CIDA currently spends two-fifths of its development aid via multilateral agencies. Given the superior analytic capacity of these agencies, it makes sense for Canada to continue delivering a sizeable fraction of aid this way. Even in the context of bilateral aid, informal multilateralism can arise. The leading donor countries in a particular country on occasion form consortia and propose joint projects to the host government. Such projects are more likely to enjoy stable medium-term financing. Relative to individual donors, a consortium of donors will likely have more ability to influence host-country governance, improve initial project design, and assure outcome evaluation.

Recommendation Two

Concentrate bilateral aid in a few countries only, and devote resources to analyzing the economies and politics of these countries.

This is an endorsement of what the International Policy Statement has--in a limited way--proposed. In recent years, CIDA has organized projects in over 100 countries (Canada 2005d). Delivering aid in so many countries, CIDA has never been able to evaluate the effectiveness of aid in individual countries. By 2010, the statement promises, at least two-thirds of bilateral aid will be devoted to 25 countries. Other countries will still be eligible for humanitarian aid. This is an exceedingly gradual shift in CIDA spending priorities. As a share of bilateral aid expenditures, the share for the targeted 25 should rise more quickly and, arguably, the list should be further pared.

In their forthcoming Commentary, Danielle Goldfarb and Steven Tapp make the point that, having historically spread aid over many countries, Canada's aid is a small fraction of aid received by any particular country. As a donor offering relatively little to any host country, Canada has relatively little diplomatic weight in influencing host-country decision making.

No doubt, an element of arbitrariness entered into CIDA's selection of targeted countries. The statement defines the rationale as having been "the poorest countries, irrespective of their size, where effective programming to address the [Millennium Development Goals] is possible and where Canada can add real value" (Canada 2005a, 22). If the criterion is the poorest, why include Ukraine? If the criterion is a Canadian comparative advantage relative to other donor countries, why exclude Haiti, a desperately poor francophone country close by in the Western Hemisphere, a country to which Canada has important links via past aid projects and the size of its Haitian immigrant population? Admittedly, a reason to exclude Haiti is its very low quality of governance rankings.

Having designated target countries, CIDA has an obligation to study realistically their domestic economies and politics, and engage Canadians beyond government in the analysis. What are the goals of country-specific CIDA programs? What evaluation of outcomes will take place? What do we know about the quality of host-country leaders and institutions?

Very few Canadians--whether in CIDA, the diplomatic service, university faculties, or NGOs--have a reasonable understanding of the targeted countries. What knowledge is available tends to be an oral tradition among seasoned aid workers and diplomats, is buried in confidential government reports, or is contained within the relevant immigrant community. One vehicle for rectifying our collective Canadian ignorance is the newly created Canada Corps. (See Box 2.)

In late 2004, CIDA launched the Journal of Development Policy with the following mandate:

   to provide a platform for dialogue between and among development
   scholars, policy makers, practitioners and the academic community
   on important contemporary issues in international development. [The
   journal] aims to publish a wide range of views and innovative
   perspectives and facilitate understanding of emerging issues in the
   field of international development. (Canada 2004.)

These are fine words, but the journal has become a symbol of CIDA's inability to sustain a dialogue on development policy. The first issue was the last; the project has been suspended.

Recommendation Three

Target aid expenditures on education and basic health services, sectors subject to complex "market failures. "Again, be prepared to impose conditionality

These are sectors in which private market supply is rarely a satisfactory substitute to provision by the state or by relevant NGOs, and improved quality of host-government institutions is, in general, a prerequisite for better outcomes. One tactic in the context of weak host-country governance is to deliver aid via NGOs. But NGOs too can be infected with problems of poor governance and corruption, and such a strategy may weaken the donors' ability to influence the host government.

By contrast, Canada's quality of administration of health and education services is high on a comparative basis of countries. Furthermore, Canada has long experience with delivery of these services by regional (provincial) governments. Many developing countries suffer from acute over-centralization and consequent impotence of sub-national governments. To improve education and health outcomes, they need to improve the quality of regional and local government. CIDA could make better use than it has to date of the managerial expertise among Canadian provinces in delivering health and education services.

Fourteen of the 25 CIDA-targeted countries are in sub-Saharan Africa. The difficulties of delivering reasonable quality health and education services in the region are monumental. The latest major report to document comprehensively the problems is that of the UK-led Commission for Africa. It acknowledges that "[w]ithout progress in governance, all other reforms will have limited impact" (United Kingdom 2005,14). As did Jean Chretien in 2002, Tony Blair used his presidency of the G-8 in 2005 to concentrate attention on the African continent. The need for additional development funds is obvious, but to be effective they must be accompanied with intrusions in host-country governance that, because of the scope envisaged, raise the ghost of 19th-century colonialism. Canada should contribute financially to such multilateral initiatives, but will do so, hopefully, with a political realism often lacking in the past.

