BUDGET SETTING AUTONOMY AND POLITICAL ACCOUNTABILITY. Susana Peralta. FEUNL and CORE-UCL. Abstract

Transcription

1 BUDGET SETTING AUTONOMY AND POITICA ACCOUNTABIITY Susana Peralta FEUN and CORE-UC Abstract The autonomy of local overnments in decidin their revenue level varies a lot worldwide, and is very low in Portual. We analyze the consequences of this autonomy from the viewpoint of political accountability. We study a two-period ame in which elections take place at the end of the first period, in a model where local officials may be public or self interested. We show that a reater autonomy improves selection (i.e., votin out bad incumbents), while it decreases discipline (i.e., ivin incentives to the bad incumbent). Electoral turnover is expected to be hiher with reater autonomy. We analyze the effect of tax settin autonomy on expected voter welfare. This paper was written for presentation at the conference Desenvolvimento Económico Portuuês no Espaço Europeu, to be oranized by the Portuuese Central Bank in February I am rateful to the Comité Científico for useful comments and to Banco de Portual for financial support. I thank Isabel Dias at the Comissão Nacional de Eleições for providin me with data on Portuuese local elections, Fátima Teixeira at the Direcção Geral das Autarquias ocais for help with municipal finance data and Sofia Fernandes for research assistance. Campus de Campolide P isboa. Phone: peralta@fe.unl.pt

2 66 Susana Peralta 1. Introduction Fiscal decentralization is a policy objective advocated by international oranizations such as the World Bank (World Bank, 2000) and the OECD (OECD, 2001, 2002). Moreover, it has become a dominant trend in several countries (Epple and Nechyba, 2004). owever, the deree to which it is implemented, both on its expenditure and tax collection aspects, varies a lot. Different institutional arranements on the sharin of competencies between central (or federal) and local (or state) overnments exist (see, for instance, Ter-Minassian, 1997, and OECD, 1999). A recent study by the OECD (OECD, 1999) has looked in detail at the tax settin autonomy of local overnments in 19 countries. It focuses both on taxes collected at the local level and tax-sharin arranements. These latter refer to taxes collected by the central overnment, whereof part of what is collected by (or arisin from activities located in) a local overnment's territory accrues automatically and unconditionally to it. Taxes and taxsharin revenues are classified into eiht cateories of decreasin autonomy. The larest autonomy occurs when local overnments are free to chose both tax rates and bases, while the lowest level refers to centrally set tax rate and base. Table 1 summarizes the results of the study. For each level of local overnment in each country, the percentae of revenues fallin into each cateory is computed. Table 1 displays, for each country, the unweihted averae across all local overnment levels (i.e., for Portual, autarquias, on the one hand and autonomous overnments from Madeira and Azores, on the other). Table 1 allows one to conclude that local overnments are far from havin full control of their fiscal revenue. It is noteworthy that transfers from the central overnment and supra-national institutions like the European Union, virtually totally out of local overnments's control, are outside the scope of the OECD study. Also, in some countries the central overnment decides on an allowable rane for the tax rates on revenue sources fallin into cateories (a) and (b). ence, if anythin, the fiures in Table 1 are too optimistic about local overnments' autonomy. Table 1 also makes clear that there is considerable variation amon countries. In this study, Portual shows up as one of the countries where local overnments enjoy less revenue settin autonomy.

3 Budet Settin Autonomy and Political Accountability 67 The question of revenue settin autonomy falls into the more eneral debate on whether overnment should be decentralized or not. The literature on decentralization is abundant, but it usually treats public ood provision and financin toether, i.e., either both are left to the central overnment or both are decentralized. One notable exception are papers studyin transfers from the central to local overnments. They build a case for local overnments' lack of revenue-collection autonomy both on efficiency (avoidin interjurisdictional externalities) and equity (avoidin fiscal imbalances) rounds (see, e.., Bucovetsky and Smart (2004) for a recent contribution where both objectives are shown not to be incompatible). This paper seeks to look at revenue autonomy from an alternative viewpoint, namely, political accountability. One may think that ivin more autonomy to local

4 68 Susana Peralta overnments increases the opportunities for corrupt local officials to extract rents for private purposes. As an illustration, one may read in the web-based edition of Inside Indonesia that corruption in the provinces stems in part from local politicians access to the budet. 1 The fact that local overnments may be more prone to corruption is the basis of the analysis of decentralization by Bardhan and Mookherjee (2000), who also reconize it to be a prominent issue in the political debate about decentralization, datin back to the foundation of the US (see Bardhan and Mookherjee (2000) and the reference to the Foundin Fathers therein). The potential daner of decentralization due to the poor quality of local politicians is also reconized in the political science literature (see, for instance, Bird, 2000). On the one hand, if one accepts the idea that local overnments are more prone to corruption than the central one, revenue-collection centralization can be seen as a way to fiht corruption, in that it hampers the capacity of local officials to extract rents. On the other hand, if the central overnment sets the local budet, we may reasonably suppose that voters observe it only imperfectly. In fact, fiscal laws are often complex and confusin (Bird, 2000), and it may be in the interest of politicians to hyde information from voters and render the budet unclear (this is suested by Besley (2005, pae 193), in the context of the use of debt). When the voters actually pay for the public ood with local taxes, their awareness of its actual cost increases. For instance, in a World Bank report about interovernmental fiscal relations in the Czech Republic (World Bank, 2003), one may read (...) local authorities have little autonomy over revenue. (...) predictable and transparent preparation of local overnment budets has been limited by lack of synchronization with the central overnment budet, uncertainty about basic budet parameters, and insufficient information on central overnment uarantees and continent liabilities. (...) Budets would be more predictable if basic parameters (...) were defined in oranic laws rather than the annual budet law. This lack of transparency is also a recurrent issue in the Portuuese political debate. This is clear in the followin quotation, taken from the Portuuese newspaper O Publico (November 11, 2005, pae 10), citin the intervention of the Minister of Finance in a parliamentary debate everyone who has assumed overnmental positions or has had any contact with local overnments knows that, besides 1

