On September 21, 2003, the World Bank unveiled its annual flagship publication, the 2004 World Development Report, entitled "Making Services Work for Poor People." The WDR's main premise is that basic services - primary education, basic health care, water and electricity services -fail to reach the poor because too many governments lack sound and representative institutions of governance. Ironically, the report expresses strong confidence in the ability of these same unaccountable governments to regulate private service provision.

In addition to deficient institutions, the WDR attributes failing public services to regressive budgets (that benefit mostly the middle class), petty corruption, and bureaucratic inertia. Among the many solutions it proposes for these problems are more progressive resource allocation, greater transparency, more competition, and citizen-based monitoring. It also encourages policy-makers to consider decentralizing service provision and -particularly in the areas of water and electricity – implementing private provision and higher user fees.

While the WDR is enthusiastic about the potential of markets to improve some essential services – often too enthusiastic, as discussed in this critique -it also makes clear that there are limits to private provision. Since the Bank's critics accuse the institution of promoting indiscriminate privatization, the World Bank has been using the WDR in public and press relations as evidence to counter such attacks.

The WDR's qualified support for public provision is an important statement to have on record in the public debate over reform of basic social services, like primary education and health care. The importance of the state's role in service provision is even more emphatic in the 2003 Human Development Report (HDR), entitled "Millennium Development Goals: A compact among nations to end human poverty." The UN Development Programme report urges a larger role for government in the direct provision of social services, and concludes that: "The supposed benefits of privatizing social services are elusive, with inconclusive evidence on efficiency and quality standards in the private relative to the public sector. Meanwhile, examples of market failures in private provisioning abound." (p. 113)

Notwithstanding its support for public provision of education and basic health care, the WDR endorses private provision of infrastructure services, such as water, sanitation, and energy. The report seeks to justify the World Bank's past and present efforts to privatize utilities by neglecting the mounting evidence of unacceptable risks of private provision in these sectors – especially in countries with weak regulatory capacity.

While the WDR carefully avoids a "silver bullet" for improving basic services, many of the reforms that it proposes are informed by an important underlying principle; that government should not both regulate and provide service provision. Indeed, the WDR suggests that public services tend to fail poor people because the fusion of these functions represents a conflict of interest. "When the policymaker takes a separate role from the provider, it is easier to say "I don't care what your problem is, just tell me the vaccination rates. Or the test scores. Or crime rates." When roles are mixed, bureaucracies become insular and tend to hide mistakes." (p. 98) The WDR does concede that most basic services in developed countries are delivered by government providers which are overseen by government regulators – and that this arrangement works just fine. But it rejects the same path for today's developing countries. Why?

Rich countries benefit from a long evolution of the relationships between the state and frontline providers. Almost all services provided directly to individuals in the now -rich countries were originally provided privately. They were eventually absorbed or consolidated by a state institution that had been separate from the existing provider organizations. The state began as an independent outside monitor and regulator of private activities. It largely retained that independence as a monitor after the same activities became public. For the developing world the desire for rapid expansion of public financing and provision short-circuits this historical development. Both the monitoring and the provision are taking place simultaneously. This is not necessarily a bad thing-the poor might otherwise have to wait much longer for services to reach them. But it does show that the current institutional features of rich countries may not transfer directly to poor countries without the establishment of a complementary regulatory structure, a structure that may need to be established beforehand. (p.99)

This explanation represents much more than a novel theory of development for academics to debate about. The imperative to separate policymakers (i.e., regulators) from service providers justifies the replacement of government service provision.

The present critique rejects the WDR's argument for two reasons: First, the report does not account for variation in outcome. As the WDR shows in numerous places, the developing world is filled with examples of effective and accountable government services which are regulated by the same government. Why wasn't good governance "short-circuited" by late and rapid development in these cases? Second, and most fundamentally, because basic public services are susceptible to market failures, there is widespread agreement that government must be able to regulate private providers. Yet if poor countries don't have "institutional features of rich countries" needed make government services work, they can hardly be expected to establish complex regulatory structures needed for private service provision.

This paper is organized into three sections: The first focuses on the WDR's analytical framework, the second on its policy implications, and the last on the relationship between the report's main messages and actual World Bank operations. The author thanks Nancy Alexander for her review and many substantive contributions to this critique.

A. Analytical framework
Anecdotes, not analysis
If one looks to the WDR to provide analytical guidance about how to choose among options for reform – especially between government and private provision -the report is not particularly useful. That is, it fails to provide policy-makers with the kinds of tools needed to evaluate the costs, risks and trade-offs inherent of different policy options in the context of particular institutional and economic constraints.