Recommendation Four

Canada should lower the barriers it imposes to efficient developing-country policies.

In recent years, the Center for Global Development (Roodman 2005) and the journal Foreign Policy (2005) have published a "commitment to development index" ranking the relative performance of developed countries across several relevant dimensions--aid/GDP (adjusting down for tied aid), trade barriers against typical developing-country exports, requirement that developed country firms respect host-country environmental concerns and refrain from bribery, and so on. (In the most recent ranking, Canada tied--with UK--for 10th position among 21 donor countries. The top-ranked country was Denmark; second through fourth positions were occupied respectively by Netherlands, Sweden, and Australia.) The value of the index is less its overall ranking of countries than its aggregation of data on the myriad ways whereby developed countries can enhance or limit the returns to efficient developing-country policies.

One such barrier is high tariffs against typical developing-country exports. At early stages, their comparative advantage is often in labour-intensive manufacturing goods (processed foods, textiles, and so on), sectors subject to high levels of protection in many developed countries. (12) Lowering this protection is politically contentious within donor countries because it endangers low-value-added manufacturing employment. Proponents of the status quo in, say, the textile sector object to importing tee-shirts from Bangladesh made under "sweat shop" conditions. Sweat shop may be an accurate description of working conditions, but Bangladesh garment workers, the great majority of them women, earn more than their counterparts who stay in the village, and garments generate three-quarters of the country's foreign exchange. The general point is that export-oriented sectors usually pay above-average wages, encourage productive entrepreneurship, and provide the foreign exchange to import usable technology.

Quite reasonably, in comparing aid/GDP across donor countries, the index discounts the value of aid tied to purchases in the donor country. Tied aid raises the cost of host-government imports and muddles the rationale for providing aid. Historically, Canada has been among the donor countries with a high ratio of aid tied. CIDA has in recent years undertaken to reduce such practices (Canada 2002).

The index also attempts to measure quality of foreign investment in developing countries. Countries rank higher if they impose codes of conduct for their national companies investing abroad. The market for bribes obviously requires both briber and bribed, and developed country firms can stymie economic growth of a developing country by abetting local corruption. Reforms in this domain are hard to monitor, but developed countries are now legislating codes of conduct for national companies operating abroad.

Conclusion

In guise of conclusion, I offer a sketch--I hesitate to claim Figure 3 as anything more--to help in thinking about development aid. The intermediate variable that deserves central attention is quality of government in the host country. In general, the performance of the host government is crucial in analyzing improvements to the three other intermediate variables: education services, health services, and the legal/regulatory framework. Via its effect on these latter three variables, quality of governance has a large impact on per capita GDP and poverty reduction. It also has an effect via other variables not here illustrated. (The weight of the various arrows implies my assessment of relative importance.)

[FIGURE 3 OMITTED]

As set out in the accompanying sketch, the variables that matter are the following:

* "History": By "history" is meant the heritage of ethnic differences and culture surrounding the conduct of business and politics. A country with "good" history is more likely to experience good governance; a country that has experienced, in recent times, civil war or other episodes of widespread violence, or is fractured along tribal, religious, or ethnic lines will probably experience weak governance.

* Geography: With the dramatic exception of Singapore, there are few examples of prosperous industrial economies close to the equator. It is inherently more difficult in tropical conditions to assure the basic public health services required to sustain a productive healthy population. This has become a fortiori true of sub-Saharan Africa, given the curse of HIV/AIDS.

* Random events: A new regime may significantly improve matters (for example, Museveni's assumption of power in Uganda). A new regime may do the reverse (for example, Ukraine's post-1990 regimes).

* Development aid: The final independent variable is aid. Its impact is likely to be modest. It may even be counter-productive.

The analysis and--admittedly very general--recommendations of this Commentary introduce the painfully complex questions as to why the poor are poor and what we who are not might do about it. There are no readily applicable rules of thumb for an effective development aid policy. Success requires realistic country-by-country assessment and a willingness to persevere.

I do not want to close with a dirge. The message from Rodrik's colleagues (recall Box 4) is the value of host countries making incremental policy decisions pointing in the right direction. If Canada targets a few countries, learns about them, exploits our comparative advantage in delivering health and education services, we can help improve the quality of life for millions of people beyond our borders.

