• Open Access

How the World Bank funds protected areas

Authors


  • Editor
    Rudolf de Groot

Valerie Hickey, 1818 H Street, NW, Washington DC 20433, USA. Tel: 202-473-6343. E-mail: vhickey@worldbank.org

Abstract

The World Bank is the largest international funder of biodiversity conservation. It invests in protected areas to conserve species and spaces, protect ecosystems, and provide food, shelter, and other ecosystem services to local communities. It spends on average, $275 million annually supporting parks in developing countries. We examined their protected areas investment portfolio from 1988 to 2008 to understand how they allocate these funds. We found that more money is allocated to countries with progressively larger GDPs. Many, but not all, of these investments correlate with consensus opinions of high biodiversity priorities. But the World Bank's investments are not proportional; poorer countries receive relatively more funds than richer ones, regardless of biodiversity importance. We suggest that these investments focus on supporting parks that provide benefits to local communities, particularly in poorer nations, rather than on biodiversity priorities in a vacuum. This mirrors their mission to work for a world without poverty.

Introduction

With nearly 13% of the Earth's terrestrial surface formally put aside for conservation (Jenkins & Joppa 2009), protected areas are crucial for conserving biodiversity. Here, we consider how the World Bank allocates its funds for protected areas. It is the largest single international funder of biodiversity conservation projects (The World Bank 2003, 2006). It spends, on average, $275 million annually supporting protected areas in developing countries (The World Bank 2008), including managing about half of the Global Environment Facility's (GEF) park portfolio. It exerts large influence on conservation worldwide and has a key role to play in supporting the 2010 target to reduce the rate of biodiversity loss globally (Brooks et al. 2006). So, what does it fund and how are its decisions influenced?

Who spends what, where?

Globally, there are few recent or reliable data on current expenditures for biodiversity conservation in general, or protected areas in particular (Lapham & Livermore 2003; Emerton et al. 2006). With any reasonable level of certainty, we simply do not know who spends what, where. The United Nations Environment Programme and its World Conservation Monitoring Centre published the most recent global survey of protected area budgets and shortfalls in 1999, based on data collected in 1993 and 1995 (James et al. 1999). They estimated that $6 billion is spent annually on managing the global protected areas network (James et al. 2001). These expenditures include national government budgets for recurrent management expenses and capital outlays, as well as foreign government, multinational, and nongovernmental funding. Developed regions account for about 88% of global protected area spending (James et al. 2001). Several recent and influential articles on conservation funding continue to refer to this increasingly outdated database (e.g., Brooks et al. 2006; Balmford & Whitten 2003).

Developing regions spent $695 million annually on protected areas in the 1990s (James et al. 2001). One study suggests that national governments provided additional support to protected areas in an amount between $1.3 and 2.6 billion annually (Molnar et al. 2004). There are also many investments made by innumerable conservation financing mechanisms (entrance fees, departure taxes, payments for ecosystem services), small businesses, private landowners, and local communities that contribute to protected areas.

Overseas development assistance also supports the protected area agenda. The USA spent $22 million annually, Germany, France, the Netherlands, and the United Kingdom, $31.5 million annually, and Japan $3 million annually, in the period 1998 to 2000 on international projects involving biodiversity, including protected areas (Lapham & Livermore 2003).

The GEF allocated over $2 billion between 1991 and 2006 for more than 790 biodiversity projects in 155 countries (GEF 2009). These projects support the sustainability of protected area systems; mainstream conservation and sustainable use into production landscapes; build country capacity to implement the Cartagena Protocol on Biosafety, and prevent, control, and manage invasive alien species; and support the implementation of the Bonn Guidelines on Access to Genetic Resources and Benefit sharing.

International conservation organizations support protected areas in developing countries with an unknown amount of money. At least until recently, the three NGOs most active in conservation priority setting (World Wildlife Fund, Conservation International, and Birdlife International) could not provide data documenting their spending in each country because they did not track spending at the national or even regional level (Halpern et al. 2006). There are few data on the scale, pattern, and distribution and consequences of their existing levels of support (though for an example of this, see Brockington & Scholfield 2010).