Especially with respect to the social sectors, the message of the WDR seems to be "let a thousand flowers bloom." The WDR is filled with seemingly random claims that private, public or decentralized provision has worked well somewhere in the past. There is an example somewhere in the document to support the viability of just about any kind of known reform. For example, under a properly pro-poor government: "an easy to monitor services such as immunization could be delivered by the public sector, or financed by the public sector and contracted out to the private or non-profit sector."

One could excuse a policy-maker for asking: aside from government, firms and NGOs, who else could deliver this service? And more importantly, how does a government decide which one to go with? At the end of the section on how to scale up service reform, the WDR Overview concludes: "In addition to creating and disseminating information, other reforms to improve service delivery will require careful consideration of the particular setting . . . Everything from publicly financed central government provision to user-financed community provision can work (or fail to work) in different circumstances." Unfortunately, the WDR's main analytical framework does not focus on country-specific circumstances, but rather generic qualities of services themselves. (An exception would be whether the political environment is "pro-poor" or not – a slippery and undefined concept that can be endlessly manipulated according to one's preferred policy outcome.)

Shaky analytical framework
The WDR attempts to provide an analytical framework -a highly technocratic approach to selecting service reforms. The recommendations are based on three sets of criteria:

1. Are client preferences homogeneous or heterogeneous?
2. Are politics "pro-poor" or "pro-rich?"
3. Are outputs easy or hard to monitor?

Some rather odd policy recommendations emerge from the WDR's framework.^[1] For example, there is only one set of circumstances in which central government provision is recommended: pro-poor politics, homogeneous clients and difficulty in monitoring outcomes. It stretches credulity that, out the eight possible combinations of criteria, the public service option that has been used most often and most effectively in the wealthy northern countries appears to have extremely limited applicability in developing countries.

The three dualistic, "black and white" criteria are unhelpful since the real world is colored in countless shades of gray. For example, few political scientists would have the temerity to assign a categorical pro-rich or pro-poor classification across service sectors to countries they have studied closely. Moreover, because of political, ethnic or religious considerations – and of course pervasive gender bias – politics may prove to be quite "pro-poor" for certain groups or areas but not for others. An event so fleeting as a change in government can significantly can significantly improve or worsen the political status of poor people. Given that the WDR's policy-recommendations differ significantly depending on whether a country is deemed to have pro-poor politics, the many difficult ies in measuring the criterion present serious challenges for policy-makers.

In the case of heterogeneity, even if there is little deba te over the clarity of the criterion, its role in the analytical framework is overly simplistic. When consumer preference is found to be "heterogeneous"—that is, when different users have different tastes or needs – the WDR supports some variant of local government or even community control. However, decentralization is a national policy, affecting all local governments – even those that are manifestly corrupt and deaf to the needs of the poor. Giving more responsibility to weak or unaccountable government is unlikely to provide more choices for heterogeneous clients. Contrary to WDR claims that decentralization improves the welfare of the poor outcomes, empirical research suggests that "the notion that there is a predictable or general link between decentralization of government and the development of more pro-poor policies or poverty alleviation outcomes clearly lacks any convincing evidence." ^[2]

Sustainability of Community-Driven Development
The WDR is enthusiastic about decentralized, community-driven development (CDD) projects and acknowledge the importance of strengthening local government to administer them. However, the Bank's CDD resources typically bypass local government and, by the Bank's own estimation, have very low sustainability rates. For instance, a significant sample of CDD (Social Fund) water operations revealed a likely sustainability of 24 percent of operations. In addition, the poverty reduction impact of World Bank financing of community-driven development initiatives is weak. The Bank's own evaluation department concluded that Social Funds, one of the main financing mechanisms for community development projects, did not improve most welfare indicators relative to a control group. In fact, it found that non-poor communities were more likely to benefit from Social Fund expenditures than the poor.* For a critical review of the rise of social funds in development aid, Judith Tendler," why Are Social Funds So Popular?" in Local Dynamics in the Era of Globalization (Shahid Yusuf, ed.) Oxford University Press, 2000.
*OED, Social Funds: Assessing Effectiveness, World Bank, 2002

And what about communities? The WDR itself concedes that "communities are not homogeneous – problems of exclusion and elite capture can be the same as in government systems." One would hope that such a caveat would come with guidance on how to assess the readiness of communities to make investment decisions and allocate resources accountably. Yet notwithstanding extensive evidence of corruption, patronage and social exclusion within local governments and communities, the WDR touts highly risky policies as bringing services closer to the people.