Appendix A: A Survey of the 25 CIDA-targeted Countries

Figure 1 highlights the World Bank government effectiveness index of the 25 developing countries that CIDA has designated as targets for future bilateral Canadian aid. Table A1 gives more detail: the country rank, for each of the 25, on each dimension of governance as measured by the 2004 World Bank governance indicators. The table also gives the respective percentile changes in rank since 1996, where the changes are deemed statistically significant. For example, Tanzania probably has improved performance (relative to other countries) in "voice and accountability," "government effectiveness," and "control of corruption." On the other hand, Bangladesh has almost certainly experienced deterioration in "political stability and violence," "quality of government regulation," and "control of corruption."

Ukraine is the sole European country targeted by CIDA. Per capita GDP in 2003 was only 56 percent of the peak, realized in 1989. Ukrainian economic decline since independence is an extreme illustration of the disappointing performance in many newly independent countries of the former Soviet empire.

Among the 14 targeted sub-Saharan countries, average government effectiveness, as measured, is virtually identical to the average for the 10 countries in Asia and the Americas. Hence, to the extent government effectiveness determines growth, there should be little difference between the two sets of countries. Sub-Saharan Africa has experienced civil and regional wars and coups d'etat, but so too have CIDA-targeted countries elsewhere. The average for the political stability index in 2004 among the 14 sub-Saharan countries is higher than that among the 11 targeted countries elsewhere. For any given level of government effectiveness, however, regression (3) in Table 2 implies the typical sub-Saharan country has an average per capita growth rate 1.28 percentage points lower than a targeted country in Asia or Americas. Plausible interpretation of this negative effect of geography is the damage wrought by tropical diseases. AIDS is the most dramatic, but certainly not the only disease to afflict Africans disproportionately.

A more comprehensive measure than per capita income of a country's development is the Human Development Index (HDI) published by the UN. The HDI is a measure derived from three elements: the country's per capita income, education performance, and life expectancy. (13) The 2003 HDI values for the targeted countries are shown in Figure A1. The trend lines indicate an obvious positive relationship between HDI and government effectiveness as measured, albeit a much weaker trend among the sub-Saharan African countries than among those in Asia and Americas.

[FIGURE A1 OMITTED]

Ghana deserves special mention. It is among the best-governed of African countries. It has average life expectancy of nearly 60 years, literacy above 70 percent, per capita income above US$2000, and the highest HDI ranking among the targeted African countries. Among the targeted countries in Asia and the Americas, Sri Lanka has the highest HDI ranking--despite enduring a civil war over the previous two decades. Guyana is a close runner-up.

Figures A2--A5 illustrate the components entering into calculation of the HDI values for each country. In Figure A2, trend lines imply that a one standard deviation increase in the index of government effectiveness (from, say, -1 to 0) increases per capita GDP by about US$900 among the African countries and by US$4200 among those in Asia and Americas. As illustrated in Figures A3 and A4, government effectiveness appears to have a positive impact on literacy and enrolment rates among the targeted countries in Asia and Americas. However, the scatterplots show a purely random distribution for the African countries (with statistically insignificant negative trend lines). In the case of life expectancy, government effectiveness matters a great deal--both among the African countries and those elsewhere. A one standard deviation improvement in the government effectiveness index implies an increase in average life expectancy at birth of roughly 12 years among countries in Africa, Asia and Americas. A reasonable interpretation of this result is the crucial importance of governance in establishing effective public health policies. Holding constant the level of government effectiveness, the devastating effect of AIDS and other regionally prevalent diseases has been to lower African life expectancy by 18 years. Among the African countries, two have life expectancy below 40 years; another nine have life expectancy between 40 and 50; only three have life expectancy above 50.

[FIGURES A1-A5 OMITTED]

Table A1: Governance Indicators of CIDA-Targeted Countries
(rank in 2004 and significant changes in index values, 1996-2004)

Country                 Code      Voice and Accountability

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN        93
Burkina Faso            BFA       131
Cameroon                CMR       176
Ethiopia                ETH       170
Ghana                   GHA        88           16.9
Kenya                   KEN       126
Malawi                  MWI       139
Mali                    MLI        91
Mozambique              MOZ       118
Niger                   NER       116
Rwanada                 RWA       168
Senegal                 SEN       101
Tanzania                TZA       127           13.7
Zambia                  ZMB       129

Asia

Bangladesh              BGD       148
Cambodia                KHM       156
Indonesia               IDN       133           19.7
Pakistan                PAK       183
Sri Lanka               LKA       122
Vietman                 VNM       190

Americas

Bolivia                 BOL       110
Guyana                  GUY        71
Honduras                HND       112
Nicaragua               NIC       107

