Capitalism and Democracy

Transcription

1 Capitalism and Democracy Piergiuseppe Fortunato and Paolo Masella July 2004 Abstract This paper provides a theory to explain why in di erent historical experiences di erent coalition of social groups have been at the hearth of the democratization process. This issue is addressed using an occupational choice model were four di erent social groups interact. As time goes, the capital accumulation process leads to changes in both the composition of society and the labor market equilibrium. In turn, these changes a ect the interests of the di erent groups and leads to the formation of various possible coalitions in support of democratization. Fortunato: Dipartimento di Scienze Economiche, Università di Bologna. pfortunato@spbo.unibo.it. Masella: London School of Economics and Sticerd. p.masella@lse.ac.uk. 1

2 1 Introduction Social scientists have always been concerned with the study of the di erent processes which transform the type of political regime in a country expanding (or reducing) the rights to participate in election and public o ces, and the openness to public contestation ( inclusiveness of a system according to Dahl (1971)). Various students focused their analysis on the focal role played by the capitalists in leading these processes (see Huntinghton (1991) among the others). In fact, the existence of an increasingly powerful social class whose interests were linked mainly to the manufacturing sector has been seen as a necessary condition for the implosion of an oligarchy and the consequent change of political regime (see, among the others, Llavador and Oxoby (2004) and Ticchi and Vindigni (2003)). In particular, these theories suggest that the emerging class whose interests are crucially concentrated in the modern sector could strategically form a coalition with one of the other social group in order to impose a new political system better suited to defend their economic interest. Less attention has instead been devoted to the possibility that the other social classes could try to stop the rising of the emerging capitalistic class channeling the reforming of political institutions in a direction which advantages the traditional sectors of production. This paper proposes a general theory of democratization analyzing the di erent possibilities that could arise along the process of institutional evolution, and characterizing the conditions under which di erent coalitions emerge. Moreover, while the great majority of previous mentioned studies are inherently static, we propose a dynamic theory of democratization stressing the linkages between the structural change that characterizes an economy and the institutional transformations. Crucially, these institutional transformations arise endogenously triggered by those classes whose interest are no more defended under a certain institutional set up. In the spirit of the traditional analysis of Banarjee and Newmann (1994), Galor and Zeira (1993), and Ghatak and Jang (2000), we construct an occupational choice model in which individuals can choose both the sector (agriculture or manufacturing) in which they operate and the kind of activity that they perform (worker, entrepreneur or landlord). According to the political structure, the model 2

3 bases on a styled representation of oligarchies and democracies, whose main di erence is represented by the size of the constituency which take part to the electoral process. Clearly, in an oligarchic society there is just one social group taking the political decisions while under democracy all the social classes have equal standing in the decisional process. Furthermore, we constraint the political debate to a one dimensional decision over a public good whose e ects are non neutral with respect to the social groups. In particular, we assume that this public good bene ts the modern sector more than the traditional so that depending on the occupational choice outcome di erent groups show di erent preferences over the public good provision. In such a framework the development of the economy coincides with a shift from the traditional to the modern sector with more and more people being able to a ord the x costs needed to become a capitalist. This shift leads to political transitions which could be of di erent types. The following section introduces the economic environment. Section 3 presents the results of the occupational choice problem. In section 4 we describe the political set up. Finally section 5 studies the dynamics of the system, and section 6 concludes with a detailed discussion of the historical evidence. 2 The Economy 2.1 Individuals The economy is populated by an in nite series of subsequent generations of individuals. Each generation is denoted by subscript t andisformedbyacontinuumofindividualsi. Each individual has a single parent and a single child so that population size N t = N is constant overtime. Individuals produce a unique commodity, which can be consumed or bequeathed. In period t, each individual i is endowed with 1 unit of labor and an initial capital k i;t,moreoveronlyafraction± of the population is endowed with an amount of land equal to L, e the landowners. This group at time t is described by a distribution function G t (k), which gives the measure of landowners with capital less than k. While the distribution of capital over the rest of population is given by the function F t (k). Individuals derive income selling their endowments in terms of labor, land and capital. This income 3