Available data from the large international conservation NGOs with near-global reach do not allow us to calculate spending on protected areas, nor even on field projects, as opposed to national campaigns that have no particular geographic aim. For example, in 2007, World Wildlife Fund-United States (WWF) spent $133 million on conservation actions (WWF 2008); Conservation International (CI) spent $113 million on its conservation programs, of which $70 million was spent on regional programs in the Neotropics, Africa and Madagascar, and Asia and the Pacific, with the rest spent on global programs, including the Center for Applied Biodiversity Science and the Center for Environmental Leadership in Business (CI 2009); Wildlife Conservation Society (WCS) spent $62 million on its global conservation programs (WCS 2009); and, The Nature Conservancy spent almost $56 million internationally (USAID 2008). Birdlife International spent just over $17 million (8.67 million UK pounds) in 2007 (Birdlife International 2007).

The World Bank spends, on average, $309 million annually on biodiversity projects, of which $114 million comes from GEF sources, but is managed by the Bank. This makes it the largest single international funder of such projects (The World Bank 2003, 2006, 2008). In contrast, all conservation NGOs together spent $140 million on all conservation initiatives in sub-Saharan Africa in 2006 (Brockington & Scholfield 2010). Of its total, the World Bank spends an average of $275 million annually supporting protected areas in developing countries. Of this, $100 million annually are from the Bank's own sources, $60 million come from the GEF (but are wholly managed by the Bank), and $115 million are leveraged through cofinancing.

Investment allocation

The biodiversity value of a region is likely only one of a number of factors that influence where conservation funds should be allocated (Bode et al. 2008). Multiple factors may include donor wishes, historical relationships, in-country spending by other organizations or government agencies, geographic specialization by organizations, political stability, and opportunity. Here, we outline hypotheses that may affect how the World Bank invests its money.

Are they investing in consensus conservation areas?

The most often-voiced spending hypothesis is that countries, states, and agencies prioritize funds for species recovery by focusing on the relative burden of threats and prioritizing investment in those species and spaces at most immediate risk of extinction (e.g., Possingham et al. 2002; Rodrigues et al. 2006; Wilson et al. 2009). Others have pointed out that global conservation prioritization has had little success in informing actual conservation implementation (e.g., Brummitt & Nic Lughadha 2003; Brooks et al. 2006). Halpern et al. (2006) concluded “overall spending is predominantly in countries containing priority area(s)” but cautioned that “global priority models are having little effect on how money is distributed among countries containing high priority area(s).”

An assessment of donor support to conservation projects in Latin America and the Caribbean similarly found that some high-priority regions were relatively neglected and recommended that the distribution of funding across regions be reviewed (Castro et al. 2000).

Another conservation goal, preserving the functioning of the planet's ecosystems to the maximum extent possible, is not necessarily incompatible with this desire to save species. However, Kareiva & Marvier (2003) suggest that “If we measure success simply by tallying up total species protected, we risk the folly of allowing major ecosystems to degrade beyond repair simply because they do not provide lengthy species lists.” However, because of the broad overlap between ecosystem services and species, but the lack of a specific, and widely accepted map of such services, we are limited in our ability to explore this hypothesis.

Are they investing where there is most need?

The effect of national economic wealth on protected areas is one of the most important in securing conservation outcomes. Total spending on protected areas is proportional to gross domestic product (GDP) among nations. In addition, the number and total areas of protected areas are positively correlated with GDP, which is itself correlated with national expenditures on conservation (McKinney 2002). Wealthier nations spend more on conservation. As a result, international funding transfers may be spent in the least developed, and poorest, places (IMF 2008). We use GDP data to test for this.

Are they investing to alleviate poverty and induce development?

The economic value of each species to its local and national constituencies may be a significant consideration when deciding on resource allocation (e.g., Leader-Williams & Albon 1988). This is particularly true given the Bank's mission to alleviate poverty. Any of its investments where conservation can add value to development issues (as well as social justice and governance regimes) would appear most palatable. We test this hypothesis here by examining whether countries that depend on nature to attract tourists, especially in countries where tourism makes up an important share of the national economy, are more or less supported.

Are they investing where there is most chance of success?

With markedly lower conservation costs and generally greater conservation benefits, field programs typically have far higher benefit-to-cost ratios in less developed parts of the world (Balmford et al. 2003). So, rather than trying to identify the most threatened species in consensus areas on a map, the Bank may want to reward effective and inexpensive actions on the ground as they happen (Kareiva & Marvier 2003). So where conservation is inadequately funded, what funding does exist may be concentrated in areas that have proven successful. This is a difficult hypothesis to test as it introduces success into the equation. Thus, we do not test for it, but instead consider factors such as cost structure—how much “bang” is available, in terms of capital and labor inputs, “for the buck”—in anecdotally analyzing its relevance to investment decision-making.

