Countries making resource discoveries often find themselves facing a number of important and difficult decisions. One such decision regards whether to set up a sovereign wealth fund – or more generally, determining a time profile for the usage of the resource income. A large increase in government spending following resource discoveries often entails political risks, inefficient investments and increased volatility.
Setting up a sovereign wealth fund with a clear spending constraint may decrease these risks. On the other hand, in a capital scarce developing economy like Guyana; with limited access to international borrowing, such a spending constraint may lower welfare by reducing domestic capital accumulation and hindering consumption increases for the currently poor. These two contradicting considerations pose a dilemma for policy makers in deciding whether to set up a sovereign wealth fund with a spending constraint.
A standard recommendation is that a capital-scarce developing country, facing a borrowing constraint, should use a large share of its resource revenue to boost current spending (van der Ploeg and Venables 2011). This is of course intuitive given that both marginal utility from consumption of the currently poor generation and the return from domestic investments is relatively high. Thus, a “spend-as-you-go” scheme – where all of the oil revenue is spent as it arrives – may seem appropriate to increase spending both on current consumption and domestic capital formation.
On the other hand, a large inflow of income from natural resources often leads to corruption and various negative political effects as politicians, officials and elites try to get part of the resource rents (Vicente 2010). These effects, in turn, trigger economic stagnation, inequality and sometimes even armed conflicts.
While Guyana, specifically, seems to be rather stable, there are worrying signs where the process around exploration rights lacks transparency and the appropriate checks and balances. An example of lack of transparency is the sharing agreements of oil revenues in Guyana: they were not public, at the initial stages. This had the effect of limiting the ability of the media and the citizens of Guyana to scrutinize the agreements, investigate whether they contain problematic elements and whether they address all the important issues in a proper way. Furthermore, to avoid considerations not in line with the welfare of the Guyanese people, i.e., an undue influence on the oil exploration and the extraction process, it would be preferred if the rules governing the formation and signing of agreements were drawn by the minister, approved by parliament and executed by an independent authority.
A closely related problem in resource-rich countries is that large revenues may make spending decisions worse from a social point of view (Ades and Di Tella 1999). Adverse partisan influence over these decisions is often hard to avoid. First of all, oil income creates the risk of using more of the revenues for public spending in election years to boost the popularity of incumbent politicians. There is substantial evidence that government consumption in developing countries is pro-cyclical and that this increases business cycle volatility (Ilzetzki and Vegh 2008). This is likely to have negative effects on growth, in particular for Guyana which has a less developed financial markets. A related issue is the selection of specific investment projects.
Earlier research has documented that spending of resource revenues quite often goes toward projects with low returns, motivated by pleasing various political groups or electorates (Isham and Kaufmann, 1999). This is related to the time profile of spending as the capacity to absorb large funds for investment is often insufficient in developing countries (van der Ploeg and Venables 2012), which calls for postponing the domestic usage.
Given these potential negative effects of high spending, for a developing country such as Guyana, it may be important to either constrain the extraction of oil or the government’s usage of the proceeds (Segura 2006). Even though part of the motivation for this recommendation tends to lie in the theoretical explanation forwarded by the permanent income hypothesis, the concern about negative political consequences is also cited as an important justification for constraining spending. However, this is of course inconsistent with the recommendation that a poor country should increase its current spending in order to mitigate capital scarcity and low consumption by the current generation. Which of these two contradicting aspects should be prioritized by Guyana remains an open question. To take a step towards answering this normative question, Guyana should analyse quantitatively the potential loss of constraining aggregate spending from resource revenues. Spending can be constrained in two ways, either by:
- postponing extraction
- postponing the usage of the proceeds.
The first option, of postponing extraction, will in most realistic circumstances reduce the total economic value of the oil. If Guyana analyses how large, quantitatively, the potential losses are of deviating from the profit-maximizing extraction path, the results will highlight that postponing extraction will generally entail large losses relative to the maximum value that can be obtained. Hence, this seems to be an unattractive solution to the problem.