4 y i;t is divided between consumption, c i;t, and bequests to the next generation, b i;t. For simplicity individuals utility is a Cobb-Douglas over consumption and bequests, u i (c i;t ;b i;t )=c s i;tb 1 s i;t where s 2 (0; 1). As a consequence each individual leaves a constant fraction of income as bequest (i.e., k i;t+1 = sy i;t 8i; t), and consumes the rest. We assume that the bequests received can be only invested in capital and that capital investments can only derive from bequests, so that k i;t+1 = b i;t. 2.2 Production There are three production technologies available in the economy: a subsistence technology (s), a capitalistic technology (c), and a rural technology (r). Crucial is the existence of a public good G (say education) that enters in both the subsistence technology and the capitalistic one. Production using the subsistence technology takes place using one unit of labor and without any capital. Moreover, it is increasing with the public good G Y s = f(g) with, f(0) > 0; f 0 ( ) > 0 and f 00 ( ) < 0 Production using the capitalistic technology could be started only after the investment of a x cost I>0. It takes place using one unit of labor and one unit of supervision public good, and it is increasing with the public good G Y c = g(g) 4

5 with, g(0) > 0; g 0 ( ) > 0 and g 00 ( ) < 0 Let us consider the following assumptions, g(g) > 2f(G) 8G and 2 f 0 ( ) >g 0 ( ) >f 0 ( ) (A1) The rst part of the assumption over the two production functions simply implies that the net payo of the capitalists is always bigger then the income got by the subsistence technology. The second one, as will become clearer later, is needed in order to guarantee that the median voter always owns to the group of entrepreneurs. Moreover, we assume also that, (1 ±)F t (I) < 1 2 (A2) As will become clearer later on, this assumption implies that the share of population without land and capital is always less that 50%. Finally, the rural technology uses one unit of labor and an amount of land L e to produce q. Let us assume that, g(0) ri q f(g) (A3) The assumption simply states that the capitalistic technology is always more valuable than the rural one (even without any provision of public good) and in turn the rural technology is always more valuable than the subsistence one (even with in nite provision of public good). 5

6 2.3 Individual incomes Given the existing production technologies, there are four possible di erent occupations. The entrepreneur provides the capital and the unit of supervision for the capitalistic technology; the employed worker provides the unit of labor for the capitalistic technology; the self-employed provides the unit of labor for the subsistence technology; and the landlord. Assuming that there are exogenous credit market imperfections such that borrowing is impossible and that the rate of interest is r, we can write the di erent incomes related to the di erent occupations as follows, 8 y i;t (w s )=Y s + rk i;t T t >< y i;t (c) =Y c w t + r (k i;t I) T t y i;t (w c )=w t + rk i;t T t y i;t (r) =q + rk i;t T t where T t is the lump sum taxation needed to pay the public good, and w t isthewageoftheworkers in the capitalistic technology. Finally, given the wages, we can evaluate what will be the bequests given by each agent depending on her occupational choice 8 >< k i;t+1 (s) =b i;t = s(y s + rk i;t T t ) k i;t+1 (w) =b i;t = s(w t + rk i;t T t ) k i;t+1 (c) =b i;t =(Y c w t + r (k i;t I) T t ) k i;t+1 (r) =b i;t = s(q + rk i;t T t ) With sq I sq 1 rs, which simply means that (with no public good provision) the landlords slowly (not all together) will become entrepreneurs; and with I sf(g) 1 rs ; which simply means that with no public goods no workers will be able to be entrepreneurs 1. 1 with s ¼ ri I s ¼ ri rs? qui scoppia il casino perche in questo caso data la prima parte della condizione qualche entrapreneur potrebbe non farcela a rimanere entrapreneur; e questa e una condizione che ci portiamo dietro anche quando discutiamo di full developed society nelle altre sections 6