Europe

Ukraine                 UKR       142

Number of countries
ranked                            207

Country                 Code      Political Stability

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN       134          -54.8
Burkina Faso            BFA       132
Cameroon                CMR       163
Ethiopia                ETH       174
Ghana                   GHA       112
Kenya                   KEN       171
Malawi                  MWI       133
Mali                    MLI       104
Mozambique              MOZ       121
Niger                   NER       141
Rwanada                 RWA       167
Senegal                 SEN       123
Tanzania                TZA       135
Zambia                  ZMB       122

Asia

Bangladesh              BGD       183          -15.2
Cambodia                KHM       144
Indonesia               IDN       188          -18.8
Pakistan                PAK       194
Sri Lanka               LKA       178           8.0
Vietman                 VNM       100

Americas

Bolivia                 BOL       148
Guyana                  GUY       139
Honduras                HND       152
Nicaragua               NIC       119

Europe

Ukraine                 UKR       129

Number of countries
ranked                            207

Country                 Code      Government Effectiveness

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN       126
Burkina Faso            BFA       136
Cameroon                CMR       148
Ethiopia                ETH       175
Ghana                   GHA       101
Kenya                   KEN       162
Malawi                  MWI       161
Mali                    MLI       115
Mozambique              MOZ       127
Niger                   NER       169
Rwanada                 RWA       139
Senegal                 SEN        97
Tanzania                TZA       125           32.0
Zambia                  ZMB       166

Asia

Bangladesh              BGD       154
Cambodia                KHM       168
Indonesia               IDN       124          -26.7
Pakistan                PAK       140
Sri Lanka               LKA       114
Vietman                 VNM       116

Americas

Bolivia                 BOL       147
Guyana                  GUY       106
Honduras                HND       151
Nicaragua               NIC       152

Europe

Ukraine                 UKR       150

Number of countries
ranked                            209

Country                 Code         Regulatory Quality

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN       141
Burkina Faso            BFA       118
Cameroon                CMR       158
Ethiopia                ETH       180
Ghana                   GHA       120
Kenya                   KEN       131
Malawi                  MWI       146
Mali                    MLI       117
Mozambique              MOZ       121
Niger                   NER       154
Rwanada                 RWA       128
Senegal                 SEN       122
Tanzania                TZA       144
Zambia                  ZMB       140          -32.5

Asia

Bangladesh              BGD       177          -13.2
Cambodia                KHM       114
Indonesia               IDN       129          -28.8
Pakistan                PAK       172
Sri Lanka               LKA        82
Vietman                 VNM       148

Americas

Bolivia                 BOL        95          -27.6
Guyana                  GUY       108
Honduras                HND       124
Nicaragua               NIC       110

Europe

Ukraine                 UKR       136

Number of countries
ranked                            204

Country                 Code           Rule of Law

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN       123
Burkina Faso            BFA       140
Cameroon                CMR       174
Ethiopia                ETH       173          -28.8
Ghana                   GHA       106
Kenya                   KEN       172
Malawi                  MWI       114
Mali                    MLI       120
Mozambique              MOZ       136           27.5
Niger                   NER       166
Rwanada                 RWA       164
Senegal                 SEN       110
Tanzania                TZA       126
Zambia                  ZMB       129

Asia

Bangladesh              BGD       162
Cambodia                KHM       170
Indonesia               IDN       165          -19.0
Pakistan                PAK       154          -11.9
Sri Lanka               LKA        97
Vietman                 VNM       134

Americas

Bolivia                 BOL       131
Guyana                  GUY       125
Honduras                HND       137
Nicaragua               NIC       143

Europe

Ukraine                 UKR       160

Number of countries
ranked                            208

Country                 Code       Control of Corruption

                                             Change in
                                  Rank,     percentiles,
                                  2004       1996-2004

Africa

Benin                   BEN       109
Burkina Faso            BFA       112
Cameroon                CMR       151
Ethiopia                ETH       159
Ghana                   GHA        99
Kenya                   KEN       165
Malawi                  MWI       156
Mali                    MLI       126
Mozambique              MOZ       154
Niger                   NER       163
Rwanada                 RWA       114
Senegal                 SEN       116
Tanzania                TZA       131           26.6
Zambia                  ZMB       149

Asia

Bangladesh              BGD       183          -26.3
Cambodia                KHM       177
Indonesia               IDN       167          -17.1
Pakistan                PAK       162
Sri Lanka               LKA        97
Vietman                 VNM       146

Americas

Bolivia                 BOL       152
Guyana                  GUY       111
Honduras                HND       142
Nicaragua               NIC       110

Europe

Ukraine                 UKR       166

Number of countries
ranked                            204

Source: Author's calculations from interactive data from Kaufmann
et al. (2005)

Note: Changes in percentiles between 1996 and 2004 are reported
only where the country's relevant governance index value changed
in a statistically significant manner. Statistical significance is
calculated at the 10 percent level, on the simplifying assumption
that each statistic is an unbiased normally distributed estimator
of the underlying dimension of governance in the country, and that
the variables for 1996 and 2004 are independent. For example, the
"voice and accountability" index value for Ghana increased
significantly. Thus in turn raised Ghana's ranking among the
countries evaluated, from 114th (among 167 countries ranked in
1996) to 88th (among 207 countries ranked in 2004). This improved
the percentile position of Ghana by 16.9 percentage points (from
40.8 to 57.7).