5 Budet Settin Autonomy and Political Accountability 69 the transfers predicted by the aw of ocal Finances, there are numerous additional mechanisms to transfer funds for local overnments, (...) transfers that are outside of the control of the Minister of Finance and are performed without any transparency. (...) The proposal in this year's budet aims at imposin a maximum value and increasin the transparency of such transfers. This lack of transparency at the local level is likely to have an impact on voters' capability to discipline local politicians. The idea that financin expenditures by raisin revenues locally leads to reater accountability is explicitly put forward by the World Bank (World Bank, 2003). 2 What is the impact on corruption and accountability when budets are under the full control of local overnments? Such is the object of this paper. We build on the model in Besley and Smart (2003). A public ood is provided at the local level. The provision cost may be hih or low. There are two types of politicians, public-interested and self-interested, unobservable to voters. If revenue is collected at the local level, voters observe both the public ood and the revenue level. If it is collected centrally, the central overnment knows the true provision cost, sets the budet accordinly, and voters cannot observe it, i.e., they only know the public ood level. The corrupt politician can misbehave in two ways, as in Besley and Smart (2003) and indriks and ockwood (2004). Firstly, he may provide no public ood and extract maximal rents, thereby revealin is type (separatin equilibrium). Secondly, when the true provision cost is low, he may pretend it is hih and extract the equivalent (smaller) rent (poolin equilibrium). In the former case, voters vote him out of office for sure, which is not true of the latter. As usual in this type of models, there is a trade-off between selection (the bad incumbent is revealed) and discipline (the bad type mimics the ood one). While voters want to discipline a bad politician (incentive), his ood behavior hampers the voters' capability to identify him as bad, thus they are less likely to vote him out of office (selection). We bein by comparin the two reimes in terms of rent extraction and obtain that none is better under all circumstances. Rents are hiher in the decentralized reime under the separatin equilibrium, but lower when the bad incumbent pools. Indeed, when the local official has full control of the budet and decides to reveal his type, he extracts 2 One may also find explicit references to this idea in the ocal and Municipal Governance and Finance proram (sub-proram of the Governance and Anti-Corruption earnin Proram), details on

6 70 Susana Peralta maximal rents. This is in line with the common wisdom that increased autonomy leads to more rent diversion. On the other hand, if the politician decides to pretend that the provision cost is hih, he takes advantae of the lack of voters' information in the centralized reime and extracts more rents in that case. It is not a priori clear whether it is better to have the budet set by the central or local overnment. Given the above, it is clear that the politician looses less from poolin in the centralized reime (conversely, has more to ain from separatin in the decentralized one). This explains our main result, namely, that separation occurs under a reater parameter rane in the decentralized reime than in the centralized one. Conversely, the centralized reime disciplines bad incumbents under a reater parameter rane. This means that, in terms of expected voter welfare, no reime is clearly preferred. In fact, for the parameter rane under which discipline and selection are the same in both reimes, centralization dominates decentralization. More interestinly, when the parameters are such that decentralization improves selection (at the cost of discipline) when compared to centralization, we obtain that this latter is preferable if the averae quality of politicians is low and the future is heavily discounted. Intuitively, the value of selection comes from the possibility of findin a ood politician to replace a bad one in the future. iterature review This paper is related to the literature on (fiscal) decentralization, whose roots date back to the seminal contributions by Tiebout (1956) and Oates (1972). Traditional approaches focus essentially on inter-reional spillovers and preference heteroeneity and suppose, in eneral, areate utility maximizer overnments. As we have already pointed out, the literature on fiscal competition has provided a rationale for decreasin the autonomy of local overnments (see, for instance, the survey by Wilson, 1999). Fiscal base mobility may be detrimental, as it increases the marinal cost of taxation, thereby creatin a downward pressure on the public sector. This constitutes a welfare loss if overnments are benevolent and is a reason to centralize tax collection. This paper is more in line with the recent contributions that take a political economy viewpoint: benevolent overnments are abandoned. The relative merits of centralized and decentralized systems have been analyzed, inter alia, by Bardhan and Mocherjee (2000), Besharov (2002), Besley and Coate (2003), and indriks and ockwood

