The Oil-to-Cash policy initiative by the Center for Global Development to help the lot of Africa’s poor presumes so. This speculative but highly plausible initiative may actually prove useful as a fiscal policy instrument that African governments may deploy to bridge the ever-growing income inequality in the continent; and better yet, help create the ever-useful middle class. The research paper by Marcelo Guigale and Nga Thi Viet Nguyen entitled “Money to the People: Estimates of the Potential Scale of Direct Dividend Payments in Africa,” is both informative and timely. A summary version of this work is presented here.

Marcelo Giugale and Nga Thi Viet Nguyen:

Historical data shows that large natural resource endowments have not translated into better quality of life in Sub-Saharan Africa (“Africa” for short). The problem is becoming more urgent, as new exploration technologies are rapidly expanding the number of countries whose fiscal revenues will grow, in many cases massively, with new oil, gas, and mineral discoveries. A search is on for innovative approaches in managing this commodity bonanza. This paper focuses on the distribution of resource rents as cash transfers to citizens, so-called “Direct Dividend Payments” (DDPs). It expands on recent related literature by calculating such transfers, whether universal or targeted, for every African country for which data is available, and compares them to measures of poverty depth under both national and global definitions. Furthermore, it extends the analysis to a different kind of resource flow enjoyed by most African countries— foreign aid. We found that DDPs can account for a large proportion of the income Africa’s poor need to step over the poverty line.


The discovery of oil in a developing country is potentially beneficial and, simultaneously, potentially calamitous. While countries could put oil revenues toward building much-needed schools and roads, fixing and staffing health systems, and policing the streets, many resource-rich states fare little better —and often much worse—than their resource-poor counterparts. Too often public money is misallocated and funds meant to be saved are raided, and those living in poor resource-rich countries pay the price. While this so-called resource curse is well established in the literature, solutions to counteract its corrosive effects remain highly elusive. CGD’s Oil-to-Cash initiative is exploring one policy option that may address the root mechanism of the resource curse: using cash transfers to hand the money directly to citizens and thereby protect the social contract between the government and its people. Under this proposal, a government would transfer some or all of the revenue from natural resource extraction to citizens in universal, transparent, and regular payments. The state would treat these payments as normal income and tax it accordingly—thus forcing the state to collect taxes, and adding additional pressure for public accountability and more responsible resource management.

This paper by Marcelo Giugale and Nga Thi Viet Nguyen, commissioned by CGD as part of Oil-to-Cash, calculates the potential scale of resource-linked transfers for every African country (for which data is available) and compares these levels to poverty depth estimates. They make a similar calculation for inward aid flows. Thus the authors make a contribution to the literature by providing a sense of how important such transfers might be, at least theoretically, to increasing incomes of Africa’s poor over the poverty line.

  1. Introduction

The past 20 years have witnessed fast and sustained economic growth in Africa, especially in resource-rich countries, thanks to improved macroeconomic policies, buoyant commodity prices, and new mineral resource discoveries [World Bank (2013); Hostland and Giugale (2013)]. Despite such progress, poverty levels remain stubbornly high, and recent studies have shown that the current pace of economic growth and poverty reduction will not be enough to bring extreme poverty below 3 percent by 2030, neither globally nor in Africa [see Dabalen and Nguyen (2013); Chandy et al. (2013)]. More fundamentally, basic human development outcomes have been particularly dismal among African countries with large natural wealth. Even when their per-capita GDP grew more slowly, resource-poor countries in Africa outperformed their resource-rich peers in extreme poverty reduction and, controlling for income level, did better in measures like life expectancy and child mortality. For example, given their similar per capita income levels, life expectancy in oil-exporter Cameroon is on average 7.7 years shorter than in Senegal. And primary school completion rate in Chad is less than half of Rwanda’s despite the countries’ similar per capita income [World Bank (2013)].

This seeming inability to turn resource rent into poverty-reducing development has intensified the search for new approaches to manage commodity revenues. One of those approaches is Direct Dividend Payments (DDPs), that is, the distribution of part of the resource rent that would otherwise accrue to governments, directly to citizens. The idea is not new, and the literature contains arguments for and against DDPs. They are ably reviewed in Moss and Majerowicz (2013). The debate is mainly about technical and political feasibility, individual identification, conditionality, public goods production, macroeconomic implications, effect on governance, progressivity, taxation, fiscal sustainability, social impact, and behavior of the beneficiaries.