Moreover, even where decentralized government is accountable, all too often it is also broke. The trend toward decentralization has created "unfunded mandates" in country after country, especially as fiscal transfers from the central government are reduced – often under direction from the same multilateral lenders who thought decentralization was the solution for better services. The WDR framework asks only if the service is appropriate for local government, and fails to examine whether the local government has sustainable financing and capacity to deliver services.

The criterion monitoring is also much more slippery than the authors admit. For example, the WDR casually claims that water and electricity are "easily monitored services." (p. 14) It is true that checking the existence of a household connection to a utility network is easier to verify than knowing whether a child is being properly taught. However, this does not mean that monitoring water and electricity services is "easy," particularly for governments that lack regulatory capacity.

Avoiding market failure in private infrastructure provision requires accurate monitoring of both performance and costs. World Bank economist Vivien Foster has argued that a "regulated company will have a strong incentive to abuse this strategic advantage [its control over information] by under-supplying information or distorting the information supplied." ^[3] Private sector providers have incentives to maximize profits and provide minimal compliance for contractual commitments that serve social rather than revenue-generating goals. In order to monitor their behavior, regulators must gather extensive data at considerable cost.

Generic advice
Policy-makers would do well to focus on specific, observed constraints to implementing private provision. Changing the analytical framework to focus on actual context instead of generic attributes could help focus national and local deliberations about policy options. Instead of using mechanical and largely abstract notions of service types, policy-makers should focus instead on the known problem . The WDR itself identifies four distinct ways that services fail poor people:

1. Too much spending on the non-poor;
2. Failure of expenditures to reach front-line providers;
3. Corruption and disincentives among service workers themselves; and
4. Lack of demand among the poor, due to economic, informational or cultural factors.

As discussed below, this critique proposes adding a fifth problem: inadequate budget.
The main questions driving the WDR's analytical framework are: "What kind of service do I have, and what kind of reform is best for that kind of service?" But a quick glance around the world shows the folly of this approach. Countries successfully employ every imaginable variation of service provision: decentralization; and community control; monitoring systems (ranging from citizen participation to high tech computerization); and private provision (through corporations, small businesses and NGOs). (And of course, there have been failures with each of these approaches as well.) Successes are obviously not inextricably tied to the nature of the service. Rather, they were achieved because of the actual country context – that is, the constraints and opportunities specific to the country or locality.

An alternative question that would drive the analysis in reforming a failing service is: "What is the root of the existing problem?" Unlike general attributes, the answer to this question is not "given." The service must be investigated, the perspectives of managers, workers and users understood. Stakeholders would also provide information about assets and capabilities that are worth building on. By identifying the institutions that provide technical competence, accountability, and pro-poor advocacy, services could be reformed based on an understanding of context-specific opportunities. By using this kind of analytical framework, there is little doubt that many of the reforms proposed in the WDR

– including those involving private participation
– would emerge as appropriate solutions. But these solutions would result from on-the-ground consultation.

B. Policy implications for private provision and resource allocation
Bad advice on privatization of infrastructure
The WDR is enthusiastic about direct private provision of infrastructure services. Because the WDR's analytical framework does not include an assessment of the well-known risks associated with asymmetric information,^[4] it offers extremely questionable advice on infrastructure services. The report maintains that government's role in these sectors should generally be limited to regulation and subsidies. The reason is that: "There are few advantages to the government's providing the [infrastructure] service itself, which explains why the past decade has seen many privatizations, concessions and the like in water and energy."^[5]

Are water and Electricity Public Goods?
The WDR's distinct policy advice for social and infrastructure services reveals that its authors do not regard water and electricity as "Public goods." For example, Box 2 (Overview), entitled, "Service-a public responsibility "refers only to health and education. Why? First these services are replete with market failures-with externalities, as when an infected child spreads a disease to playmates, or a farmer benefits from a neighbor's ability to read. So the private sector, left to its devices, will not achieve the level of health and education that society desires. Second basic health and education are considered fundamental human rights," Later, the report justifies charging user fees "for water, electricity and other services whose benefits are enjoyed mainly by the user…" (p. 10) (Strangely, in the same paragraph the WDR recognizes water quality as a public good, because of public health externalities. Why water quality would be considered as a public good but water access would not remains unclear, given that access has an obvious impact on public health. Indeed, it is lack of access to safe drinking water that leads poor people to contaminated water resources.)

These points are certainly valid, but they apply with equal force to water and electricity. Water and sanitation services obviously provide tremendous positive externalities, preventing the spread of waterborne diseases (such as cholera, which killed hundreds of South Africans following the imposition of user fees.) Electricity also provides important benefits that extend well beyond the consumer, including higher economic productivity and growth, and promoting public education bye enabling children to study at night. As for what can happen to electricity service when the private sector is "left to its devices," one need recall the traumatic price hikes and blackouts that plagued California and the US northeast. Finally, the UN has explicitly identified access to safe drinking water as fundamental human rights, once again making the distinction between water and health care seem arbitrary.