7 Budet Settin Autonomy and Political Accountability 71 (2004). The first two focus on whether one of the overnment levels is more prone to capture by special interests, the first usin a model of political competition and the second a menu-auction one. In Besley and Coate (2003), centralized decision is undertaken by an assembly of locally elected representatives. indriks and ockwood (2004) use a political aency model. Voters have incomplete information about the quality of politicians and must choose whether or not to reelect an incumbent, usin past performance to infer his quality. A decentralized system (each reion run by a politician) is compared with a centralized one (one politician for both reions). A key feature of decentralized systems is the possibility for yardstick competition: with correlated economic contexts, voters can condition their reelection stratey on policy outcomes in the other reion. This is studied by Belleflamme and indriks (2002), Besley and Smart (2003) and indriks and ockwood (2004), amon others. The remainder of the paper is oranized as follows. The next section presents a brief description about the Portuuese system of local public finance. Section 3 introduces the model and equilibrium under each reime. Section 4 looks in detail at the relative merits of both reimes in terms of discipline and selection. Section 5 compares voters' expected welfare under the two reimes. In Section 6 we present some preliminary data on Portuuese local elections. Section 7 concludes. 2. A closer look at Portuuese local finance Portuuese local overnment is divided into two levels: municipalities (municípios) and a smaller unit called freuesia. 3 These two overnment levels are called autarquias. Additionally, the archipelaos of Madeira and Azores have their own autonomous overnments. As of 2004, there were 304 municípios and 4281 freuesias. The averae município has an area of 299 square km and 34 thousand inhabitants. Freuesia is a smaller unit, with an averae area of 21.6 square km and 2438 inhabitants. A closer look at the data also allows one to conclude for the existence of considerable heteroeneity in the sizes of municípios and freuesias (Direcção Geral das Autarquias ocais - DGA, 2004b). Both overnment levels have their own elected officials, with elections takin 3 The Portuuese Constitution mentions an additional level of local overnment, the Reions, encompassin several municipalities, which have not, to date, been implemented.

8 72 Susana Peralta place every 5 years. In the OECD publication mentioned in the introduction (OECD, 1997), Portual shows up as one of the countries conferrin a smaller autonomy to local overnments in the determination of their fiscal revenue. Takin a closer look at the Portuuese official publications (DGA, 2004a, 2004b) one may actually conclude that autonomy is even more restricted than what is suested by OECD (1997). We have attempted a reclassification, takin into account the detailed description of local overnments' tax settin autonomy in DGA (2004a, 2004b) and Baleiras (1997). The result is Table 2. This table refers only to autarquias, i.e., Madeira and Azores autonomous reions are excluded. It is noteworthy that taxes only represent 28% of total revenue (fiure for 2002), as shown in Table 3, which displays the composition of revenues of Portuuese autarquias. This raises the question of the capacity of local overnments to influence the amount of funds they receive as transfers. As is apparent from a close inspection of the transfers' rules, local overnments have little or no power to influence the amounts. ocal overnments are entitled to 30 % of the averae revenues of personal income tax, corporate income tax and

9 Budet Settin Autonomy and Political Accountability 73 value added tax collected in the country two years before. This total amount is then distributed to local overnments as follows % for the Municipalities General Fund (Fundo Geral Municipal), split accordin to a formula that accounts, in decreasin order of importance, for the number of inhabitants, hotel and campin occupation rates, youner than 15 inhabitants, eoraphic area, number of freuesias and amount of personal income tax collected in the municipality % for the Municipalities Cohesion Fund (Fundo de Coesão Municipal), which tarets the poorest municipalities, i.e., the ones with a per capita fiscal revenue below the national averae and the ones performin worse in terms of quality of life indicators (health, life expectancy, and education) % for the Municipalities Base Fund (Fundo de Base Municipal), equally split amonst the municipalities; 7.6 % is distributed to freuesias. ocal overnments are also entitled to conditional transfers (i.e.,tareted for specific projects) from the central overnment, varyin from 60% to 90% of the total cost in situations like unexpected public calamity, projects related to urban reconstruction, environment and natural resources, transportation infra-structures, subsidized housin, amon others. Finally, municipalities may apply to European Structural Funds and Reional Development Funds. In both types of conditional transfers, the scope of local overnments to influence the total amounts received relies almost exclusively on their capacity to submit projects which are in line with the fundin priorities defined by the central overnment or the European Union. The brief description above allows one to conclude for a weak autonomy of local overnments vis-à-vis the desin of their own revenue sources as tax proceedins are virtually insensitive to local policy-makers decisions concernin tax parameters and, as reards transfers, the discretionary autonomy of policy makers is very limited (Baleiras, 1997, paes 4 and 5). ocal overnments in Portual do seem to enjoy considerable freedom in choosin the provision of local public oods of which they are in chare (a nonexhaustive list includes parks, transportation and road system, sports and leisure, consumer protection, housin; see Baleiras (2002) and DGA (2004b) for a detailed description).

10 74 Susana Peralta 3. The model The base model is adapted from Besley and Smart (2003). A local overnment decides on the quantity of public ood G. The cost θ of the public ood is uncertain and it can be hih ( θ = ) or low θ = ), with >. The probability of a hih cost is q. Voters derive utility from the consumption of the public ood and dislike hih local budets. The local budet is denoted x [0, X ], i.e., there is a no-debt constraint and a maximum budet size. The utility function is W ( Gx, ) = G Cx ( ), where C is a strictly increasin and strictly convex function. The C function is meant to capture inefficiency costs of tax collection. We define G and G as follows G = ar max G C( θg) θ and let xθ = θgθ. Convexity of the C function ensures that x > x, i.e., the public sector is optimally larer when the provision cost is low. This, toether with <, implies that G > G, i.e., the quantity of the public ood provided decreases with the provision cost. We also suppose that X > x. There are two types of politicians, ood ( ) and bad ( b ). Good politicians always pursue the interests of the electorate, while bad ones care about the rents r they manae to extract. The proportion of ood politicians is π. This may be interpreted as a measure of the quality of the polity.

