Corruption, centralization, and the shadow economy.

• Author: Echazu, Luciana

1. Introduction

   Bureaucratic corruption and a high incidence of unregulated, and often illegal, productive activities in what is conventionally denoted as the shadow economy, are two features that characterize many developing economies. (1) The interrelations between them have been examined in numerous studies. (2) These studies typically analyze the equilibrium incidence of bribery in the formal (or the regulated or legal) sector when the informal or shadow sector provides an avenue of escape from predatory bureaucratic demands. But bureaucratic corruption exists in the informal sector as well--in the form of extortions by enforcement officials in charge of apprehending illegal firms. While this aspect is relatively understudied in the literature, it can, as this paper shows, complicate the sectoral interactions sufficiently to invalidate established results in two areas of the literature on corruption. In the area of the "industrial organization of corruption," our analysis generates very different conclusions concerning the benefit of bureaucratic centralization. (3) The same analysis also shows that the complementary relationship between the sizes of the formal sector and the shadow economy may no longer be valid when the informal sector is corrupt and characterized by less productive firms.

   Shleifer and Vishny (1993) examine the impact of the structural organization of bureaucracy on official corruption and find that a centralized bureaucracy results in lower bribes than a decentralized one. The intuition behind their result is that when permits are issued by different officials, bureaucrats do not take into consideration the negative impact of their independent decisions on the corrupt incomes of other officials. A centralized bureaucracy that maximizes total bribes internalizes this external effect of one official's action on another, thereby lowering corruption. Because higher bribes reduce the viability of otherwise profitable firms, a decentralized bureaucracy results in a lower level of both economic activity and social welfare than a centralized one.

   We find that this result loses its general validity when we widen our analysis to consider economies with formal and informal sectors. When corrupt officials operate in both sectors, the benefits of bureaucratic centralization critically depend on whether it occurs within or across these two sectors of the economy. In particular, bureaucratic centralization is beneficial, as per Shleifer and Vishny (1993), only when it is confined to the formal sector. When corruption is present in both sectors, our analysis reveals that cross-sectoral centralization of the bureaucratic apparatus can result in higher bribes and lower welfare.

   Our model portrays three corrupt bureaucratic agencies, two granting permits in the formal sector and a bureaucrat-regulator (henceforth, simply denoted as the regulator) who monitors the informal (or shadow) economy. We consider four different kinds of bureaucratic organization. First, with decentralized corruption each bureaucrat maximizes his individual payoff. Second, vertically centralized corruption exists when a bureaucrat in the formal sector is also in charge of monitoring activities in the informal sector. (4) Third, with horizontal centralization there is one bureaucrat in the formal sector and a separate and independent regulator in the informal sector. Finally, totally centralized corruption is present when there is only one agent in charge of all the sectors.

   In practice, perhaps the most mentioned case of decentralized corruption is that of postcommunist Russia, while horizontal centralization was more prevalent in the Soviet Union where there was one agency charging bribes (see Schleifer and Vishny 1993). In Argentina, there are situations where the same agency in charge of granting a permit--for example, to operate nightclubs--is also in charge of monitoring for illegal operations; this is a typical example of vertical integration. (5) Other examples of vertical centralization are environmental offices like the U.S. Environmental Protection Agency and its counterparts in other countries such as South Africa (Water Institute), Costa Rica (Secretaria Tecnica Nacional Ambiental), and Argentina (Autoridad Regulatoria Nuclear): These agencies issue environmental permits and also monitor compliance activities of potential polluters. (6)

   We show that horizontal centralization lowers corruption and improves social welfare. Bribes chosen by different officials, as long as they operate within the formal sector, are strategic substitutes for each other, and replacing them with a monopoly reduces total bribery. This is the same as the external effects in Shleifer and Vishny (1993), and produces identical results; we term this as the "S&V (Shleifer and Vishny) effect."