Given the disastrous record of privatized water and energy, one would expect some elaboration on the sweeping dismissal of public provision. The WDR does not mention what the "few advantages" of public service might be, or how those advantages can be compared to those of private provision. As this critical assessment is left to the reader's imagination, a few advantages of direct public provision are offered here:
No need to regulate firms with incentives and capacity to withhold information needed for pricing and investment decisions. While government provision does require regulation, achieving accountable supervision of a public utility is qualitatively different – and much less costly – than for a private one. No monopolistic rent-seeking or market manipulation. Under government provision of a natural monopolies or erratic markets like energy generation, there is no incentive to charge monopolistic prices or distort supply. Similarly, government provision requires no profit margin, which increases prices for consumers.

No corrupt bidding process or sweetheart deals on private concessions.
When service is poor, service users can exercise demands more effectively as organized citizens, rather than atomized customers.

World Bank research has recognized the importance of regulation when it comes to private provision of monopoly services, especially network-based sectors that are not subject to competition. The WDR itself includes a section on the importance of competent and autonomous regulation in ensuring that providers – public or private – be held accountable. Yet even though most of the Bank's borrowing governments have relatively weak regulatory capacity – especially in terms of monitoring and sanctioning powerful private sector firms, the WDR embraces privatization of water and energy.^[6]

Privatization Paradox
WDR's facile advice in favor of infrastructure privatization underscores a fundamental dilemma of development that the World Bank fails to confront: The same accountability limitations that make public services fail will almost certainly undermine the effectiveness of private provision and decentralized services. The simple reason is that, no matter who provides a service, the state is ultimately responsible for regulating the provider and protecting consumers. The report observers that: "Providers end up being more accountable to policymakers than to clients, which breaks the long route of accountability" which depends on the collective political power of the citizenry. Importantly, this is not a problem, by definition, when serving the public interest is the policy-maker's priority.

The WDR argues that there can be a conflict of interest when government policy-makers regulate government providers. It strongly underscores how private provision is a solution because it separates these functions. However, the report does not ask why so many countries succeed in putting regulators and providers "under one roof." More seriously, it implausibly suggests putting regulators and providers "under one roof." More seriously, it implausibly suggests that consumers (including the poor) will be better off when the same, ostensibly unaccountable policy-makers regulate private providers instead of government providers.

Avoiding Tough Question About Equity
The WDR encourages governments to attract private investment by liberalizing and deregulating the services they provide. It's true that opening up essential services may attract investor. However, the private sector often engages in "cherry picking," or "cream-skimming." Firms have little incentive to invest in "unprofitable people." In the case of utilities, private providers like to expand household access (e.g. hook up water pipes, make connection to electrical grid) or upgrade services in urban area, especially where middle class consumers demand more and better services.

They are less likely to go into peri-uban, slum or rural areas, where topography is more difficult, per capita consumption is less, and most importantly, incomes are lower. Cherry picking also constrains investment in the health care sector. Private hospitals prefer to serve drop those who develop conditions requiring medical attention. As a result, liberalizing services is increasingly associated with the establishment of a two-tired services supply with a corporate segment focused on the healthy and wealthy and an under-financed public sector focusing on the poor and sick.

As noted above, the WDR's main premise is that services fail to reach the poor because too many governments lack sound and representative institutions of governance. So it is a mystery why the report expresses such confidence in the ability of these same unaccountable governments to regulate private water and energy concessions. The report explains that as a result of weak citizen voice, "pub lic services often become the currency of political patronage and clientelism."

(p. 7) But how will this same neglected citizenry force its own corrupt government to avoid enriching itself through shady deals with private firms with strong rent-seeking incentives and resources? The report even warns: "If there is no good legal system and the civil service is subject to bribes . . . private sector contracts might be a major source of corruption." Yet how many of the Bank's borrowing members - especially those among the least developed countries

- do not have these characteristics?

The WDR's general description of proper regulatory features for the water sector is sensible enough. However, what would be more useful than making the uncontroversial argument that good regulation is necessary, is an ex-ante appraisal of regulatory capacity. Surely even the World Bank would have to admit that there are some places in the world in which meaningful regulation – and therefore private provision of a monopoly network -is simply not an option. If it were possible to benchmark standards and measure regulatory capacity needed for different kinds of private provision, policy-makers would have more confidence in their decisions to privatize or reform existing public services.