11 Budet Settin Autonomy and Political Accountability 75 The timin of the model is as follows. There are two periods. In the first period, there is an incumbent in place, who implements a iven policy. At the end of the first period, an election takes place. The incumbent is either approved by the electorate and stays in office one further period or he is voted out, in which case nature randomly selects a politician to be in office in the second period. The ame ends at the end of the second period, with no further elections. The future is discounted by β. The local budet x may be set by the local overnment or by the central one. We shall denote the former reime as D (mnemonic for decentralization) and the latter as C (standin for centralization). When the budet x is set locally, local officials decide x and G, and both are observed by the electorate. If the budet x is set by the central overnment, local officials only decide G. In this situation, voters have less information about the size of the budet. This may be because the law is unclear or too complicated to be understood by the electorate, with lots of exceptions, or even because there is no clear written rule as to the size of the budet. We capture this idea in its more straihtforward form, by supposin that voters do not observe x. The central overnment knows the true provision cost and sets the budet accordinly, that is x = x if θ = and is x = x if θ =. ence, voters know that the budet is x ( x ) with probability q (1 q ) Preliminary results Voters observe the policy implemented by the incumbent and use Baye's rule to compute the posterior probability Π that he is ood, iven the observed record. If Π is reater than the probability that the randomly selected official is ood, π, the incumbent is reelected. Otherwise, he is voted out of office. What about the politicians? As usual, we solve the ame by backwards induction. We bein by lookin at reime D. A ood politician implements ( Gθ, xθ ) in both periods. ence, any other policy vector perfectly sinals a bad politician. A bad politician extracts maximal rents in the second period (since he is no loner concerned by re-election), i.e. x = X and G = 0. We now look at the bad politician's behavior in the first period. e may implement one of three policy vectors: ( G, x ), ( G, x ) or (0, X ) 4 To simplify notation 4 Any other policy vector is a perfect sinal of his type and is dominated by (0, X ).

12 76 Susana Peralta we shall refer to the policy vectors only by the G component. Also, let Π G denote the posterior probability of the ood type iven an observed quantity of the public ood of G. If the provision cost is, the bad incumbent ets a neative rent (equal to ( G ) ) by implementin G, no rent by choosin G and the maximal rent when he implements 0. The two former strateies are dominated by the last one. 5 If the provision cost is low, a similar dominance arument allows one to eliminate G. The politician either implements 0, extractin a rent of X, or G, extractin a rent of ( G ). We let λ denote the probability that he takes the latter action. We may summarize the behavior of the bad politician as follows. e never provides the correct amount of the public ood. If the cost is hih, he extracts maximal rents and is voted out of office. When the cost is low, he may use one of two possible strateies. e either separates himself from the ood type by providin no public ood and bein voted out. Or he pools with the ood type under a hih cost and keeps his chance of re-election. In so doin, he foreoes some current rents in return for a probability of reelection (hence, future rents). From the above, we immediately see that G can only be implemented by the ood politician, hence Π = 1 > π and the voters reelect the incumbent. Also, G Π 0 = 0 < π, hence the incumbent is voted out of office. Finally, we have that π q Π = G πq + (1 π)(1 q) λ As reards reime C, a ood politician will, aain, implement G θ in both periods. The bad politician will aain extract maximal rents in the second period (i.e., implement G = 0 and et a rent equal to the budet set by the central overnment). As for the first period, when the provision cost is hih, the central overnment sets a budet of x. A dominance arument of the kind used above allows one to conclude that the politician implements G = 0 and extracts eliminate the dominated action (with a rent of ( G G )). r = x. If the provision cost is low, aain we may G and we are left with G = 0 (with a rent of x ) or G 5 Indeed, 0 yields a payoff of X, whereas the other two actions ive at most ( G ) + β X and β X, respectively.

13 Budet Settin Autonomy and Political Accountability 77 Voters can only observe the level of G. They do not reelect when G = 0 and reelect with probability one if the incumbent is ood is aain iven by G = G. Conditional on observin G, the probability that Π. G Comparin reimes C and D, one concludes that maximal rents are hiher in the latter, which is not surprisin, since the local official has full control of the budet. owever, ( G G ) = x G > ( ) G = x G (since x > x ) and the rent he extracts when poolin his hiher in reime C. This is due to the lack of transparency, since voters can only observe the level of G, and not the budet. We will use the followin notation D r = ( ) G C r = G ( G ) We have the followin preliminary result reardin the bad politician's behavior. Proposition 1 When the bad politician sinals his type to the electorate, he extracts hiher rents under the decentralized reime. When the bad politician mimics the ood one, he extracts hiher rents under the centralized reime Equilibrium avin outlined the strateies of the players, the equilibrium of the ame is straihtforward to obtain. et ˆ σ denote the probability that voters approve the incumbent after observin G and ˆλ the probability that the bad politician implements G when the cost is low, for reime D. For reime C, σ and λ have analoous meanins. Equilibrium in reime D is described in the followin lemma. emma 1 (Besley and Smart, 2003) When the budet is set at the local level, an equilibrium exists for all values of parameters and is enerically unique. 1. A poolin equilibrium, with ˆ λ = ˆ σ = 1, exists if and only if 1 and D q r (1 β ) X 2 2. A hybrid equilibrium, with ˆ D λ = q /(1 q) and ˆ σ = ( X r )/ β X, exists if and