7 3 Labor Market We have now to consider two di erent scenarios depending on the value of Y c and q. Inparticularwe have to distinguish between the case (Y c RI)=2 >qand (Y c RI)=2 <q. The rst case clearly refers to a situation in which there are huge productivity advantages for the modern sector, while the second one depicts a scenario in which the cross sector productivity are comparable. We will start analyzing the labor demand and labor supply curves and then we will study the equilibrium con gurations. 3.1 Labor Demand Case 1: (Y c RI)=2 q Under this con guration, we have two distinct situations. When the wage is lower or equal than (Y c RI)=2, this implies that the income of the entrepreneur is bigger than the one of the remaining two categories (landlords and workers). Therefore, independently on the land possession, all the individuals endowed with enough capital to run a rm will do it (i.e. the labor demand is (1 ±)[1 F t (I)] + ± [1 G t (I)]). Vice versa, with higher wages, everybody will strictly prefer being a worker than running a rm, so the labor demand will be equal to zero. 8 >< (1 ±)[1 F t (I)] + ± [1 G t (I)] when w (Y c RI)=2 0 otherwise Case 2: (Y c RI)=2 <q Under this other con guration, if the wage is lower or equal than Y c RI q,thisimpliesthat the income of the entrepreneur is bigger than the one of the remaining two categories (landlords and workers). Therefore, independently on the land possession, all the individuals endowed with enough capital to run a rm will do it (i.e. the labor demand is (1 ±)[1 F t (I)] + ± [1 G t (I)]). With higher wages (but still lower than (Y c RI)=2), the income of the entrepreneurs will be bigger than the one of workers, but lower than the one of the landlords. This means that only the capitalist without land will run a rm. With higher wages, everybody will strictly prefer being a worker than running a rm, so the labor demand will be equal to zero. 7

8 8 (1 ±)[1 F t (I)] + ± [1 G t (I)] when w Y c RI q >< (1 ±)[1 F t (I)] when Y c RI q<w (Y c RI)=2 0 otherwise w>(y c RI)=2 3.2 Labor Supply Case 1: (Y c RI)=2 q Under this con guration, clearly an agent without enough capital and without land will o er her labor endowment if the wage is bigger then the subsistence income; in the some way an agent without enough capital but with land will o er her labor endowment if the wage is bigger then the rural income; nally if the wage is bigger than (Y c RI)=2 (that in this case is bigger than the rural income), everybody will be willing to be worker. 8 0 when w Y S >< (1 ±)F t (I) when Y S <w q (1 ±)F t (I)+±G t (I) when q<w (Y c RI)=2 1 when w>(y c RI)=2 Case 2: (Y c RI)=2 <q Under this other con guration, clearly an agent without enough capital and without land will o er her labor endowment if the wage is bigger then the subsistence income; if the wage is bigger than (Y c RI)=2 but less than the rural income (this case is possible under this parametric con guration) all the agents without land will be willing to o er labor; nally, with higher wage everybody will be willing to be worker. 8 0 when w Y S >< (1 ±)F t (I) when Y S <w (Y c RI)=2 1 ± when (Y c RI)=2 <w q 1 when w>q 8

9 3.3 Labor Market Equilibria Case 1: (Y c RI)=2 q (i) (1 ±)[1 F t (I)] + ± [1 G t (I)] < (1 ±)F t (I) < (1 ±)F t (I)+±G t (I) Equilibrium wage, w = Y S Occupational choice, L 8>< w =(1 ±)[1 F t (I)] + ± [1 G t (I)] L S =(1 ±)F t (I) L w L C = L w L R = ±G t (I) Under these distributions, the amount of agents without capital and land is bigger than the amount of capitalists in the economy (with or without land); therefore the equilibrium wage will be the lowest possible and still we will get an ecess of labor supply (the labor in excess will be employed in the subsistence sector). (ii) (1 ±)F t (I) < (1 ±)[1 F t (I)] + ± [1 G t (I)] < (1 ±)F t (I)+±G t (I) Equilibrium wage, w = Y L Occupational choice, 8 >< L w =(1 ±)[1 F t (I)] + ± [1 G t (I)] L S =0 L C = L w e L R = ±G t (I) [L w (1 ±)F t (I)] In this second case, there is no more excess of supply, now the capitalist are more then the agents 9