Do budgets matter?
The WDR's answer to the question seems to be: only marginally. One of the report's central arguments is that more money won't be enough to improve services when budgets are regressive (e.g., benefit primarily the non-poor) or when service providers lack incentives to reach the poor. That's fair enough. But while the report makes an important case for reforming the way services are delivered, it fails to acknowledge the damage that budget cutting -often under imposed structural adjustment loans -has caused public services over the last 20 years. Today the IMF still ties much of its financing to conditions that national and local governments cut back resources available for delivering public services.

At World Bank seminar in May 2003, economist Jeff Sachs chastised the international organizations, including the World Bank, for unfairly lowering the expectations of what is possible in the water and sanitation sector. Indeed, he blamed the institutions for putting developing countries on a starvation diet for twenty years and insisted that huge amounts of public funding are required to achieve functioning water supply systems. Sachs said that "if public-private partnerships are a means to get the private sector to do work that only the public sector can pay for, it will be a sad experiment."

Not even the most ardent defenders of government services argue that more money is the whole solution. Service providers can be made more responsive and efficient through better manageme nt techniques, greater autonomy to frontline workers, and more transparent information about resources and performance. But surely the budget is often a major part of the solution as well.

Ironically, the World Bank's own press release, "Key Services Often Fail Poor People," highlights this fact through examples: "Indonesia used its oil windfalls to build new schools and hire more teachers, doubling primary enrollment to 90 percent by 1986.

A program in Mexico that gives cash to poor households if they visited a clinic regularly and their children attended school reduced illness among children by 20 percent, and increased secondary enrollment by 5 percentage points for boys and 8 for girls."

The Indonesian case might suggest that all developing countries should use their vast petroleum wealth to improve social services, good advice that regrettably may go unheeded in countries with more limited natural resource endowments. And while the Mexican program may be innovative, its key element appears to be "cash." In other words, budgets do matter. The point is so obvious that it should hardly need emphasizing, but the WDR's skepticism over increasing budgets makes it necessary to re-state the obvious. The WDR is right to caution against throwing good money after bad and insisting on the most efficient and accountable public services possible. But it should make this point in the context of guidance for how to assess resources gaps and suggestions for overcoming those gaps.

The WDR's position on budgets is also problematic because it arbitrarily separates the issue of resources from the level of performance. Any analysis of failing services should address the potential relation of the overall sector budget to other shortcomings of government providers. Reformers should start by asking the simple question of cause and effect: Are corruption, indifference and inefficiency the independent causes of failing services, or are they the outcomes of a fiscal straightjacket? When service providers perform poorly, the WDR takes a general position that the problems are entrenched within the organizational culture or wider political system. In analytical terms, it treats budgets, productivity and accountability as conceptually separate issues. The possibility that greater resources can, by themselves, improve productivity and accountability, is not seriously considered.

The WDR largely ignores the possibility that starving public services of funds can lead directly to inefficiency, corruption and professional apathy. As the Overview argues, "As resources become more productive, the argument for additional resources becomes more persuasive." Of course in many cases existing resources are badly mismanaged. But in making this claim, the WDR takes a categorical position that the productivity of services is not related to the level of resources available. Yet in the case of infrastructure, low budgets postpone investment and maintenance, leading to physical deterioration that inevitably undermines productivity. In the social services, budget cuts have led to reduced salaries, shabby working conditions and inadequate supplies. These conditions increase incentives for petty bribery and also demoralize professionals who see little chance of making a difference.

Finally, one of the WDR's most important progressive policy recommendations is lifeline rates or free minimum services to poor people (see discussion below). Especially where poverty is pervasive, making such subsidies possible will obviously require tremendous increases in service budgets.

User fee issues
The WDR makes a useful analytical contribution on the controversial subject of user fees. In the context of widespread criticism of Bank-financed policy reforms that impose commercial pricing on the poor, the WDR emphatically states that there should be "no blanket policy on user fees." (See Box 4.4) It also provides an interesting analytical framework for helping policy-makers determine when user fees are appropriate, and with what kinds of complementary measures to protect the poor. The WDR also supports the imposition of user fees for services not used primarily by the poor, to ensure that the middle class is not the primary beneficiary of public resources.

Three particular recommendations, however, merit special scrutiny – not because they are wrong, but because they are highly susceptible to political discretion.

Lifeline rates. When it is difficult to distinguish the poor from the non-poor, the WDR promotes user fees with "lifeline" rates that provide a base level of service for free or at a highly subsidized price. This is clearly a progressive option, and is likely to avoid administrative difficulties associated with targeted subsidies, especially for meter-based services like water and electricity. South Africa adopted a national law allowing 6 kl of water and 20-50 kwh of electricity per month free per family.