14 78 Susana Peralta only if 1 and D q < r (1 β ) X 2 3. A separatin equilibrium, with ˆ0 λ and ˆ σ = 1, exists if and only if D r Proof See Besley and Smart (2003). (1 β ) X The equilibrium stratey of the bad incumbent is ruled by the trade-off between current (obtain r D and foreo X ) and future rents ( β X ). When behavin accordin to D voters' interest, the incumbent bears a cost of X r, with a ain of βσ X. If the cost is very hih, he prefers to reveal his type by extractin maximal rents and be voted out of office. If the cost is low enouh, he mimics the ood type to ain reelection. While future rents are iven by X in reime, irrespective of the state of nature, this is no loner true when the budet is set by the central overnment. In this case, the bad incumbent will extract a rent of x when the cost is hih and x when the cost is low. To compute equilibrium strateies, we have to make an hypothesis about the realization of the shock in the second period. To keep thins simple, we suppose that the provision cost is the same in both periods. ence, the rent extracted by the bad incumbent in the second period, if he keeps in office, is equal to x. emma 2 When budet is set at the central level, an equilibrium exists for all values of parameters and is enerically unique. 1. A poolin equilibrium, with λ = σ = 1, exists if and only if 1 and C (1 ) q r β x 2 2. A hybrid equilibrium, with λ = q /(1 q) and σ = ( x r )/ β x, exists if and only if C 1 and C (1 ) q < r β x 2 3. A separatin equilibrium, with λ = 0 and σ = 1, exists if and only if

15 Budet Settin Autonomy and Political Accountability 79 Proof See Appendix. C r (1 β ) x The intuition underlyin the equilibrium is the same as in the D reime. It amounts to a trade-off between current and expected future rents. Equilibrium of both reimes is summarized in Fiure 1, where i = C, D and max rent i stands for X when i = D and x when i = C. When rents from poolin are too low, separation occurs. If rents are hih, some poolin occurs at equilibrium. If the probability of a hih cost is hih enouh, voters are willin to believe that G was implemented by a ood politician and they reelect the incumbent with probability one. Otherwise, both voters and the politician play mixed strateies. 4. Discipline and selection While both reimes are equivalent in the equilibrium structure, they do no yield i the same payoffs. A closer inspection of Fiure 1 reveals that r, max rent i and σ differ D C between reimes. We know from Proposition 1 that max rent = X > max rent = x and r C D > r. It is therefore natural to expect that the separatin equilibrium arises more under reime C, which is exactly what one sees in Fiure 2, where the ( q, β ) space is divided

16 80 Susana Peralta into five reions, accordin to the prevailin equilibria. As the value of the future, β, decreases, the bad politician prefers to extract maximal rents in the first period and be voted out. This will happen for lower values of β under reime C. It is now clear that, when the budet is set centrally, less separation occurs at equilibrium. ence, lack of autonomy in budet settin improves discipline, at the cost of decreasin selection. 6 Proposition 2 The discipline of the bad incumbent is improved when the budet is set at the central level. We now turn to the selection issue, i.e., bad incumbents bein voted out of office. In a separatin equilibrium, selection is perfect: the bad incumbent sinals his type and is voted out. In a poolin equilibrium, there is no selection, since the incumbent is always reelected, irrespective of his type. ence, when β is intermediate and q > 1/2 we may safely conclude that selection is improved in the D reime. When q < 1/2, that is still the case. A bad incumbent is always voted out of office in case D and only sometimes in reime C. 6 Note that, under the hybrid equilibrium, ˆλ = λ, hence both reimes yield the same discipline.

17 Budet Settin Autonomy and Political Accountability 81 When the equilibrium of both reimes is hybrid, we have c D D C x r X r xr Xr σ ˆ σ = = < 0 βx β X βx X And aain the probability that the bad incumbent is voted out ( 1 σ ), is reater in the decentralized reime. 7 The fact that σ < ˆ σ is a result of the mixed stratey equilibrium. Given that in reime D the ain from separatin is hiher, voters must reelect with a hiher probability to keep the incumbent indifferent between current and future rents. Proposition 3 Elections are more effective in selectin out bad incumbents under reime D. We may interpret devolution of budet settin power to local overnments as makin more information available to voters. Our results would then imply that improvin voters' information is bad for discipline but ood for selection. This is reminiscent of the result in Besley and Smart (2003). In their model of local budet settin, they introduce a probability that voters become informed about the provision cost and show that increasin this probability improves selection but decreases discipline. Propositions 2 and 3 hihliht the basic trade-off between the two budet-settin systems. It is not clear, a priori, which system is better. In the next section we compare expected welfare under both reimes. One implication of this proposition is that we should observe more turnover when budets are set locally. We present some preliminary evidence reardin turnover in Portuuese local elections in Section 6. 7 For the sake of precision, one must reckon that in a hybrid equilibrium there is a probability to vote out the ood incumbent, when the cost is hih. This is equal to q(1 σ ). The probability to vote out the bad incumbent is q+ (1 q)( λ(1 σ) + 1 λ). ence, the net selection effect of elections is equal to 2 q 1 λσ + qσ = 1 σ, i.e., the reater is σ, the lower is selection. 1 q