   Vertical centralization, on the other hand, increases bribery and reduces welfare. As our analysis demonstrates, an increase in the bribe charged by the former induces relocation of firms to the shadow economy, where returns to corrupt activity increase. This favorable impact of formal sector bribes on the official's payoff in the informal sector is the crucial factor--we call it the "vertical effect"--that distinguishes the outcome of vertical centralization from that of Shleifer and Vishny (and horizontal centralization). (7) Because our model differs from that of Shleifer and Vishny in that we consider the shadow economy, we do not see our findings as contradictory but rather as a more in-depth extension of their research.

   The analysis of total centralization yields additional interesting results. Not surprisingly, it turns out that the relative strengths of the vertical and the S&V effects determine whether total centralization of the bureaucracy leads to improvements in social welfare over decentralization. Our findings suggest that the central authority should establish a centralization policy confined to only one sector because cross-sectoral centralization always results in higher corruption.

   A number of papers on corruption have explored the link between bribes in the formal sector and the size of the shadow economy (e.g., Rose-Ackerman 1978) and provide evidence of how higher corruption in the formal sector increases the size of the informal sector (Johnson, Kaufmann, and Zoido-Lobaton 1998; Friedman et al. 2000). In an interesting analysis of corruption, competition, and democracy, Emerson (2006) also provides empirical evidence that shows how countries with a high level of corruption have low development and emphasizes the negative relationship between corruption and competition. Choi and Thum (2005), on the other hand, demonstrate that the presence of the shadow economy is beneficial because it reduces corrupt bureaucratic extractions and consequently, enhances economic activities in the formal sector. Our results contrast those of Choi and Thum (2005) when we consider bureaucratic centralization across the two sectors. The model that we analyze demonstrates that the complementarity between the formal sector and the shadow economy no longer holds. In fact, our analysis shows that if relocation to the informal sector by a firm involves loss of productivity, then the existence of corruption in the shadow economy may induce higher corruption in the formal sector, and a smaller size of this sector, when the bureaucrats in charge of granting permits in the formal sector also monitor the informal sector.

   The next section presents the model and shows the conditions under which a firm chooses to operate in the informal sector. Section 3 introduces the issue of the complementary relationship between corruption and the shadow economy, and analyzes the simple case with only one bureaucrat in the formal sector and a regulator in the informal sector to show the impact of vertical centralization. Section 4 includes two bureaucrats in the formal sector and a regulator in the informal sector under the three different instances of centralization. The last section of this paper, section 5, presents the concluding remarks.

2. The Model

   Consider a continuum of firms of mass 1 that decide between operating legally in the formal, or official, sector and functioning illegally in the informal sector. Let v [less than or equal to] 1 be the proportion of firms entering the underground economy. To operate legally, a firm needs to acquire mandatory permits from bureaucrats who are corrupt and demand bribes. Bribes in the formal sector cannot be avoided; that is, if the firm refuses to pay a bribe, the bureaucrat does not grant the necessary license or permit. Let [b.sub.i] denote the bribe charged, where i = 1 or 2. Assuming the actual cost of the license is zero, the ith official, therefore, earns a net income of [b.sub.i] per license.

   Firms operating in the formal sector receive a profit of V, net from operating costs but gross of bribes. In addition, as in Waller, Verdier, and Gardner (2002), these firms receive an idiosyncratic return [epsilon] on investment, which is distributed uniformly over the unit interval, that is, [epsilon] ~ U[0,1]. (8) The magnitude of [epsilon] is private information to a firm and is unavailable to other firms or officials. Alternatively, if the firm decides not to enter the legal sector, it operates in the underground economy. All illegal firms are homogeneous and earn a profit V that is net of operating costs but gross of bureaucratic extortions. In this sense, we follow Waller, Verdier, and Gardner (2002) in assuming that firms in the informal sector, because of lack of access to formal sources of credit or import channels, cannot invest in productivity-enhancing technology; therefore, informal firms are less productive than formal firms and miss out [epsilon]. This loss of productivity from...