Alternatively, because conservation outcomes may be correlated with good governance, conservation investments may focus on wealthier countries where outcomes are more assured. We ask whether, when corrected for GDP, corruption predicts spending.

An analysis of World Bank spending

Methods

Between July 1988 and June 2008, the World Bank approved 598 projects that wholly or partially supported biodiversity conservation and sustainable use (The World Bank 2008). Each project is associated with a single funding source within the World Bank. This means that actions on the ground supported by separate funding sources (such as GEF grants and an International Bank for Reconstruction and Development loan) count as separate projects. Projects invested in activities ranging from support to protected areas, conservation policies, and capacity building, to combating invasive species, and nurturing sustainable financing mechanisms. All told, the World Bank spent over $3.4 billion on conservation actions, and leveraged another $2.7 billion in cofinanced conservation dollars (The World Bank 2008). Of these 598 biodiversity projects, 329 involved direct support to protected areas in 97 countries and through 5 global and 12 regional projects. We collated all park projects in each country, and correlated this spending with information on economic output (IMF 2008), corruption indices (TI 2008), and tourism revenues (WTTC 2008). The latter indicators serve as proxies to address need, the chances of success, and an urge to alleviate poverty We used a price level index to distinguish current from constant prices to flatten any possible oscillations between purchasing power in different countries (The World Bank 2005).

We also lumped countries into regions a priori (Table 1)—Eastern Europe and Central Asia; Central and South America; Middle East and North Africa; Southeast Asia; Savannah Africa; Small Island Developing States (SIDS); South Asia; and Tropical Africa—based on their similar ecosystems across largely contiguous geographies and the existence of all of these regions but two as regions in the World Bank's organizational schema. The two differences include splitting Africa into two according to their stark ecological differences and the role of conservation, particularly as identified with tourism in their economic profile. We also include SIDS as a separate region to avoid losing their biodiversity import amidst their larger regional counterparts. Though an island, Madagascar is clearly not a small island developing state so we placed it within “Savannah Africa” because of its geographical proximity and the importance of nature to its tourism profile and as a result, its economic output. The regions and countries are listed in Table 2.