18 82 Susana Peralta 5. Welfare comparison We bein by computin expected welfare (at the beinnin of the first period) for both reimes, EW C and EW D. Recall that the utility function is iven by G C( x). For the decentralized reime, this poses no problem, as voters observe both G and x. This is no loner true of reime C, as voters only observe G. owever, the central overnment's budet is funded by taxes, a part of which is born by the local voters. Our objective is to concentrate on the trade-off between selection and discipline. ence, we prefer not to make any explicit assumption about the reional distribution of the central overnment tax collection. In order to keep the two reimes fully comparable, we shall use the same utility function for both reimes. Voters may care about the budet size even if they do not pay for it entirely, or they do not foresee the link with their tax bill. Alternatively, one may interpret the utility function as reflectin the trade off between public ood provision and budet size which a benevolent planner would take into account for the purpose of comparin the two reimes. W b θ Before proceedin, it is useful to introduce some notation. et Wθ Gθ C( xθ ) b i i i = Cx ( ) and W = C( X). Also, denote EW = qw + (1 q) W. Now let θ X =, o W C and o W D denote the expected per-period utility in the absence of elections under each reime, i.e. W = πew + (1 π) EW and W = πew + (1 π) W o b o b C D X Expected utility in reime C is iven by o EW = (1 + β) W + (1 π)(1 q) λg + C C b b ( q W W q W W ) β(1 π) π σ( ) + (1 )(1 σλ)( ) While expected utility in reime D equals o (1 ) (1 )(1 ) ˆ b EW = + β W + π q λ( W W ) + D D X b ˆ b ( q ˆ W ˆ WX q W WX ) β(1 π) π σ( ) + (1 )(1 σλ)( ) (1) (2) The two last terms in (1) and (2) ive the impact of elections on voters' welfare. The first one is the discipline effect, the increased utility obtained when the bad incumbent refrains from extractin maximal rents, which happens with probability λ when the

19 Budet Settin Autonomy and Political Accountability 83 incumbent is bad (probability 1 π ) and the provision cost is low (probability 1 q ). The second term is the selection effect, representin the increased utility from identifyin and votin out bad incumbents. This may happen when the cost is hih or low. In the former case, a ood politician is voted out with probability 1 σ and a bad one with probability 1. ence, selection improves voters' welfare with probability σ. In the latter case, a bad politician is replaced by a ood one either if he pools and is voted out ( λ(1 σ) ) or if does not pool (1 λ). The total probability equals 1 λσ. One should note at this stae that incentive and selection only matter to the extent that some bad politicians exist. If π = 1, the existence of elections has no impact on welfare. This is a natural result in a model where there is consensus, i.e., all voters aree on their preferred policy. The welfare difference between the two reimes may be sub-divided into three parts. The first one is the baseline difference, i.e., what would obtain in the absence of elections, ( C X qc x q C x ) (1 π)(1 + β) Δ C = (1 π)(1 + β) ( ) ( ) (1 ) ( ) (3) The second one pertains to the discipline effect of elections ( ˆ ) (1 π) Δ D = (1 π)(1 q) λg λ( W + C( X)) (4) The third one is related to selection [ ( ˆ ) (1 π) βπδ S = (1 π) βπ q σ( W + C( x )) σ( W + C( X)) + ˆ ( ˆ σλ σλ ) (1 q) (1 )( W + C( x )) (1 )( W + C( X)) Summin up the three effects we obtain ] (5) ( C D S ) EW EW = (1 π) (1 + β) Δ +Δ + βπδ c d A few observations are in order. First of all, in the absence of bad politicians ( π = 1 ), both reimes would be equivalent in terms of welfare, since ood politicians have the same behavior in both. Secondly, in the absence of elections, reime C dominates

20 84 Susana Peralta reime D, since there is no discipline at all and only maximal rents matter. Finally, ΔC, ΔD and ΔS are independent of π. The averae quality of the polity, π, then determines the relative weiht of selection on the welfare difference. Indeed, the ain from votin out a bad incumbent is only ood to the extent that there is a hih probability that he be replaced by a ood one. Table 4 summarizes the relative performance of each reime in ivin proper incentives and selectin bad incumbents. Even if reime C is better in the absence of elections, reime D is, in eneral, more effective in ettin the most out of elections. Excludin the positive effect of discipline in S and SP (and selection in, which we cannot sin), reime C is always outperformed by reime D. Note that this is not incompatible with Proposition 2. It is still true that reime C provides more discipline, in the sense that it does so for a reater rane of parameter values: in reions S and SP, D provides no discipline at all while C provides some. owever, when both reimes provide discipline, i.e., in SS, and PP, the expected impact on voters' welfare is hiher under reime D. This is because the stakes of disciplinin the incumbent, i.e., the difference between maximal rents and rents extracted when poolin are hiher in reime D. Given the difference in maximal (i.e., second period) rents, the stakes of votin out a bad politician (the selection effect) are also hiher under reime D. This, toether with the result in Proposition 3, immediately implies that the impact of selection on welfare is hiher under reime D. We summarize our conclusions in the followin two propositions. Proposition 4 When equilibrium is separatin or poolin in both reimes, expected welfare is hiher in the centralized reime.