While free and lifeline rates are good ideas in principle, they are tough to do right in practice. Simply determining "how much" of a service satisfies basic needs of the poor is fraught with political and economic controversy. Indeed, many argue that South Africa's pioneering laws do not come close to providing sufficient water and electricity for poor families. For governments facing budget pressures and MDB requirements to reduce deficits (or increase primary surpluses), the path of least resistance is to ignore such pro-poor measures or make them so limited as to leave the poor highly vulnerable. This brings us back to the central issue of resources. Despite the WDR's insistence than "more money isn't enough," poor countries will obviously need a lot more money to make lifeline rates a reality.

Targeting. When it is possible to distinguish the poor from the non-poor, the WDR promotes user fees while targeting poor people with direct money transfers or exemptions from payment. This may be progressive in principle, but in practice can be extremely difficult to do with precision and equity. Studies of "targeted" subsid y and voucher schemes show high levels of "leakage" (benefits going to the non-poor) and exclusion (of the poor). Particularly in countries or areas where the vast majority of people are poor, targeted subsidies are likely to be administratively overwhelming and economically unfeasible.

User fees. Fundamentally, the WDR accepts user fees as a "necessary evil" when the service will not be "adequately delivered without user fees." ^[7] The argument is tautological: User fees must be used when services cannot be delivered without user fees. The underlying question, of course, is how to determine when this situation exists. According to the WDR, the situation "calls for an honest appraisal of the ability [of government] to deliver services along the long route [where citizens hold policy-makers accountable for making sure that service providers do their job]. If teachers or medical providers cannot be supervised and medical stores not maintained by government, then clients, by default, must bring purchasing power to bear." (Box 4.4)

Similarly, the WDR argues that when politicians increase their power by providing subsidized electricity services mainly to wealthier constituencies: "Reducing those rents by raising tariffs or having the private sector provide electric ity, even if it violates the principles of equity – they are already violated in the existing system – may be the only way of improving electricity services to the poor." (p. 13)

To paraphrase the WDR's argument, user fees are acceptable when policy-makers or public service providers are not accountable. The problem with this position is that it involves a tremendous judgment call that can easily be influenced by fiscal considerations or World Bank pressure. When is the public sector "beyond hope?" By cla iming that user fees are justified when policy-makers pursue political goals or don't hold providers accountable (and are therefore unresponsive to their own citizens) the WDR provides a potential loophole for full cost recovery for virtually any poorly performing service. (In the case of patronage-based subsidies, it would seem that corrupt and self-interested policy-makers who are pressured into privatizing a public service would have every incentive and opportunity to enrich themselves further at the pub lic's expense, and little interest in ensuring that the firm serves poor people.)

Before turning to regressive user fees, shouldn't policy-makers attempt to make providers more responsive? The WDR identifies plenty of ideas for how to make this happen, including more transparent disclosure of information and getting citizens involved in monitoring the service. If reaching the poor is truly the goal of the World Bank, then it stands to reason that it would encourage borrowing governments to first attempt to improve accountability – or ensure proper allocation of lifeline resources to the poor -before imposing regressive user fees. By using accountability as a pretext for commercial pricing (incidentally, a common step prior to privatization) the Bank essentially "throws up its hands" and admits defeat in the attempt to improve the lives of the poor.

Preconceptions about workers distorts judgment ^[8]
Much of the report treats workers as a problem—often the problem. It is filled with many critical references to bad worker behavior, such as absenteeism or even hitting patients. It sees greater discipline, and even the insecurity of the labor market, as methods for ‘dealing' with these problems. But for the most vital public services, like education and healthcare, the workers are the service – the teachers, nurses, and others. Their income, capacity, and motivation are key factors in the performance of public utilities. The report fails to explore these issues properly – and distorts the understanding of some of its own case studies of good practice.

For example, it references Uganda's successful increase in primary school enrolment which resulted from cutting school fees, but fails to add that education quality has suffered because no extra teachers have been employed, leading to classes with over 100 students. It also refers to improvements in primary healthcare performance in Cambodia, which it presents as an example of what can be achieved by contracting-out to NGOs. But its own information clearly shows that the two key measures were abolishing user charges, and paying higher wages to workers, reducing their incentive to seek payments from patients.

The report duly notes the great advances made in health care reforms during the 1980s and 1990s in state of Ceara, Brazil. But it fails to mention the importance of the treatment of the workforce. The major work on these reforms, written by MIT professor Judith Tendler and cited in the WDR's bibliography,^[9] explains that "the workers in these programmes showed high dedication to their jobs [and that] government itself fed the high dedication of these workers with repeated public demonstrations of admiration and respect for what they were doing . . . All this contributed to a new respect for these workers by the public – remarkable in a time of widespread contempt for government." Tendler pointedly contrasted these findings with "development advice in which the public servant is presumed guilty of self-interest unless proven otherwise" – a description that fits the WDR all too well.

