فساد بزرگ به جای تعهد؟ تجدید نظر در زمان تناقض سیاست های پولی

This paper suggests that inflation may be affected differently by grand corruption compared to its positive nexus with petty corruption. In an extended Barro and Gordon (1983a) model grand corruption may serve as a quasi-commitment device: a cheating (expropriating) government may actually deter a monetary authority from cheating (reneging). Furthermore, Rogoff”s (1985) conservative central banker has an unambiguously beneficial effect; she reduces the inflationary bias even more while also rendering fiscal policy more effective. The model nests the standard fiscal–monetary interaction logic with and without expropriation as well as the diametrical “symbiosis” result obtained by Dixit and Lambertini (2003a).

What is the nexus between corruption and inflation? Theory and empirical evidence suggest that there may be a positive effect of petty corruption on inflation, but not much is known about the impact of grand corruption on inflation. As for empirical evidence, there seems to be some indication of a positive link, overall, although the body of literature is sparce, for instance Al-Marhubi (2000) and Smith-Hillman (2007).1 A conceptual problem is that empirical studies do not take account of the difference between petty corruption on the one hand, and grand corruption (henceforth also called expropriation) on the other hand.2 In fact, the corruption indicators used typically capture petty corruption.3 From a theoretical point of view the effects of petty corruption and grand corruption may well be very different.4 If petty corruption leads to tax evasion and firms going underground, then policymakers may find it optimal to lean towards higher inflation taxation. This is also the model-theoretic argument employed by Huang and Wei (2006) who contend that monetary policy should be allowed to be more expansionary, if there is a great deal of administrative inefficiencies (or petty corruption). Based on a similar framework, but modelling grand corruption instead of petty corruption, this paper allows for the opposite result; government expropriation could even lead to lower inflation. It is argued that grand corruption may actually produce better monetary policies, if fiscal policymakers are known to attribute some (though possibly limited) value to expropriating the public. Given such an objective, the government determines its optimal level of output expansion (equality of marginal gain from output stimulation and the marginal gain from expropriation) and uses all additional resources for direct expropriation. If the central bank were to go for more output expansion by producing a higher surprise inflation, the government would like to withdraw its output-enhancing fiscal stimuli and use them for expropriation instead. Thus, an equilibrium with lower inflation than in the non-expropriation setting is possible, if not likely.5 The time-inconsistency problem of monetary policy and the exploitation of the short-run Phillips curve have first been studied by Kydland and Prescott (1977). The argument is that the central bank can renege on its promise of stable monetary policy and use surprise inflation (leading to lower real wages) for reducing unemployment below the natural rate of unemployment.6 Anticipating agents render the government's machinations ineffective despite increased inflation. Barro and Gordon, 1983a and Barro and Gordon, 1983b argue that a commitment device is required for preventing such deviating policies, restoring private agents' trust in the policymaker, and thus ensuring an optimal outcome. Rogoff (1985) posits that an independent, inflation-averse (“conservative” in his terminology) central bank can establish a reputation for non-inflationary monetary policy and thus act as such a commitment device. Other commitment devices are central banker contracts – suggested by Persson and Tabellini (1993) and Walsh (1995) – and inflation targets – suggested by Svensson (1997) and discussed by Beetsma and Jensen (1999). Mishkin and Westelius (2008) show that optimally determined inflation band targets can achieve almost optimal monetary policy while being more realistic than writing optimal inflation contracts and appointing a “conservative” central banker. Pegging the exchange rate is yet another commitment device. However, Aizenman and Glick (2008) argue that – given a large enough shock – any fixed exchange rate regime will collapse and produce great losses in welfare. A limitation of all aforementioned papers is that they ignore fiscal policy. Alesina and Tabellini (1987) and many others show that the inflationary bias of monetary policy carries over, in principle, even if fiscal policy is included in the analysis. Nonetheless, the magnitude of the inflationary bias may be affected by behavioural and institutional assumptions.7Huang and Wei (2006), for instance, maintain that central bank conservatism should be limited, if the institutional quality of the government is rather poor.8 In addition, introducing fiscal policy without a strict (and unrealistic) budget constraint for each period leads to a deficit bias on top of the inflationary bias. For instance, Agell et al. (1996) and Demertzis et al. (2004) include a balanced budget objective in the government loss function and show that optimal fiscal policy is always expansionary because costs incurred by the deficit are offset by the benefit from stimulated output. 9 This paper introduces expropriation as a government objective. Fiscal and monetary policymakers share the typical Barro and Gordon, 1983a and Barro and Gordon, 1983b type inflation and output objectives, but the government has additional fiscal objectives. The central bank controls inflation and would like to create surprise inflation to stimulate output. The government determines fiscal policy while considering the trade-off of fiscal stimulation between the gain, in terms of output, and the loss caused by an unbalanced budget. Deficit spending is thus possible, but costly. There is yet another trade-off because, additionally, the government has an expropriation objective. It may choose to expropriate some of the government revenue and, thereby, forego the output stimulation effect. Overall, the results of the model confirm, qualitatively, what has been found in the literature without expropriation; there are deficit and inflation biases. The difference from earlier results of models including fiscal policy only becomes clear when the findings for the model with expropriation are compared to the results for the model without expropriation. This is possible because the political economy model presented here actually nests the standard fiscal–monetary interaction logic without expropriation. First, if output expansion is an important goal, ignoring the expropriation objective leads to an increase in the inflationary bias. Conversely (and ironically), it is only when fiscal policymakers are modelled to “cheat” the public (by expropriating government revenue) that monetary policymakers may be constrained in their “cheating” of the public (which they would do by reneging on a promise to the public). Expropriation, therefore, has an effect similar to those commitment devices mentioned previously (such as optimal central bank contracts, optimal inflation targets, or exchange rate pegs), i.e. it helps alleviate the time-inconsistency problem of monetary policy. Second, expropriation renders fiscal policy partially ineffective. There is still a deficit bias, but the output stimulation effect may be lost because resources can be expropriated instead. In addition to these findings, the approach is very general and could be applied to many previously discussed issues. By introducing expropriation as a government objective, this paper moves the discussion away from a benevolent government setting to the perspective of a government with selfish interests.10 This is done without losing the fundamental properties of earlier models, while acknowledging a role for expropriation and, thereby, widening the scope of the time-inconsistency literature. Issues discussed in earlier papers can now be reconsidered within a more general framework; the effects of differences in preferences (different weights and/or bliss points in the objective function) between policymakers and society or amongst policymakers themselves, for instance central bank conservatism; or alternative forms of (strategic) interaction between central bank, government and/or wage setters, for instance cooperation, non-cooperation, or Stackelberg leadership. In Section 5 of this paper, the model is employed for discussing previous “solutions” for the elimination of the time-inconsistency problem of monetary policy. First, a conservative central banker a la Rogoff (1985) is introduced and found to be unambiguously beneficial both in terms of monetary and fiscal policy. Second, the “symbiosis” result obtained by Dixit and Lambertini (2003a) is emulated. It is shown why, in their model, the introduction of fiscal policy alone suffices for eliminating the inflationary bias. The paper is structured as follows. Section 2 presents a general political economy approach to modelling grand corruption in a time-inconsistency setting. In Section 3, the results for the case without expropriation are obtained and interpreted. Section 4 discusses the impact of expropriation on both the inflation and deficit biases by comparing the model results with and without expropriation. In Section 5, examples of applications of the general approach are presented. Section 6 concludes.