Note: No "control of corruption" index values were calculated for
Senegal and Rwanda in 1996.

I have benefited in writing this Commentary from comments on earlier drafts and discussions with many people, some of whom probably prefer to remain unnamed. I thank explicitly Richard Bird, Dominic Cardy, Finn Poschmann, Danielle Goldfarb, Owen Lippert, Jack Mintz, Alimullah Miyan, Nancy Olewiler, Bill Robson, and Stephen Tapp.

References

Ahmed, M. and S. Nath. 2005. "Quality With Equity: The Primary Education Agenda." Education Watch Report 2003-04. Dhaka: Campaign for Popular Education, Bangladesh (CAMPE).

British Broadcasting Corporation (BBC). 2006. "Q&A: Uganda's Besigye cleared." Available on line at http://news.bbc.co.uk/2/hi/africa/4781528.stm

Bruns, B., A. Mingat, and R. Rakotomalala. 2003. A Chance for Every Child: Achieving Universal Primary Education by 2015. Washington: World Bank.

Canada. 2002. "Tied Aid." Canada making a difference in the world: a policy statement on strengthening aid effectiveness. Ottawa: Canadian International Development Agency. Available on line at http://www.acdi-cida.gc.ca/aideffectiveness#15

Canada. 2004. Journal of Development Practice and Policy. Available on line at http://www.acdicida.gc.ca/journal

Canada. 2005a. A Role of Pride and Influence in the World: Development. Canada's International Policy Statement, development section. Ottawa: Canadian International Development Agency. Available on line at http://www.acdi-cida.gc.ca/INET/IMAGES.NSF/vLUImages/IPS_PDF_EN/ $file/IPS-EN.pdf

Canada. 2005b. A Role of Pride and Influence in the World: Diplomacy. Canada's International Policy Statement, diplomacy section. Ottawa: Foreign Affairs and International Trade. Available on line at http://www.dfait-maeci.gc.ca/cip-pic/IPS/IPS-Diplomacy.pdf

Canada. 2005c. "CIDA announces new Development Partners: developing countries where Canada can make a difference." News release (19 April 2005).

Canada. 2005d. Statistical Report on Official Development Assistance--FY 2003-2004. Ottawa: Canadian International Development Agency. Available on line at http://www.acdicida.gc.ca/cida_ind.nsf/8949395286e4d3a58525641300568be1/ 79243df347894a9e85256acb005f1e01?OpenDocument

Canada. 2006. Afghanistan/Overview. Available on line at http://www.acdicida.gc.ca/CIDAWEB/webcountry.nsf/VLUDocEn/ Afghanistan-Overview

Central Intelligence Agency (CIA). 2006. "Uganda." The World Factbook. Available on line at http://www.cia.gov/cia/publications/factbook/geos/ug.html

Department for International Development (DFID). 2005. "UK cuts direct budget aid to Uganda by 15 million [pounds sterling], withholds further 5 million [pounds sterling]." Available on line at http://www.dfid.gov.uk/news/files/pressreleases/uganda-reduction.asp

Dollar, D. and V. Levin. 2004. "The Increasing Selectivity of Foreign Aid, 1984-2002." World Bank Policy Research Working Paper 3299. Available on line at http://wwwwds.wortdbank.org/servlet/WDS_IBank_Servlet?pcont=details& eid=000009486_20040615151147

Easterly, W. 2003. "Can Foreign Aid Buy Growth?" Journal of Economic Perspectives. 17:3:23-48.

Farooq, O. 2004. "Free electricity crisis looms." Available on line at http://news.bbc.co.uk/1/hi/world/south_asia/3601860.stm

Foreign Policy. 2005. "Ranking the Rich, 2005" (September/October). Available on line at http://www.foreignpolicy.com/story/cms.php?story_id=3221

Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: Free Press.

Geertz, C. 2003. "Which Way to Mecca?" New York Review of Books. Parts I and II. 50:10, 50:11 (12 June 2003, 3 July 2003).