10 without land and capital, so the age has to be a bit higher in order to convince some landowner without capital to become worker. (iii) (1 ±)F t (I) < (1 ±)F t (I)+±G t (I) < (1 ±)[1 F t (I)] + ± [1 G t (I)] Equilibrium wage, w =(Y c RI)=2 Occupational choice, 8 L w = 1 2 >< L S =0 L C = 1 2 L R =0 in this nal case, the number of agents without capital (both with and without land) is lower than the number of agents with capital (both with and without land). Therefore the equilibrium wage is the highest possible. Case 2: (Y c RI)=2 <q 2 (i)(1 ±)[1 F t (I)] < (1 ±)[1 F t (I)] + ± [1 G t (I)] < (1 ±)F t (I) Equilibrium wage, w = Y S Occupational choice, L w =(1 ±)[1 F t (I)] + ± [1 G t (I)] L S =(1 ±)F t (I) L w L C = L w L R = ±G t (I) 2 in order to rule out some less interesting case we assume that F 0 (I) >

11 considerations very similar to the ones of the case 1(i) apply. (ii)(1 ±)[1 F t (I)] < (1 ±)F t (I) < (1 ±)[1 F t (I)] + ± [1 G t (I)] Equilibrium wage, w =(Y c RI) Y L Occupational choice, 8>< L w = (1 ±) 2 L S =0 L C = L w L R = ± under these distributions, the number of capitalist with or without land is bigger than the number of agents without capital and land. There is an excess of demand of labor. This disequilibrium is solved by an high wage. The wage has to be high enough to make the landowners with capital indi erent between running a rm and using a rural technology. 4 The Political Arena 4.1 Preliminaries We assume that the political choices are of two di erent types. First, people have to choose the constitution, which does not represent a choice on speci c public policy but on the rules that will be followed to pick up those public policies. Second, there is the proper vote on the public policies. Four di erent social classes (characterized by con icting interests) emerge from the above described economic structure. Landowners endowed with enough capital to invest, poor landowners who can only choose between rural technology and working, landless endowed with capital bigger than I, and nally landless who are obliged to look for a job in the market. There are two possible di erent constitutional arrangements. The rst one is the dictatorship of just one class; the second is the full franchise (all the social classes take part to the political decisions). 11

12 Moreover, we assume that to keep the exclusive power (dictatorship) the elite has to pay a cost ". One can think to this cost as military expenses to avoid rebellions, or as a moral cost related to the unjust distribution of power. Once a constitution has been chosen, elections take place and the policies preferred by the median voter are implemented. 3 The only available policy in our framework is the provision of the productive public good that we described above, and that is characterized by linear production costs, C(G) =G Each social class will have a preferred G. Let us now assume that the choice of the constitution depends on the informal political power of the di erent groups, with this informal political power of each group at time t being equal to its respective share of income. In particular, assume that if the income share of one class is bigger than 1/2 then this social class is able to govern the country alone implementing its preferred constitution (de ne such a class, a dominant class); otherwise there is a call for coalition. Two or more social classes make a coalition in order to choose the constitution. Such a coalition, and the constitution implemented, is stable if there is no other coalition able to block it (and of course if there is not a dominant class). Moreover, in order to get rid of some not interesting multiplicity of equilibria, we are going to assume that a coalition in power that implements a particular constitution at time t will be able to change that constitution if and only if all the members of the coalition agree on that. The chronological sequence of the economic and political is as follows. The individual born with a certain inherited endowment in terms of capital. First, they decide which coalition to form in order to implement a certain constitution; second, given the institutional environment, they vote over the public policy; and nally they make the occupational choice. Production takes place and the consumption/bequest decision is taken. 3 noi supporemo che la classe sociale dominante e quella dei landlord senza capital, ci potrebbero essere casi in cui dopo la democratizzazione la classe capitalista diventa classe dominante. dovremmo citare in nota questi casi, sottolineando che il core del paper e comunque la transizione verso la democrazia. viri 12