As with every topic related to public services, the WDR avoids being completely categorical. The Overview offers the following: "Many, often the majority [or frontline service providers are driven by intrinsic motivation to serve." If the WDR authors truly believed this to be true, it's a wonder that they dedicate so little of their report to supporting and rewarding that intrinsic motivation, and so much to controlling and disciplining worker behavior. Moreover, the report skirts the issue of abrogation of workers' rights when privatization processes violate labor agreements.

C. World Bank operations
In terms of its operational impact on World Bank activities, the WDR's most important words may be those that follow the copyright information three pages before the first word on Page One. Although the WDR is routinely identified as a World Bank publication and suggestive of its current thinking on development policy, one may be surprised to discover that its "findings, interpretations, and conclusions . . . do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequences of their use."

As the WDR authors so rightly point out, accountability in public institutions is hard to achieve.

When it comes to public relations and press coverage, the World Bank Group is actively taking credit for the progressive and sensible recommendations of the WDR: the need for more pro-poor budgeting, lifeline rates and free services for the poor, greater transparency, the need for strong regulation of private service provision, the principled rejection of user fees for primary education, and doing away with the "project implementation unit," which was found to fragment government capacity. However, as the publication disclaimer suggests, when it comes to actual lending practices, evidence shows that the same organization often flouts the WDR's good advice whenever and wherever it chooses.

But perhaps no area is as controversial as World Bank conditionality, criticized by citizen groups, academics and (usua lly former) government officials for undermining democratic processes and imposing policies that just don't work. In the chapter "Donors and service reform," the WDR states that "donors attempt to [promote citizen voice by] imposing conditions and setting performance criteria on aid flows where voice is weak . . ." (p. 209, emphasis added.) The claim is remarkable in two ways: first, it turns out that conditionality has always been about empowering the poor, even though the poor have been the most vociferous opponents of World Bank conditions. Second, although conditions are attached to virtually all adjustment loans in all borrowing countries, they have been imposed only "where voice is weak."

To its credit, the WDR concedes that "donor conditions are fundamentally different than citizen voice." Although they were imposed with the best of intentions and only with the true interests of the poor in mind, "traditional conditionality does not work well." But if one expects such an insight to result in a new approach to lending, disappointment awaits. The WDR's explanation for why traditional conditions fail is lack of political commitment – not the appropriateness of the reforms themselves.

"There is ample evidence today that conditions based on promises do no t work well, because they undermine ownership of the reform program. When policy-makers are not encouraged to develop their own positions on, say, privatization of water supply and other services, but rely on donor conditions in taking action, they can more easily deny responsibility for a later failure." (p. 209)

The tortured logic of the statement merits closer attention. The World Bank "encourages" policy-makers to develop positions identical to its own – from water privatization to trade liberalization - through a variety of instruments, such as technical assistance, capacity-building, and public statements confirming the wisdom of the government's liberalization efforts.^[10] The quote also suggests that when privatization fails – and it has failed a lot – the reason is not flawed policy, but rather insufficient commitment on the part of a government pressured into doing it.

Ownership: The gap between rhetoric and reality
The WDR observes that "Donors still underestimate how difficult it is to influence reform without undercutting domestic constituencies." (p. 203) That's an understatement. In June 2003, the World Bank released a survey by Princeton Survey Research Associates based on interviews with 2,600 "Opinion formers" in 48 countries that found a "Consistent and overwhelming" view (From all regions and countries) that the World Bank forces its agenda on developing countries. Only 22 percent of respondents reported that the Bank is doing a good job on reducing poverty.
(1) Not surprisingly, the WDR's chapter on donors contains very little on how the World Bank and other international institutions undermine country ownership through everyday lending practices, which in addition to policy conditionality, include: Starving the public sector. In a study of electricity reform in six countries (Argentina, India, Indonesia, Bulgaria, Ghana and South Africa), the World Resources Institute found that reforms were always driven by the immediate need for capital, usually as a result of the withdrawal of international donor support for the power sector.
(2) Issuance of Decrees. IMF and World Bank adjustment programs routinely call for the Executive to issue decrees to privatize resources and services, or recoup costs through tariffs and user fees. In reference to Latin America, three veteran World Bank officials note the importance of rules and legislation that strengthen the office of the presidency in relation to the legislature, including "powers to rule by decree" and "an unassailable presidential veto."
(3) Some of these reforms appear to reverse democratic progress in representative institutions achieved with great struggle over the last decades. Public Relations Campaigns. A recent communications "toolkit" Published by the World Bank's external relations department identifies democratization as a constraint to privatization.
(4) It never suggests that Bank or Government actually assess the merit of arguments voiced by citizens opposed such policies. (They are simply dismissed as propaganda from rent-seeking groups.) Moreover, in what may be a direct violation of its own Articles of Agreement, which prohibits interference in domestic political affairs of borrowers, the Bank's toolkit instructs Bank task team leaders on how to work with national legislatures to overturn parliamentary opposition to privatization, ad offers guidance on how to build legislative coalitions in favor of the policy.