This paper contributes to that debate by providing a piece of information that has so far been missing: a simple but comprehensive calculation that illustrates how DDPs would actually look in Africa. It extends previous work by Devarajan and Giugale (2013) in three ways. First, it covers all resource-rich countries in the region for which data is available. Second, it compares DDPs to both national and global poverty lines—the latter being 1.25 PPP dollars per person per day in 2005 prices. And third, it calculates DDPs from both natural-resource fiscal revenues and foreign aid flows. The objective is not to prescribe DDPs for any one country or of any one size. Rather, we explore how relatively-modest, universal DPPs (say, ten percent of fiscal resource revenue) compare with poverty levels, and how costly, in terms of foregone fiscal revenue, DDPs would be if they were perfectly targeted to raise the income of the poor up to the poverty line. We then replicate the calculation using official development assistance (ODA), rather than resource revenues, as the means of funding the DDPs.

  1. The Literature on DDPs

Sala-i-Martin and Subramanian (2003) were among the first to call for DDPs in the context of a resource-rich developing country, in their case as a means to compensate for the poor governance of oil revenues in Nigeria. Their underlying reason is that resource revenues go directly from extracting companies to governments, without citizen involvement—people do not have full information about the rent that is being extracted. This weakens their incentive to scrutinize government expenditures and, thus, fosters corruption. The process is reinforced by the fact that the larger the resource revenues, the less need for taxation and, thus, lesser accountability to taxpayers for the use of public funds [Bornhorst et al. (2009); McGuirk (2010)]. This lies behind the proposal by Devarajan et al. (2012) that resource-rich governments transfer some or even all of their resource revenues directly to their citizens and then tax them back to finance public spending. The case is further made by Arezki et al. (2012). They find that, as the size of the resource windfalls increases in countries with weak administrative capacity, the optimal spending policy should put more emphasis on redistribution and less on public investment. This is based on the assumption that adjustment costs, reflecting the limited administrative capacity, increase with the size of the resource windfalls.

Falkinger and Grossmann (2005) take a different tack. They submit that a more equal distribution of resource rents promotes economic growth and structural change by facilitating investments by credit-constrained entrepreneurs. This shifts the distribution of political power from public officials toward a new business class, resulting in an economic and political environment more favorable to productivity gains. The idea is given indirect backing by Segal (2011) who uses a global dataset on resource rents and distribution of income to claim that, under certain conditions, DDPs could cut the number of people living under US$1 a day by up to 66 percent. A number of more recent studies also argue that resource-rich countries, including Iraq, Nigeria, Uganda, and Ghana, should adopt DDPs as a way to accelerate political and economic transformation and a new social contract [see, for example, Sala-i-Martin and Subramanian (2013); Gelb and Majerowicz (2011); Moss and Young (2009); Sandbu (2006); Birdsall and Subramanian (2004); Palley (2003)]. These papers all carry an implicit sense of urgency with regards to Africa: with the help of new technologies in exploration and extraction, over the next ten years the region is likely to experience a massive wave of new oil and gas discoveries from the East African Rift Valley to West Africa’s Gulf of Guinea [Diamond and Mosbacher (2013)].

II. Calculating DDPs in Africa: Methodology

The calculations presented in this paper are only meant to provide an order of magnitude to possible DDPs in Africa. As such, they ignore any improvement in governance that DDPs may trigger, assume a zero opportunity cost for the funds used to pay for DDPs, and do not incorporate the economy-wide impacts of putting money in the hands of the poor. In other words, they ignore the net impact on baseline poverty of possible improvements in the quality of public expenditures, contractions in the quantity of public investment, distributional effects on aggregate consumption, and the related changes in relative prices.

Still, because of data paucity, even order-of-magnitude calculations are challenging in Africa. We present our data in Annex 1. We used the World Development Indicators (WDI) 2013 as the primary source on country population, GDP in current US dollars, net ODA per capita in current US dollars, poverty headcount ratios, and poverty gaps under either national or the international poverty lines. National poverty lines are as provided by the countries’ national statistical offices.