21 Budet Settin Autonomy and Political Accountability 85 Proof See Appendix. When equilibrium is separatin under both reimes, bad incumbents are always voted out, and second period utility is the same as first period one. As bad incumbents always extract maximal rents, reime C is better. As reards case PP, it is easy to see that maximal rents also play an important role. On the one hand, bad incumbents are never voted out and always extract maximal rents in the second period. On the other hand, the probability of a hih cost q, is hih, and bad incumbents also extract maximal rents in the first period when the provision cost is hih. It is interestin to notice that the advantae of the centralized reime stems in both cases from the welfare difference in the absence of elections. Indeed, inspection of Table 4 reveals that the impact of elections on welfare, i.e., Δ I +ΔS is neative in both cases SS and PP. We know, however, that a chane in the tax settin reime may induce a chane in the equilibrium outcome of the ame. In particular, it is possible that selection is improved in the decentralized reime (cases S, SP and ). We now investiate what is the likely impact of this equilibrium chane in the expected welfare of the voters. Proposition 5 When selection is improved in the decentralized reime, expected welfare is hiher under this latter if the value of the future is hih and the averae quality of the polity is low. Proof See Appendix. The intuition behind proposition 5 is easy to rasp. In the two cases in which Δ D >0 and Δ S < 0, i.e., S and SP, we just have to recall that the weiht of the selection effect is increasin with the averae quality of politicians and that selection matters when the weiht iven to the future is hih enouh. ence, when politicians are on averae very ood (and future is not too heavily discounted), the reime that performs better in terms of selection (D) is better than C. As reards the hybrid equilibrium, we have that, excludin the selection effect, the centralized reime outweihs the decentralized one. Imposin one further condition (amountin to an upper bound on the rent difference of the decentralized D reime, X r ), we ensure that the impact of selection on the expected utility is neative, and aain the intuition oes throuh.

22 86 Susana Peralta 6. Turnover in Portuuese local elections Our analysis has hihlihted several implications of the lack of tax settin autonomy. One of those is related to electoral turnover, which is expected to be lower than with a hiher deree of local autonomy. In this section, we present some preliminary data on turnover in Portuuese local elections. A richest empirical analysis would entail either an international comparison (countries conferrin different derees of autonomy to local overnments) or a comparative analysis of Portuuese municipalities, explorin the variation in revenue sources. Even if the eneral level of discretion is very low, there are some revenue sources for which municipalities enjoy a reater autonomy. A full-fleded empirical analysis bein outside the scope of this paper, we simply show some descriptive statistics to suest that turnover at the local level seems indeed to be restricted in Portual. We present our descriptive statistics in Table 5. ocal elections in Portual took place in 1976, 1979, 1982, 1985, 1989, 1993, 1997, 2001 and Table 5 ives the number of different presidents that have been in office in each municipality durin the last three decades. It refers to the identity of the president, as opposed to his party. 8 Althouh the Presidente da Câmara is not the only elected official in each municipality, he is the chief of the executive branch of the local overnment, the Câmara Municipal, and seems to enjoy a considerable latitude in decision makin. Indeed, the 8 This distinction is relevant, althouh not very important empirically, as it has happened that an incumbent runs for reelection representin a different political party.

23 Budet Settin Autonomy and Political Accountability 87 followin quote, adapted from DGA (2004b, paes 45-47), suests that the Presidente da Câmara enjoys considerable power. The Câmara Municipal is a permanent executive office, in chare of the oranization and functionin of the municipal services, urbanism and public works, as well as relations with other local overnment bodies. It executes the decisions of the Assembleia Municipal [leislative and consultative body], administers the employees and the patrimony, decides the local budet, concedes licences [for construction and economic activities], and ives support to Freuesias. (...) It is the Presidente da Câmara who decides the division of policy areas amonst the elected members of the Câmara Municipal (vereadores). Within certain limits, he may also decide the number of vereadores who are actually responsible for a policy area, as opposed to merely attendin the meetins, and whether they work part or full-time. e is also competent to supervise the administration of the employees at the service of the municipality. Table 5 allows one to conclude that most Portuuese municipalities have indeed had a very low number of different presidents. Indeed, with a history of 9 elections, the averae number of different presidents is 3,76. Around 73% of the municipalities have had 4 or less presidents. This means more than two mandates per president. No municipality at all has chaned president in every election. This piece of evidence is to be taken with care, iven that it does not take into account several important aspects. First of all, it may well be the case that turnout is low because the incumbent is, on averae, ood. If this is the case, however, our welfare analysis suests that increasin local autonomy will have a positive impact on voters' welfare. More importantly, our model is one of political consensus, that is, all the voters aree on the best policy to be implemented. In reality, the Presidentes da Câmara belon to different parties and propose different platforms to the electorate and this is, admittedly, an important determinant of turnover. 7. Concludin remarks This paper looks at the relationship between budet settin autonomy and political accountability at the local level. The usual focus of the literature is on decentralization of both expenditure and revenue collection. owever, both functions are not necessarily decentralized to the same extent. In particular, there is evidence of revenue collection bein