Goldfarb, D. 2002. Trade as Aid: Freeing Access to Canada's Markets for the World's Poor. Backgrounder 60. Toronto: C.D. Howe Institute. Available on line at http://www.cdhowe.org

Goldfarb, D. and S. Tapp. Forthcoming. How Canada Can Improve Its Development Aid Effectiveness: Lessons from Other Aid Agencies. Commentary. Toronto: C.D. Howe Institute.

Habeck, M. 2005. "Reading 9/11." The American Interest. 1:1:101-108.

Hausmann, L. Pritchett, and D. Rodrik. 2004. "Growth Accelerations." Harvard University JFK School of Government Faculty Research Working Paper Series. Available on line at http://ksgnores1.harvard.edu/research/wpaper.nsf/rwp/RWP04-030? OpenDocument

Ignatieff, M. 2003. Empire Lite: Nation-Building in Bosnia, Kosovo and Afghanistan. Toronto: Penguin Canada.

Kaufmann, D. 2004. "Corruption Matters: Evidence-Based Challenge to Orthodoxy." Journal of Development Policy and Practice. 1:1:1-24. Available online at http://www.acdicida.gc.ca/journal

Kaufmann, D. and A. Kraay. 2002. "Growth without Governance." Economia 3:1:169-229. Available on line at http://www.worldbank.org/wbi/governance/pubs/growthgov.htm

Kaufmann, D., A Kraay and M. Mastruzzi. 2005. "Governance Matters IV: Governance Indicators for 1996-2004." Washington, D.C.: World Bank. Interactive data set of governance indicators available on line at http://www.worldbank.org/wbi/governance/govdata/ index.html

Lewis, B. 2005. "Freedom and Justice in the Modern Middle East." Foreign Affairs. 84:3:36-51.

Lippert, O. 2005. "Overcoming Bangladesh's Democratic Deficit." The Rajshahi Lecture. Inroads 17:88-99. Available on line at http://www.inroadsjournal.ca

Meier, G. 2001. "The Old Generation of Development Economics and the New." G. Meier and J. Stiglitz eds. 2001. Frontiers of Development Economics: The Future in Perspective. Washington, D.C. & New York: World Bank and Oxford University Press.

Murphy, R., N. Kamal and J. Richards. 2002. Electricity for All: Electrification and Rural Development in Bangladesh. Commentary 2. Available on line at http://www.iubat.edu/cpr

Organisation for Economic Cooperation and Development (OECD). 2005. Joint Progress Toward Enhanced Aid Effectiveness: Harmonisation, Alignment, Results. Available on line at http://www.oecd.org/dac/effectiveness

Putnam, R., R. Leonardi and R. Nanetti. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press, 1993.

Rodrik, D. 2004." Rethinking Growth Policies in the Developing World." Draft of the Luca d'Agliano Lecture in Development Economics delivered in Torino, Italy. Available on line at http://courses.umass.edu/econ804/rodrik.pdf

Roodman, D. 2005. "The Commitment to Development Index: 2005 Edition." Available on line at http://www.cgdev.org/content/publications/detail/3647/

Sachs, J. 2005. The End of Poverty: How We Can Make It Happen in Out Lifetime. London: Penguin Books.

Sen, A. 1999. Development as Freedom. New York: Knopf.

Shanker, R. 2004. "From the Editor." Journal of Development Policy and Practice. 1:1:v. Available online at http://www.acdi-cida.gc.ca/journal

United Kingdom. 2005. Out Common Interest. London: Commission for Africa. Available on line at http://commissonforafrica.org

United Nations Development Project (UNDP). 2005. Human Development Report 2004. National data available on line at http://hdrundp.org/statistics/data/ index_countries.cfm

United Nations Millennium Project. 2005. Investing in Development: A Practical Plan to Achieve the Millennium Development Goals. Jeffrey Sachs, director. New York: United Nations Development Program.

World Bank. 1993. World Development Report 1993: Investing in Health. Washington, D.C. and New York: World Bank and Oxford University Press.

World Bank. 1997a. World Development Report 1997: The State in a Changing World. Washington, D.C. and New York: World Bank and Oxford University Press.

World Bank. 1997b. Primary Education in India. Series: Development in Practice. Washington, D.C.: World Bank.

World Bank. 1998. Assessing Aid: What Works, What Doesn't, and Why. Washington, D.C. and New York: World Bank and Oxford University Press.

World Bank. 2004. World Development Report 2005: A Better Investment Climate for Everyone. Washington, D.C. and New York: World Bank and Oxford University Press.

(1) Politicized Islam is the subject of a vast number of analyses in recent years. Three review essays discussing recently published books are Geertz (2003), Habeck (2005), and Lewis (2005).