(1)The Global Poll: Multinational Survey of Opinion Leaders 2002, Princeton Survey Research Associates (siteresources.worldbank.org/NEWS/Resources/globalpoll.pdf)

(2) Navroz Dubash, Power Politics: Equity and Environment in Electricity Reform, WRI, 2002.

(3)Shahid Javed Burki, Guillermo Perry and William Dillinger. Beyond the Center: Decentralizing the State. World Bank, Washington, DC, 1999.

(4)Public Communication Programs for Privatization Projects: A Toolkit for World Bank Task Team Leaders and Clients, World Bank, 2003.

The resulting recommendation for how to change lending practices is extremely significant – and disturbing. Rather than rethink the soundness of privatization policies in general, or their feasibility even in particularly weak regulatory environments, the WDR argues that the Bank needs a more direct way to get borrowing governments to adopt (what the Bank has known all along to be) the right policy.

"Empirical studies show that aid finance is ineffective in inducing policy reform in a bad policy environment. What works better is choosing recipients more carefully, based on performance (country selectivity) and setting conditions that reward reforms completed rather than those promised." (pp. 209-10)

Noting that conditionality doesn't work, the WDR proceeds to endorse more coercive conditionality. A "bad policy environment" exists where governments (often pressured by citizens who do exercise voice) resist reforms supported by the World Bank itself. Sometimes those policies may be justified, such as progressive budgeting. But more often they are liberalizing policies that are known to be risky and difficult to implement effectively. In any event, the WDR's approach is not to scale back conditionality, but rather to make it far more invasive than it has been thus far. The report thus reinforces the World Bank's message to borrowing governments: in order to qualify for increased aid – or perhaps any aid – you must comply with our policy conditions before receiving resources, not after.

October 8, 2003.

[1] In a heroic effort to synthesize these criteria, the authors produced a tree diagram - modestly entitled "Eight Sizes Fits All?" - which graphically locates policy options. See Figure 7 in the Overview, p. 14.

[2] Richard Crook and Alan Sverrisson, "Decentralisation and poverty-alleviation in developing countries: a comparative analysis or, is West Bengal unique?" IDS Working Paper 130, February 2001.

[3] In "Ten Years of Water Service Reform in Latin America: Towards an Anglo-French Model" (forthcoming)

[4] The problem of asymmetric information is at the heart of the so-called "principal-agent problem." Agents are supposed to work on behalf of principals: CEOs work for company shareholders, sub-contractors and private concessions work on behalf of government. However, in practice, agents have much more information about their activities than the principals they are supposed to serve, and often withhold or manipulate that information to promote their own rent-seeking interests. The only way to make sufficient information available to principals is through monitoring. When monitoring becomes very costly or cannot be done adequately, the principal needs to rethink their entire relationship with the agent.

[5] I reality, much of the explanation for why the past decade has seen so many privatizations is direct pressure from the multilateral lenders, such as conditionalities on adjustment loans or debt relief. Governments also turn to privatization when starved for public resources. The WDR team thus celebrates as common sense the policies that have been imposed on many of the Bank's borrowing members.

[6] The WDR advocates for a "short route of accountability" between provider and consumer policy-makers are not accountable to citizens. However, if the World Bank took seriously the need for citizen accountability, one might expect that in the many privatization schemes it has financed, that it would help governments design regulatory processes that include disclosure to and input from citizens. Yet there are few, if any, avenues for customer participation in regulatory decisions and debates in the developing world.

[7] Another condition is the inability to use lifeline rates. However, as discussed above, fiscal pressures often make progressive subsidies impossible in practice, even if they are a good idea in principle.

[8] This section was contributed by Dave Hall, Director of Public Services International Research Unit

[9] Judith Tendler, Good Government in the Tropics . John Hopkins University Press, 1997

[10] A statement consistent with the World Bank's rhetoric on "ownership" might simply have read: "Policy-makers should be encouraged to develop their own position on the best way to deliver water and other services." Such as statement would be made even more powerful by adding: "The World Bank can support government ownership by de-linking aid levels. from reform decisions.