Table 1.  Ratios of regional funding estimates corrected for GDP in decreasing total investments. Thus, countries in Savannah Africa and Madagascar receive on average 325% more funds than expected for their GDP and Europe and Central Asia only 17%
Savannah Africa and Madagascar325%
Latin America146%
South Asia99%
Tropical Africa69%
Southeast Asia68%
Small island developing nations41%
Middle East and North Africa24%
Europe and temperate Asia17%
Table 2.  The regions and countries funded by the World Bank and their funding relative to regionally adjusted GDPs. Countries in boldface receive more than 200% and countries in italics less than 50% of the regionally and GDP-corrected expectations. Both protected area (PA) investments and GDP are adjusted to reflect purchasing power parity
RegionCountry  Proportional PA investmentsGDP (millions of dollars)
Europe and temperate AsiaAlbania492%47.25   24,209
Europe and temperate AsiaArmenia266%28.72   23,027
Europe and temperate AsiaAzerbaijan314%50.14   56,951
Europe and temperate AsiaBosnia and Herzegovina110% 8.48   11,978
Europe and temperate AsiaBulgaria160%33.18  138,733
Europe and temperate AsiaBelarus45% 9.38  129,745
Europe and temperate AsiaCroatia196%24.14   87,611
Europe and temperate AsiaCzech Republic24% 4.75  291,682
Europe and temperate AsiaEstonia23% 2.23   39,947
Europe and temperate AsiaGeorgia861%101.02   31,934
Europe and temperate AsiaHungary63%10.94  221,431
Europe and temperate AsiaLatvia8% 0.94   60,445
Europe and temperate AsiaLithuania21% 2.96   82,266
Europe and temperate AsiaMoldova60% 6.21   16,590
Europe and temperate AsiaMongolia250%21.16    9,480
Europe and temperate AsiaMontenegro151% 9.56    6,609
Europe and temperate AsiaPoland130%36.07  711,490
Europe and temperate AsiaRomania154%38.33  369,789
Europe and temperate AsiaRussia156%86.563,061,763
Europe and temperate AsiaSerbia158%30.49  131,126
Europe and temperate AsiaSlovakia45% 7.34  135,409
Europe and temperate AsiaTajikistan420%50.70   12,969
Europe and temperate AsiaUkraine83%29.42  494,016
Europe and temperate AsiaUzbekistan19% 4.17   72,098
Latin AmericaArgentina126%326.59  507,674
Latin AmericaBelize10%10.37   42,518
Latin AmericaBolivia209%197.04   12,830
Latin AmericaBrazil430%1640.982,377,579
Latin AmericaChile23%36.41  226,886
Latin AmericaColombia89%185.48  312,248
Latin AmericaCosta Rica140%111.98   32,663
Latin AmericaEcuador107%151.24   83,709
Latin AmericaEl Salvador81%91.78   41,093
Latin AmericaGuatemala110%117.48   52,264
Latin AmericaHonduras117%94.19   19,501
Latin AmericaMexico227%650.251,380,982
Latin AmericaNicaragua169%93.15    8,085
Latin AmericaPanama143%104.41   25,214
Latin AmericaParaguay48%55.72   29,559
Latin AmericaPeru102%188.65  208,293
Latin AmericaVenezuela97%180.78  299,020
Tropical AfricaBenin379%170.08   11,363
Tropical AfricaBurkina Faso206%86.12    7,780
Tropical AfricaCameroon136%78.77   30,213
Tropical AfricaCentral African Republic28% 6.60    2,529
Tropical AfricaChad85%40.37   11,869
Tropical AfricaEthiopia19%19.91   51,465
Tropical AfricaGabon208%112.58   25,761
Tropical AfricaGambia, The23% 6.81    1,458
Tropical AfricaGhana413%258.94   28,102
Tropical AfricaGuinea101%104.64   78,969
Tropical AfricaGuinea-Bissau148%23.29      565
Tropical AfricaLiberia101%18.22    1,160
Tropical AfricaMali82%34.48   11,257
Tropical AfricaNigeria81%101.19  245,354
Tropical AfricaRwanda138%65.52    8,545
Tropical AfricaSenegal121%55.35   15,468
Tropical AfricaZaire64%30.89   15,565
Savannah AfricaKenya230%637.28   51,665
Savannah AfricaLesotho116%86.00    2,327
Savannah AfricaMadagascar335%697.03   15,973
Savannah AfricaMalawi38%61.55    8,595
Savannah AfricaMozambique104%161.81   14,015
Savannah AfricaNamibia59%60.65    8,499
Savannah AfricaSouth Africa89%377.97  394,305
Savannah AfricaTanzania105%271.07   34,742
Savannah AfricaUganda113%239.89   19,753
Savannah AfricaZambia39%57.09   15,258
Savannah AfricaZimbabwe118%375.00   22,357
Southeast AsiaCambodia42%19.39   17,849
Southeast AsiaChina150%480.917,573,315
Southeast AsiaIndonesia324%493.07  888,938
Southeast AsiaLaos210%87.32   11,285
Southeast AsiaMalaysia3% 2.61  296,803
Southeast AsiaPhilippines614%637.82  275,169
Southeast AsiaVietnam139%151.68  196,755
Middle East and North AfricaAlgeria166%48.61  214,365
Middle East and North AfricaEgypt108%54.68  348,686
Middle East and North AfricaIran42%28.01  851,903
Middle East and North AfricaJordan609%73.74   20,238
Middle East and North AfricaMorocco132%29.75  122,299
Middle East and North AfricaSyria16% 3.77   74,228
Middle East and North AfricaTunisia307%63.30   65,587
Middle East and North AfricaTurkey118%49.47  916,546
Middle East and North AfricaYemen29% 5.92   41,739
South AsiaBangladesh137%184.71  159,862
South AsiaBhutan97%77.83   39,112
South AsiaIndia180%670.622,552,500
South AsiaPakistan40%75.28  359,861
South AsiaSri Lanka107%115.03   84,969
Small island developing nationsGrenada30% 2.25    1,092
Small island developing nationsHaiti276%68.98   14,918
Small island developing nationsKiribati260% 8.66      106
Small island developing nationsMauritius83%14.80   12,060
Small island developing nationsSeychelles164%17.64    1,697
Small island developing nationsWestern Samoa34% 2.20      670

Results

World Bank spending varied from roughly $US 1 million to almost $US 1 billion per country over the 20-year period. Larger economies receive substantial larger funds, but this highly significant correlation (P < 0.0002) shows considerable country-to-country variation. Using an ANCOVA, we find that regional differences are highly significant (P < 0.0001). Given those overall differences, there is no significant difference between regions in the slopes of the relationships between spending and GDP (P = 0.9). Figure 1 makes these points visually: the overall differences are clear, while the slopes of the within-region fits are all broadly similar.

Figure 1.