Fiscal revenue from natural resources is an indicator that needs to be taken with special caution since its definition varies widely across sources. We use the IMF Article IV Consultation Reports and Country Reports as our primary source. The IMF defines revenues from non-renewable resources as (i) royalties, (ii) income from profit sharing agreements, (iii) dividends or other payments from national resource companies, and (iv) taxes on resource profits or production. When information from the IMF is unavailable, we use the Extractive Industries Transparency Initiative (EITI), which provides resource revenues broken down by categories such as corporate tax, dividend, royalty, property rent, and licenses. When IMF and EITI figures are unavailable, we use government reports.

We picked 2011 as the reference year for all indicators except the poverty rate. Household surveys are carried out non-concurrently across countries and, on the whole, infrequently (for example, data was collected in Guinea in 2012 and Zimbabwe in 2011, but it dates back to 2003 in Botswana and Lesotho). We thus use the most up to date surveys available and assume that poverty rates, either using the national or international definitions, remained unchanged until 2011. This is, in practice, a conservative assumption, as all countries in Africa have experienced positive economic growth in the period since their last household survey.

For each country, the WDI provides data on the poverty gap as defined by Foster, Greer, and Thorbecke (1984). That gap is calculated as the sum of the distances between each poor person’s income or consumption and the poverty line, divided by the total size of the population, whether poor or not. In that sense, it represents a hypothetical average contribution that every member of society would have to make to end poverty. We use that information to compute the average poverty “depth” as defined by Devarajan and Giugale (2013). The average poverty depth is also calculated as the sum of the distances between each poor person’s income and the poverty line, but divided by the size of the poor population only. It thus reflects the transfer that the average poor person needs to receive to reach the poverty line. This makes it the right measure to compare against DDPs.

  1. DDPs in Africa: Results

As expected, wide heterogeneity in resource endowments, foreign aid flows, population sizes, and poverty depth across Africa translates into equally wide heterogeneity in how large DDPs are in relation to poverty, and how expensive in relation to fiscal revenues.

a) Natural Resources, National Poverty Line

Say that governments decide to distribute ten percent of their natural-resource fiscal revenues equally among all citizens, rich or poor. How big would these uniform and universal transfers be compared to the average poverty depth, that is, to the money needed to bring the average poor person up to the national poverty line? Only in three countries (Angola, Equatorial Guinea, and Gabon) would that transfer document a positive growth elasticity of poverty, and amount to half or more of the average poverty depth. Two more countries (Republic of Congo and Nigeria) join that group when the 10 percent DDP is distributed only among the poor.

What if the comparison is not against half or more of the average poverty depth but, say, a tenth of it? Using that standard, DDPs of ten percent of resource-related fiscal revenue would make the cut in eight African countries (Angola, Botswana, Chad, Republic of Congo, Equatorial Guinea, Gabon, Nigeria, and South Sudan) when universally distributed, and in twelve when given only to the poor (Cote d’Ivoire, South Africa, South Sudan, and Sudan would join the group). These are not negligible numbers as, depending on definitions, the total number of resource-rich countries in the region currently stands at about 30.

A related question is how expensive it would be to “eradicate” poverty. That is, what proportion of natural-resource fiscal revenues would need to be transferred in a perfectly- targeted way to raise the income of every poor person up to the poverty line? In a few cases (Angola, Equatorial Guinea, Gabon), it would be extremely cheap—six percent of revenues. In some (Botswana, Chad, Republic of Congo, Nigeria, South Sudan), it would be more expensive—between a tenth and a third of revenue. But in most, it would be unaffordable— more than 100 percent.

A point of note. Because of their country’s relatively large resource revenues and very small population size of less than one million, a universal DDP of ten percent of those revenues would give citizens of Equatorial Guinea the highest absolute payment in the region in US dollars—approximately US$ 765 per person per year. This amount would be more than 20 percent larger than the size of the average poverty depth. Similarly, only six percent of the resource revenues would be needed to bring every poor Equatorial Guinean up to the national poverty line. That would be no small achievement given that more than three quarters of Equatorial Guineans are currently living in poverty. Nigeria, on the other hand, while benefitting from resource revenues that are ten times the size of those of Equatorial Guinea, has a population that is more than two hundred times larger. Consequently, a universal DDP at ten percent of revenue would be significantly lower—around US$35 per capita per year. And yet, it would cover half of the amount needed to get the average poor person out of poverty. But it would take a fifth of the resource revenue to eliminate poverty in Nigeria all together. The point is clear: the impact of DDPs depends as much on the volume of natural resource riches as it does on demographics and the initial position of the national poverty line. The following section alters that, by using the international definition of extreme poverty, rather than the national ones.