(2) These two regions constitute approximate]y one-half of the world's population. Income levels within sub-regions of each vary dramatically, by factors of four to one. In the case of East Asia, the most prosperous regions are the eastern coastal provinces of China. In the case of South Asia, economic prosperity is primarily to be round in certain southern and western Indian states. Furthermore, 300 million people live in either Pakistan or Bangladesh, countries with average per capita incomes only two-thirds the overall Indian average.

(3) Gerald Meier (2001) provides an accessible survey of trends in development economics since mid-20th century.

(4) A classic example of this direction was the World Bank's 1993 Annual Development Report, devoted to health policy (World Bank 1993). Achieving good basic health and education outcomes does hot however guarantee economic prosperity. Illiteracy in the Indian state of Kerala has been effectively eliminated, and life expectancy at birth is 10 years greater than the Indian average. Despite having outperformed all other states on these social dimensions, Kerala's average per capita income remains slightly below the Indian average (World Bank 1997b, 64).

(5) A recent example of a successful election campaign based on a promise of "free" electricity occurred in the 2004 state election in Andhra Pradesh (Farooq 2004). Load shedding refers to a utility's management of excess demand by randomly switching off power for particular regions. See also the discussion in Murphy et al. (2002).

(6) Owen Lippert (2005) provides an accessible introduction to the non-cooperative political culture of Bangladesh and its adverse effects on the country's development.

(7) In Figure 1, the bars around the point estimates of government effectiveness for the 25 CIDA-targeted countries indicate that, with 90 percent probability, the true index value lies within the designated range.

(8) The regressions discussed here ignore many intuitively relevant explanatory variables. And, in considering regressions (3) - (5), readers should be aware of the controversy surrounding reverse causation: does better governance bring about higher GDP growth, or does higher growth induce better governance? Probably countries do experience reverse causation, but its extent is hard to measure. Using sophisticated econometric techniques, some argue it is substantial (United Nations Millennium Project 2005, ch.7); some argue it is a minor phenomenon (Kaufmann & Kraay 2002, Kaufmann et al. 2005).

(9) The illustrated growth rate for Ukraine has been augmented by 7.16 percentage points, that for each of the sub-Saharan African countries by 1.76 percentage points.

(10) Dollar and Levin (2004) examined the distribution of aid to host countries by 41 bilateral and multilateral donor agencies, including CIDA. The study analyzed distribution in each of four years, 1999 to 2002. For each donor agency, the authors regressed its aid to a particular country on the country's population, its per capita GDP, and a measure of its institutional quality. Given the equation form, the regression coefficients provide donor-specific measures (to be precise, donor-specific elasticities) of the extent to which donors are concentrating aid on countries that are poor and those that have good institutions. The overall index averages, for each donor, its score on the separate institutional quality-selectivity and poverty-selectivity indices.

(11) Uganda offers a topical example. Under President Museveni, continuously in power since 1986, the country recovered from the brutal regimes of Idi Amin and Milton Obote, and bas realized impressive economic growth relative to most African states (CIA 2006). However, corruption indices have worsened and during the recent presidential election Museveni imprisoned his principal political opponent. In response, the UK cut bilateral aid to the government, diverting it to humanitarian relief in northern Uganda where a guerrilla war festers (DFID 2005). Whether such donor pressure can bring about desired policy changes is moot (BBC 2006).

(12) Canada has in recent years eliminated all tariffs for those countries classified by the UN as "least developed" (Goldfarb 2002). Most CIDA-targeted countries are on the "least developed" list.

(13) The 2003 HDI is constructed as an equal weighting of three separate sub-indices. For each of the three sub-indices, the denominator is a predetermined range; the numerator is the distance for the particular country above the minimum. Hence, by construction, for each sub-index and the HDI itself, the minimum value is zero; the maximum is one. The first sub-index is calculated from the log of per capita GDP, measured in terms of purchasing power parity (PPP) US dollars. The denominator is the difference between log (40,000) and log (100). The numerator for country is the difference between log (per capita GDP of country i) and log (100). The second sub-index weights two indices: two thirds weight to adult literacy rate and one third to gross enrolment rate. For literacy, the denominator is 100 percent (= 100 percent--0 percent); the numerator is the literacy rate of those over age 15 in country i. For enrolment, the denominator measures total population in relevant age cohorts; the denominator combines primary, secondary, and tertiary, and measures actual enrolment at the beginning of the school year. The third sub-index is calculated from average life expectancy at birth. The denominator is 60 years (= 85 years - 25 years); the numerator is life expectancy in country i less 25 years.