24 88 Susana Peralta only partially left to the autonomous initiative of local overnments. The usual rationale for such a lack of autonomy is the internationalization of inter-jurisdictional spillovers as, e.., externalities stemmin from tax competition. This paper studies the issue from the viewpoint of political accountability. We use a political aency model to compare two budet settin reimes: the centralized (budet set by central overnment) and decentralized one (budet set by local overnment). In both reimes, expenditure is set at the local level, i.e., local overnments decide the quantity of the local public ood. No reime dominates the other in terms of rent seekin. While maximal rents, that reveal a bad incumbent, are hiher with decentralization, a corrupt local official can extract hiher rents in the centralized reime without revealin his type. ence, the relative ains from separatin vis-à-vis poolin are hiher with decentralization. Therefore, the decentralized reime outperforms the centralized one in terms of selection, while it is outperformed in terms of discipline. With the backround of the relative merits of the two reimes in terms of discipline and selection, we proceed to a comparison of expected welfare. When both reimes yield exactly the same discipline and selection (i.e., equilibrium is either separatin or poolin in both), we show that centralization is preferred to decentralization. In both types of equilibrium, maximal rents play an important role, be it because the bad incumbent always plays them in the first period (separatin equilibrium) or because he is never voted out of office, hence they are very often played in the second period. Therefore, the reime with the lowest maximal rents is preferred. The most interestin cases arise when a chane in reime switches the type of equilibrium, namely, discipline is improved in the centralized reime. That is, the decentralized one has a separatin equilibrium, with no discipline at all, and at least some discipline arises under centralization, with a hybrid or poolin equilibrium. In these cases, the decentralized reime dominates the centralized one if the averae quality of politicians is hih and if the future is not heavily discounted. Indeed, it makes more sense to invest in selection if there are a lot of ood politicians in the world and if second period utility matters a lot; conversely, if most politicians are bad, or future is very discounted, then it pays more to ive them a stroner discipline. The analysis undertaken in this paper constitutes a first look at the issue of

25 Budet Settin Autonomy and Political Accountability 89 different derees of autonomy in the revenue and expenditure functions of local overnments. It ives interestin insihts about the main tradeoffs drivin the choice of the deree of autonomy in budet settin. The analysis could be extended in a number of ways. Firstly, we could analyze a ame with more than two periods. Indeed, in Portual, local politicians are not term limited and it is not realistic to suppose that they cannot run for reelection at the end of the second period. With a finite number of periods, the last period effect (i.e., the bad politician extractin maximal rents) always arises. The ain from poolin is however hiher as the politician reckons that he may serve for more than one additional period. ence, we could expect restraint to be more often observed. This is likely to shift the balance in favor of the decentralized reime, where rents extracted when poolin are lower than in the centralized reime. If we consider, instead, an infinite number of periods, then matters become more complicated, since we are likely to run into multiple equilibria issues, and it is not clear how to compare the reimes. Another interestin extension would be to consider endoenous entry of politicians, that is, suppose that each individual in the society is either public or self-interested and may decide to become a politician. The pool of politicians is likely to be better in the decentralized reime, as public-spirited individuals know that they have a hiher chance of bein elected, due to the increased turnover, and this increases their utility from becomin politicians. One may also envisae extensions that will have an impact in the information of voters, like the introduction of debt or the possibility of yardstick competition. These can chane the relative merits of the two reimes if there are reasons to believe that they have an asymmetric impact in information available to voters under the two reimes. The most interestin predictions implied by our analysis may be summarized as follows. Firstly, there should be more turnover at the local level when budet settin power is held by local officials. Secondly, we should observe more rent seekin at the local level in countries where local overnments enjoy less budet settin autonomy. Finally, one should observe less budet settin autonomy of local overnments in countries where the averae quality of the politicians is lower. The relative merits of increased local autonomy seem to have made their way into Portuuese politics. Indeed, the Portuuese Government has recently created a Commission for the revision of the law that reulates local public finances (ei das Finanças ocais),

26 90 Susana Peralta where the idea of ivin increased autonomy to, and increasin the responsibility of, local overnments plays an important role (Secretary of State for the ocal Administration, cited by the online newspaper Portual Diário, October 25, 2005). Appendix Proof of emma 2 We look at each type of equilibrium in turn. Poolin With λ = 1, we have that π G vote aainst the incumbent when they observe incumbent is λ = 0. With q 1/2 the bad incumbent, i.e. Π if and only if q 1/2. With q < 1/2, voters G, and the best reply from the bad, we have to make sure that G dominates 0 for r + β x x C ybrid Voters must be indifferent between reelectin or not, i.e., Π = π, which, solvin for λ, yields λ = q /(1 q). The bad incumbent must be indifferent between poolin or separatin, i.e. G Separatin Playin D C + σβ σ = β x r x x r x G must be dominated 0 with σ = 1, i.e. With 0, we have G voters. r + β x x C 1, hence reelectin is a best reply from Proof of Proposition 4 For reion SS, it is straihtforward to obtain EW EW = (1 π)(1 + β(1 π)) Δ C > 0 c Under PP, straihtforward manipulation yields ( d EW EW = (1 π ) q( C( X ) C( x )) (1 q)( C( x ) C( x )) c d + (1 q) β( C( X) C( x )) + (1 π) qβ( C( X) C( x )) )