b) Natural Resources, International Poverty Line of $1.25 per day at 2005 international prices How does using the international extreme poverty line of 1.25 PPP dollars per day per

person (in 2005 prices), instead of each country’s own poverty line, change the size of DDPs relative to poverty depth and fiscal revenue? It does not change the results much.

c) Official Development Assistance (ODA), National Poverty Line

The funding of DDPs need not come from natural wealth. Conceivably, it can come from another of Africa’s resources—its donors. They contributed some US$ 43 billion, or just over 3 percent of the Region’s GDP, in 2011. This is, on average, equivalent to about a third of the fiscal revenues received from natural resource exploitation (10.4 percent of GDP in 2011). Africa’s ODA has proven fairly stable in nominal terms although, as a proportion of regional GDP, is has been in gradual decline since 2000. A universal, uniform distribution of 10 percent of ODA would represent half or more of the average poverty depth in only one country (Sao Tome & Principe). Focusing that 10 percent of ODA only on the poor would add just two countries (Cape Verde and Rwanda).

If the coverage sought is not half but a tenth of the average poverty depth, those kinds of ODA-funded transfers would make the cut in six countries (Cape Verde, Cote d’Ivoire, Rwanda, Sao Tome & Principe, Sierra Leone, and Tanzania) if they are distributed to all citizens, and in 19 countries if they are distributed only among the poor (the previously- mentioned ones plus Benin, Burkina Faso, Ethiopia, Ghana, Liberia, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, South Sudan, and Uganda).

Remarkably, there are 18 countries in which the flow of ODA would be more than enough to bring everyone up to the national poverty line. In fact, in 11 of those countries half or less of the ODA flow would suffice (Benin, Cape Verde, Cote d’Ivoire, Ethiopia, Mauritius, Namibia, Rwanda, Sao Tome & Principe, Sierra Leone, Tanzania, and Uganda).

Official Development Assistance (ODA), International Poverty Line of $1.25 per day at 2005 international prices.

Finally, in no country will a distribution of a tenth of the ODA uniformly across all citizens suffice to cover half or more of the average poverty depth, when poverty is defined as $1.25 PPP dollars per day per person. If, instead, that tenth of the ODA is distributed only among the poor, the transfer would cover half or more of the average poverty depth in six countries (Cameroon, Cape Verde, Gabon, Mauritania, Sao Tome & Principe, and Seychelles). A ten-percent DDP distributed universally and uniformly continues to cover half or more of the average poverty depth only in three countries (Angola, Republic of Congo, and Gabon). And focusing the DDP only on the poor, again adds only two more countries to that list (the two additional countries are Cameroon and South Africa). Notably, Nigeria now drops out of the list, as the 1.25 PPP dollar line is, in fact, higher than the national poverty line. The change from national to international poverty line does not alter the lists of countries when the DDP is compared with a tenth of the average poverty depth. In that case, DDPs of ten percent of revenue would “work” in seven countries (Angola, Cameroon, Chad, Republic of Congo, Gabon, Nigeria, and Sudan) when given out universally, and in twelve (add Cote d’Ivoire, Ghana, Mauritania, Namibia, and South Africa) when distributed only among the poor.

How costly is it to bring everyone up to the international, instead of the national, poverty line? Only in four countries (Angola, Cameroon, Gabon, and Republic of Congo) would it cost ten percent or less of fiscal resource revenue. For all other countries in Table 2, except South Africa and Sudan, the DDP needed to “end poverty” would represent more than a third of resource revenue.


Note (1): SSA countries with no or insignificant fiscal revenues coming from natural resources in 2011 are excluded from this list. These countries are Benin, Burkina Faso, Burundi, Cape Verde, Comoros, Eritrea, Ethiopia, Gambia, Guinea-Bissau, Kenya, Lesotho, Madagascar, Malawi, Mauritius, Rwanda, Sao Tome and Principe, Senegal, Seychelles, Somalia, and Swaziland. Among those, many are expected to have large resource revenues flowing in the near future, for example, Ethiopia, Kenya, Malawi, Mauritius, Gambia, and Senegal (see Diamond and Mosbacher, 2013).