World Bank spending per country increases as that country's gross domestic product (GDP). Note the logarithmic scales. Spending per country differs almost one thousand-fold for a given GDP and there are consistent differences between regions (defined a priori, see text). The dashed line of constant funding shows the slope of all lines for which spending would be proportional to GDP.

These results test against the null hypothesis that spending does not increase with GDP. A more interesting null is that spending increases in direct proportion to GDP—that is, the slopes of the fits in Figure 1 would be unity. They are all significantly smaller (P < 0.0001) meaning that spending is proportionately greater for each region's poorer countries.

Countries in Savannah Africa and Madagascar receive more than three times more funding than one would expect given their GDP. Countries in Europe, Central Asia, and the Middle East got relatively small amounts of funds. Table 2 also shows how each country fares relative to the region in which we include it and when corrected for GDP. Henceforth, we call these regional and GDP-corrected numbers “adjusted” funds. We highlight those countries that receive unusually high (>200%) or low (<50%) levels of such adjusted funding given the region in which they exist and their GDP.

The countries of Savannah Africa and Madagascar receive the highest overall adjusted funding. Kenya and Madagascar have the highest adjusted funding levels of all. Zambia and Malawi fared poorly, while no funds went to Botswana. Within Latin America, Mexico, Bolivia, and especially Brazil received the largest adjusted funds, while Belize and Chile, the lowest.

Tropical Africa, South Asia, and Southeast Asia all have overall similar levels of adjusted funding. Within these countries, Benin, Ghana, Indonesia, and especially the Philippines received larger than expected funds, while Ethiopia, Gambia, Cambodia, Pakistan, and especially Malaysia, did poorly. In Papua New Guinea, which received scant investments, World Bank biodiversity assistance was limited to supporting the country's National Biodiversity Strategy and Action Plan in 1999.

There was no effect of a country's corruption index once the region and GDP were taken into consideration (P = 0.87). The corruption index scales from 1 (the worst) to 6 (the best) (TI 2008) The three lowest scores (Somalia, Iraq, and Myanmar) received no funding, but Haiti is fourth lowest and received nearly three times the adjusted funding. Fifteen of the 30 poorest scoring countries (two or less) received funds. This is essentially the same fraction as the 17 of 30 countries that received funds that scored between 4 and 6.

Discussion

Our results show that the World Bank's investments are not proportional; poorer countries receive relatively more funds than richer ones, regardless of biodiversity importance. Obviously, though World Bank funding for conservation is important, it is not the only funding for biodiversity in any particular country. Low or no funding may mean simply that a country found funds elsewhere. Or that a country was ineligible for World Bank funding. Given the paucity of public records on conservation spending, this is not an easy problem to address. Nor do the correlations found in our results necessarily confer causality in the absence of more explicit records on spending. This caveat conditions the discussion we now present.

Consensus conservation areas

Tropical moist forests hold the majority of the world's terrestrial species and they form 15 of 25 of the biodiversity hotspots—areas defined to have high levels of endemism coupled with high levels of deforestation. They are: (1) Colombia, Ecuador, and Peru—which comprise the tropical Andes and Choco/Darien hotspot, (2) Indonesia, Malaysia—which comprise the Sundaland and Wallacea hotspot—and Papua New Guinea, (3) Madagascar (a hotspot), (4) Brazil, which contains the Amazon forest—the largest remaining block of tropical moist forest—and the cerrado (a dry forest) and Atlantic coast forest hotspots, (5) the various Caribbean Island countries, (6) Myanmar, Thailand, Cambodia, Vietnam, Laos, and tropical China, which constitute the Indo-Burma hotspot, (7) the Philippines, (a hotspot), and (8) the countries from Panama north to southern Mexico that constitute the Mesoamerica hotspot.

Brazil, the Philippines, and Madagascar rank 1, 3, and 6 in actual funding, and 5, 2, and 9 in adjusted funding. This suggests that conservation priority is important in funding allocation. Thereafter, the results are decidedly mixed. Columbia, Peru, and Ecuador rank 14, 15, and 22 in actual funding, 62, 57, and 62 in adjusted funding—this does not fully reflect their conservation significance. In Mesoamerica, Mexico ranks 2 in actual and 18 in adjusted funding, and this refers to all of Mexico, of course. But the other countries span from 20 (Costa Rica) to 77 (Belize) in actual funding and 25 (Nicaragua) to 94 (Belize) in adjusted funding.