The second option, to constrain government spending, can be done by investing the resource revenues in a sovereign wealth fund, and utilizing a simple and transparent rule that stipulates annual spending to be a fixed share of the fund’s total value. While such a setup may not always be in the interest of politicians and elites, it is, under some circumstances, a realistic outcome. A simple and rigid framework, such as a fund and a spending rule, partly ties the hands of future politicians by making the breaking of the rule easy to detect for media, international bodies and non-incumbent politicians (Bacon and Tordo 2006). In particular, it is simpler to implement than a theoretically optimal rule that takes into account all future revenue flows, since this requires forecasting the notoriously unpredictable oil price (Landon and Smith 2015). The uncertainties in such predictions may undermine transparency in spending decisions which is likely to create a bias towards spending and unsustainable borrowing by incumbent politicians.
Moreover, to the extent that current politicians can constrain the actions of future politicians (through, say, constitutional mechanisms), it may be in their interest to set up a spending rule to minimize over-spending by their opponents in the future. A spending constraint can be implemented in more or less binding ways. For instance, as a consensual hand-shake between parties in parliament (as in Norway) or in more binding fashion through formal rules. It should be emphasized that it is the constraint for how much has to be put in and how much can be taken out from the fund that mitigates partisan politics and other political effects. It is not the fund as such.
Clearly, for a capital-scarce and borrowing-constrained developing Guyana, where the marginal utility of consumption by the current generation is relatively high and investment needs are many, constraining current spending through spending rules involves welfare losses. But how large are these losses? To answer this question, Guyana must give consideration to building a macroeconomic framework which is well suited for the evolving economy, of capital scarcity and borrowing constraints. It must contain the intratemporal trade-off between public and private capital and the intertemporal trade-off between capital and consumption. It should further allow for population growth, investment frictions, capital scarcity, borrowing constraints and technical change – features that have significant relevance for developing the Guyanese economy.
Perhaps surprisingly, in any baseline calibration of the economic framework for Guyana (abstracting from any political side effects), and from a consumption smoothing perspective, a shift from a spend-as-you go scheme to using an oil fund along with a fairly strict spending rule will entail only a marginal, if any, welfare loss. In particular, the losses should appear small compared to losses of fairly mild political side effects that may arise as a result of increased oil spending. For instance, if the spend-as-you-go scheme implies a lag of structural transformation by one year, or if it retards annual productivity growth by as little as possible percentage points, then the fund is preferable. This implies that, considering the potential negative political and economic side effects of a drastic increase in oil spending, the case for constructing a sovereign wealth fund along with a spending rule is rather strong.
Furthermore, the most serious political side effects of resources, such as civil conflict as suggested by Garfinkel and Skaperdas (2007), should imply welfare losses of several orders of magnitude larger. Thus, to the extent that a spending constraint can reduce the risk of these events, it would be well worth the small loss of consumption smoothing. Put differently, getting the political checks and balances right is more important than getting the spending profile right.
Therefore, in developing the Guyanese economy, the merit of maximising current spending from resource revenues is that it provides an opportunity to increase the consumption of the currently poor population. It also relaxes potential borrowing constraints and thereby enables increasing investment in domestic capital with high rates of return. Thus, constraining spending by utilizing a sovereign wealth fund with a restrictive spending rule may involve a welfare loss. Hence, given that having a sovereign wealth fund along with a simple, but rigid, spending constraint provides more transparency and helps decrease the potential negative political side effects associated with increased spending, adopting such a construction seems like a well-motivated policy measure.
In light of the pronounced volatility of the oil price and the substantial evidence that government consumption in developing countries is pro-cyclical; precautionary saving provides additional motivation for using a fund instead of spending revenues as they arrive (Cherif and Hasanov 2013).
One possibility is to use two funds: one for constraining spending over the long run and one with the specific purpose of smoothing government budgets in the short and medium run. Such a construction has, for instance, been used in Ghana (van der Ploeg et al 2012) and could prove useful in Guyana, as well; whilst focusing on the long-run costs and benefits of a sovereign wealth fund. Caution must also be abstracted from other negative effects of resources, such as the potential for the Dutch Disease harming manufacturing exports or leading to increased income inequality. Incorporating these issues into the macroeconomic model of Guyana; would favour the fund and spending-constraint construction as these negative effects would be delayed.