Note (2): Poverty rates and gaps at US$1.25-a-day (PPP 2005) are not available for some resource-rich countries (namely, Botswana, Equatorial Guinea, South Sudan, and Zimbabwe). Consumer Price Index is not available for Sierra Leone in 2005. Therefore, our calculations cannot be applied to this country. More to the point, in those same 27 countries the flow of ODA is more than sufficient to raise everyone up to the international poverty line. In fact, in 14 of them, just half or less of the ODA would be sufficient (Cameroon, Cape Verde, Cote d’Ivoire, Ethiopia, Gabon, Gambia, Ghana, Mauritania, Namibia, Sao Tome & Principe, Senegal, Seychelles, South Africa, and Togo).

Putting it all together

First, for a few countries, DDPs can be both extremely large (relative to poverty depth) and extremely cheap (relative to resource revenues). In places like Angola, Cameroon, Republic of Congo, Equatorial Guinea, and Gabon, even universal DDPs that take up a tenth or less of the natural resource revenue can make a major contribution to poverty alleviation efforts—and in some cases, suffice to raise the income of the poor up to the poverty line. And third, in about a third of all African countries, ODA is more than sufficient to lift everyone’s income above the poverty line, assuming perfectly-targeted DDPs are possible. In fact, for about a dozen countries, less than half of the ODA flow would be enough. This calls attention to the funding source of DDPs, for African countries that lack natural resource rents usually get relatively large aid flows. In sum, the quantitative analysis indicates that DDPs have obvious country candidates, can help with poverty alleviation, and need not be funded by natural wealth. Second, in a few countries, DDPs that are tailored to cover exactly the poverty depth of each poor person (“perfect targeting”) can be a potentially powerful tool to cut poverty headcounts, while accounting for only a small share of revenue. That is true whether the DDPs are funded through natural resource rents or official donors.

Conclusions: Value and Limitations of the Analysis

The calculations presented in this paper suggest that DDPs, at least in terms of relative size and cost, could be a powerful new tool in poverty alleviation among African countries. But, while helpful as an indication of orders of magnitude, this analysis has both conceptual and methodological limitations.

First, transfers by themselves do not ensure poverty reduction, as they may, and probably will, have second-order effects on the income of the poor, both positive and negative. That is, of course, also true of the more traditional Conditional Cash Transfer programs (CCTs) currently deployed in some 70 developing countries, 35 of which are African. In fact, the only difference between DDPs and CCTs is that the latter require a specific behavior by the recipient—say, consuming basic health services—and are not explicitly linked to any specific source of funding. Money being fungible, CCTs may actually be funded with fiscal resource rents, especially in resource-rich countries. Second, DDPs do not “work” in all countries, in that they may be too small to make a difference to the recipients or too large for a government to afford them—especially those governments that are unable to pay for basic public goods. At the same time, for countries whose governments have enjoyed large resource rents for a long time and where poverty remains stubbornly high, DDPs could be an interesting game-changer.

Third, for the purpose of cross-country comparison, the figures shown in this paper correspond to a single point in time—the year 2011. But fiscal resource rents can, and do, fluctuate significantly. When computed for a single country, DDPs should optimally be calculated on the basis of structural, long-term flows. For most African countries, such data does not yet exist. Fourth, while ignored in this paper, the political economy of DDPs is complex. DDPs imply a reduction in the discretionary power of incumbent governments to allocate rents, say, through public employment or price subsidies. And the choice between universal transfers and transfers focused exclusively on the poor is a major societal decision. Whether in practice those issues can be arbitrated by political contestability, enhanced citizen information, or both, remains to be seen.

Finally, drastic resource price changes or major resource discoveries may quickly render this paper’s calculations obsolete. In that sense, they should be taken only as indications of potential magnitudes. While prices are not expected to rise in real terms in the medium-term, and may in fact begin to fall, quantities are bound to expand on the wake of faster, cheaper and cleaner exploration and exploitation technologies. The net effect on fiscal resource revenues as a source of DDPs is uncertain.