In Asia, Indonesia ranks 8 and 10 for actual and adjusted spending, but of its neighbors, Malaysia ranks 92 and 96 and Papua New Guinea received no money at all. Thailand and Myanmar received no funds, while Laos, Cambodia, and Vietnam ranked from 30 to 76 in actual funding and 19 to 77 in adjusted funding.

Other countries have tropical moist forests, but again no simple pattern emerges to suggest that conservation priorities are the primary driver of funding. One hotspot, New Caledonia, is considered part of France and so cannot receive funding under Bank rules.

Outside of the Mediterranean, European countries are on no international list of priorities and the amount of funding they receive is also low.

In addition to the hotspots already mentioned, there are three regions with Mediterranean ecosystems that have large numbers of species at risk. The southwest of Australia cannot receive funding. South Africa holds two of these nonforest hotspots, the Cape Floristic Province and the Succulent Karoo (Myers et al. 2000). It ranks 5 in actual funding, though only 63 in adjusted funding because countries in Savannah Africa enjoy exceptionally high levels of funding (Figure 1, Table 1). The third region is the Mediterranean itself. As Figure 1 shows, these countries receive exceptionally few funds.

Finally, the Caucasus region is a biodiversity hotspot. The two large countries involved, Georgia and Azerbaijan, rank highly for adjusted funding 1 and 11. (They are, however, in the group of European and Asian countries, which constitute the poorest funded region overall; Table 1).

Investing where there is most need

Our results clearly show that poorer countries within regions receive proportionally more funding than wealthier countries, if not in absolute terms, suggesting that the Bank does consider need in its decision-making. However, while this may hold within regions, the same is not true between regions. Although Savannah Africa and Madagascar receive the most investments overall, small island developing systems receive relatively little despite their needs.

Investing to alleviate poverty and induce development

Across all regions, the poorest countries receive the proportionally greater funds, relative to their GDP, suggesting this is a major consideration for the World Bank.

The countries in what we have called Savannah Africa enjoy the highest overall levels of actual funding (Figure 1). For example, Kenya, South Africa, Tanzania, Uganda, Mozambique, Lesotho, Namibia, Zambia, and Malawi rank 4, 5, 13, 16, 17, 18, 27, 33, 42, and 54, respectively. Botswana does not receive funds, but then it has the second highest per capita GDP in sub-Saharan Africa after Equatorial Guinea (IMF 2008). All these countries have extensive, famous game parks that attract substantial numbers of international visitors each year. On average, tourism constitutes 8.3% of these countries’ (and Madagascar's) GDP, compared to only 5.4% for Tropical African countries (WTTC 2008). Not all tourists are ecotourists, of course. Nonetheless, these data strongly suggest that the high investments in these countries reflect the likelihood that they can more easily benefit economically from conservation outcomes and affect poverty accordingly. This suggests that World Bank funding for biodiversity conservation focuses on interventions in areas where effective conservation is more likely given the direct connection between saving nature and ginning economic growth.

Investing where there is most chance of success

There is no evidence to suggest that countries with lower-cost structures receive more investments. For example, in Southeast Asia, while Laos receives disproportionally more funds than Cambodia, they have similar cost structures. Moreover, corruption levels do not correlate with spending patterns, suggesting that investment decisions are not made based on supporting countries with better governance.

In sum, while our results show that it is possible to identify method to biodiversity spending by the World Bank, our research reveals the futility in trying to simply compile data on how various institutions fund protected areas in developing countries. Given the uncertainties, we find that the World Bank allocates increasingly more funds to countries with progressively larger GDPs, but not proportionately so. Poorer countries receive relatively more funds than richer ones. There are substantial regional differences. These only poorly correlate with consensus opinions of which countries are priorities, where the most endangered species live, or where the most need has been identified. A rather more compelling explanation is that the World Bank favors countries where protected areas will likely generate substantial revenues for local and national constituencies.

More generally, there has been a shift in official donor and government priorities away from biodiversity conservation and protected areas (Lapham & Livermore 2003; Emerton et al. 2006). Following the Millennium Summit of 2000 and the 2002 World Summit on Sustainable Development, poverty reduction has become the overriding focus that guides both international assistance to developing countries and the allocation of budgets at national levels in the developing world. Accordingly, international support for biodiversity conservation is increasingly driven by social and economic objectives, and especially by its touted ability to contribute to poverty reduction (Scherl et al. 2004; Locke & Dearden 2005). Given the Bank's mission statement and this shift, our conclusion would appear to fit the data and the circumstances.

Ancillary