John Richards is a professor in Simon Fraser University's Graduate Public Policy Program and the Phillips Scholar in Social Policy at the C.D. Howe Institute.

Table 1: Population Living Below the Poverty Line, by Developing Region

                              1990   2001   Change

                               millions of people
                                 (percentage of
                              regional population)

$1 a day poverty line

East Asia                      472    271     -201
                              (30)   (15)

Eastern Europe/Central Asia      2     17       15
                                (1)   (4)

Latin America /Caribbean        49     50        1
                               (11)   (10)

Middle East/North Africa         6      7        1
                               (2)    (2)

South Asia                     462    431      -31
                              (41)   (31)

Sub-Saharan Africa             227    313       86
                              (45)   (46)

Total                         1218   1089     -129

$2 a day poverty line

East Asia                     1116    865     -251
                              (70)   (47)

Eastern Europe/Central Asia     23     93       70
                               (5)    (20)

Latin America /Caribbean       125    128        3
                               (28)   (25)

Middle East/North Africa        51     70       19
                               (21)   (23)

South Asia                     958   1064      106
                              (86)   (77)

Sub-Saharan Africa             382    516      134
                               (75)   (77)

Total                         2655   2736       81

Source: Adapted from UNDP (2005), Table 2.3

Note: Poverty lines are set in 1993 US dollars, adjusted for
purchasing power parity.

Table 2: Regression Models to Explain Average Per Capita GDP Growth
Rates among 25 CIDA-targeted Countries, 1990-2003

                                        (1)         (2)         (3)

Intercept                             2.54 ***    2.25 ***    4.08 ***

Regional compensation (b)
  Sub-Saharan Africa (l: country
    in sub-Saharan Africa; 0:
    elsewhere)                       -1.35 *     -1.51 **    -1.28 **

  Eastern Europe (1: country in
    Eastern Europe; 0: elsewhere)    -7.24 ***   -6.96 ***   -6.85 ***

Government effectiveness, 2004
  (World Bank index)                                          2.90 **

ODA/GDP (c) (percent)                             0.029

ODA/GDP interacted with
  "more effectively" governed"
  (ODA as percent of GDP: "more
  effectively" governed countries;
  0: "less effectively" governed)

ODA/GDP interacted with
  "less effectively" governed"
  (ODA as percent of GDP: "less
  effectively" governed countries;
  0: "more effectively" governed)

R square                              0.48        0.49        0.59

R square, adjusted                    0.43        0.42        0.54

                                        (4)         (5)

Intercept                             2.48 ***    2.17 ***

Regional compensation (b)
  Sub-Saharan Africa (1: country
    in sub-Saharan Africa; 0:
    elsewhere)                       -1.76 ***   -1.87 ***

  Eastern Europe (1: country in
    Eastern Europe; 0: elsewhere)    -7.16 ***   -6.87 ***

Government effectiveness, 2004
  (World Bank index)

ODA/GDP (c) (percent)

ODA/GDP interacted with               0.076 *     0.096 ***
  "more effectively" governed"
  (ODA as percent of GDP: "more
  effectively" governed countries;
  0: "less effectively" governed)

ODA/GDP interacted with              -0.037
  "less effectively" governed"
  (ODA as percent of GDP: "less
  effectively" governed countries;
  0: "more effectively" governed)

R square                              0.66        0.65

R square, adjusted                    0.60        0.60

Source: On line data available from UNDP (2005), Kaufmann et al.
(2005).

Legend: Variable coefficient is zero with probability below:

* 0.075

** 0.05

*** 0.01

Notes:

(a) Dependent variable is average per capita GDP growth rate,
1990-2003, measured in the country's local currency.

(b) Regional compensation variables adjust growth rates for
regionally specific problems, such as AIDS incidence and regime
collapse, events of particular importance in Eastern Europe and
Sub-Saharan Africa. In effect, regional compensation eliminates
consideration of any impact of other variables on Ukraine.

(c) ODA/GDP is the average of net aid to the relevant country relative
to its GDP for 1990 and 2003, the opening and end-years of the
interval.

(d) The interactive variables divide the targeted countries, based on
their ranking in terms of the World Bank government effectiveness
index. The "more effectively" governed comprise (in descending
effectiveness rank) Senegal, Ghana, Guyana, Sri Lanka, Mali, Vietnam,
Indonesia, Tanzania, Benin, Mozambique, Burkina Faso, and Rwanda. The
"less effectively" governed comprise (again in descending rank)
Pakistan, Bolivia, Cameroon, Ukraine, Honduras, Nicaragua, Bangladesh,
Malawi, Kenya, Zambia, Cambodia, Niger, and Ethiopia.