The inflationary effects of stochastic resource revenues in resource-rich economies with less well-developed financial markets

Abstract

This article develops a simple Dynamic Stochastic General Equilibrium (DSGE) model to illustrate how economies that face restrictions in their ability to alter both government spending and taxation in the short run and cannot borrow easily (perhaps because of incomplete internal capital markets) can find external fluctuations in resource revenues producing unexpected variations in their internal money supply and ultimately in their inflation rate. The main channels for these effects run through the government budget and through the country's balance of payments position. The model is calibrated to illustrate the case of Iran.

Keywords:

• monetary stability
• stochastic oil revenues
• dynamic stochastic general equilibrium model
• passive and active government policies

JEL Classification:

• C61
• E13
• E17
• E37

Acknowledgements

We thank Patrick Coe, Hashmat Khan, Akbar Komijani, Deming Luo, Makoto Nirei, Asghar Shahmoradi and Ali Taiebnia for their comments on an earlier version of this article.

Notes

¹ For a number of countries, oil provides a large share of government revenue. For example, the governments of Nigeria, Angola and Saudi Arabia derive over 70% of their budget revenue from oil. This falls only slightly to 60% for countries like Kuwait and Algeria while remaining as high as 50% for others such as Sudan and Azerbaijan. Oil provides roughly 33% of Mexican government revenue while even Russia relies on oil to fund as much as 28% of its expenditures. See Oviedo-Cruz (2005).

² The stocks are subscripted to denote their value at the beginning of period t so that the asset values chosen by households in period t are the values they will have to begin period t + 1.

³ While we allow for the possibility that , the absence of capital in the government sector in the model and the incentive to maximize lifetime utility by consuming financial wealth argue for . This condition generates problems for linearization later in the analysis. See footnote 4.

⁴ Since , it follows that which in turn equals zero if π^ss  = 0.

⁵ When b^p (ss) = 0, Equation 13 becomes equal to

⁶ This also implies that all the repatriated oil revenue, , will be used by the government to pay off privately held government debt. To the extent that some portion of the budget surplus is used to buy back government debt from the central bank, the money supply will be decreased correspondingly.

⁷ Because b^p (ss) = 0, the usual log linearization of the system cannot be undertaken (since it requires division by zero). We proceeded by taking the total differential of the equations, where the partial derivatives are evaluated at their steady state value, and then use db^p _t  = b^p _t  − b^p (ss) = b^p _t . The resulting set of differential equations then incorporates the contemporaneous, lagged and forwarded values of the percentage changes of all variables except for private holdings of government bonds.

⁸ Notes for all these derivations are available from the authors.

⁹ The DSGE model uses only a subset of the full model's parameters. Some parameters, such as μ and γ, determine the level of, but not the nature of the deviations from, the steady state.

¹⁰ Changes in the value of θ change the scale of the impulse response functions without changing its characteristic shape. Increasing the value of θ to 0.3, for example, raises the size of the initial inflationary response to a negative oil shock in period 1 from 0.4 to over 0.5 while reducing correspondingly the effect of a positive oil shock.

¹¹ The actual simulations used the method of Blanchard and Kahn (1980). The formal setup of the solution and the programs used are available from the authors. All data were collected from economic times series made available by the Central Bank of the Islamic Republic of Iran. The raw data can be found at http://www.cbi.ir/page/4275.aspx.