This paper contributes to the literature on the time-inconsistency problem of monetary policy by analysing a Barro and Gordon (1983a)-type model including fiscal policy and grand corruption. The analysed case of identical loss functions for fiscal and monetary policymakers produces four main results. First, fiscal policy (without expropriation) does not eliminate the inflationary bias, but adds an additional budget deficit bias. This confirms previous results in the literature. Second, when a government expropriation objective is considered, the inflationary bias does no longer depend on the desired level of output, but on the desire for expropriation. Even if a large output expansion is aimed for, the inflationary bias does not increase. The existence of an expropriation objective may, therefore, alleviate the time-inconsistency problem of monetary policy. This is similar to the effect of any of the commitment devices discussed in the literature such as optimal central bank contracts, optimal inflation targets or exchange rate pegs. Third, given expropriation fiscal policy is rendered partially ineffective. The deficit bias of the non-expropriation scenario is preserved, but the output stimulation effect may be lost. Fourth, central bank conservatism is unambiguously beneficial in a situation with government expropriation and rational expectations. Central bank conservatism reduces the inflationary bias while increasing the fiscal stimulation, thereby confirming the standard monetary–fiscal trade-off. At the same time, the budget deficit does not worsen, but remains unchanged because government expropriation is reduced by the same amount. These results were obtained under some more or less restrictive assumptions. A major caveat in the entire literature is that intertemporal, strategic and reputational effects are not explicitly accounted for as already discussed in footnote 9. Other than that, the model offers a general approach which can be used for studying most of what has been ignored in this paper, for instance, monetary and output shocks, differences in preferences between policymakers, the sequential order of moves, the output stimulation effect of expropriation, wage setting behaviour, social welfare, etc. Two applications of the approach have been given: Rogoff (1985) style central bank conservatism; and costless fiscal policy emulating the results by Dixit and Lambertini (2003a). There are two ways of thinking about the findings in this paper. First, suppose the model actually describes reality. In particular, it could be argued that grand corruption is always present, at least to some degree, not just in developing countries, but even in advanced Western economies. Take for instance the state corruption under the Democratia Cristiana (DC) and under Craxi's socialist-DC coalition government in Italy. Or think about the corruption allegations against George W. Bush and other Western leaders. Giving the model a positive interpretation we could argue that the inflationary bias may not be as bad as commonly suspected. In addition, having established independent, conservative central banks in most Western countries greatly alleviates the problem. From a normative point of view, we are still stuck with a number of questions. Should we really try to eradicate grand corruption? What are the welfare implications? Secondly, the findings actually raise even more questions and suggest topics for future research. Grand corruption may have an effect on the inflationary bias, but what happens, if we include tax distortions in our model as well. Does making fiscal policy more costly necessarily increase the inflationary bias? Similarly, what happens if we allow for both petty corruption and grand corruption. Huang and Wei (2006) argue that we may want to limit central bank conservatism when there are inefficiencies in the tax collection process. In this paper, we argued in favour of more conservatism. Is there a trade-off, if different forms of corruption are included in the analysis?