13 4.2 Coalition Formation In order to study the process of coalition formation we need to evaluate the preferences of each di erent social classes with respect to the production of the public good. These informations are recorded in the following lemma. Lemma 1 The preferred policies of each di erent social class depend on the initial distribution of capital and land and so on the related labor market equilibrium case 1(i) and 2(i) G L nc =0<G L = c G NL c <G NL nc case 1(ii) G L nc = G NL nc =0<G L c = G NL c case 1(iii) G L nc = G NL nc = G L c = G NL c case 2(ii) G L = nc G L = c G NL c =0<G NL nc Proof: it follows directly from the Assumption 1. Notice that in the more developed a society is (i.e., when there is more capital) the less are the di erences in the preferences for public goods. In other words, since the society as a whole relies more heavily on the modern sector all the social groups end up having similar preferences. We have now enough informations to solve the coalition formation game backward. To do this we need rst to evaluate to which group own to median voter if democratic election were held under each di erent labor market equilibrium, Lemma 2 The chosen policy under democracies in the di erent cases are as follows: case 1(i) and 2(i) 13

14 Since NL nc < 1 2,theMVisLc (equivalently NL c ) and therefore G L = c G NLc is chosen case 1(ii) The MV is NL nc (equivalently L nc ) and therefore G L = nc G NLnc =0is chosen case 1(iii) Clearly the chosen policy is G L nc = G NL nc = G L c = G NL c case 2(ii) The MV is L nc (equivalently L c or NL c ) and therefore G NL = c G L = nc G NLc =0is chosen Proof: It follows from Lemma 1 and from the con guration of the labor market in the di erent cases. We are now able to characterize which coalitions will be formed when no social group is dominant. Two elements are focal: the identity of the median voter and the preferences of each social group. Proposition 3 If there is no dominant class the coalitions formed (i.e., the chosen constitutions) in the di erent scenarios are, case 1(i) and 2(i) Since NL nc < 1 2 then all the landless (with and without capital) and the landlords with capital form a coalition supporting democracy. case 1(ii) All the landlords without capital form a coalition with the landless without capital supporting democracy. case 1(iii) Grand coalition for democracy. case 2(ii) All the landlords and the capitalists without land will form a coalition in favor of democracy. Proof: It follows from lemma 2. The previous proposition contains the main results of our model. The transition towards a democratic system could be triggered by di erent coalitions depending on the con guration of the economic 14

15 structure. In particular, if there is an excess of labor supply (case 1(i) and 2(i)), the advantages derived by public good provision will be enjoyed by entrepreneurs (via increased productivity in the modern sector) and poor landless (via externalities on the subsistence sector and therefore via an increased bargaining power vis-a-vis the entrepreneurs). On the other hand if there is an excess of labor demand 4 we may have two di erent situations. If the technology is biased towards the modern sector (case 1(ii)), then the e ect of the excess of labor demand is an increase of the wage till to q andthisimpliesthe formation of a coalition between all the individuals without capital. Viceversa, when the bias towards the modern sector is not so strong (case 2(ii)) the excess of labor demand brings to very high wages, the workers extract all the rents from the modern sector, and, therefore, a natural coalition emerge between landlords and entrepreneurs. Finally, in fully developed society, when the rural sector disappears (case 1(iii)), also all possible con icts of interests tend to disappear giving the way to the formation of a Grand coalition. 5 The Dynamics 5.1 Economic Structural Change Consider an economy which at time 0 is characterized by a very limited amount of capital. In such an economy there are just few entrepreneurs while the majority of the population is employed either as a worker or as a landlord. There is an excess of supply of labor so that we are in the situation described in the case 1(i) (or 2(i)). Let us rst consider the initial conditions described in the case 1(i), that is we consider a situation characterized by a huge productivity bias towards the modern sector. As time goes, the landlords which accumulate enough capital become capitalists so that the share of population employed as entrepreneur is strictly increasing while the opposite is true for the landlords. Moreover, the landlords which have not enough capital begin to work in the modern sector. As a consequence, the equilibrium on the labor market changes and, in particular, the wages rise and equalize the returns on land (case 1(ii)). Finally, 4 We de ne as an excess of labor demand a situation such that the number people endowed with enough capital to run a rm is great than 50%. 15

16 as more and more people accumulate enough capital to become entrepreneurs, we move to the situation described as case 1(iii): the rural sector disappears while the labor market is characterized by an excess of labor demand. These information are recorded in the following lemma, Lemma 4 There exists a unique t 1 such that an economy characterized by a labor equilibrium 1(i) will move to a new equilibrium 1(ii). Moreover, there also exists a unique t 2 such that the equilibrium changes from 1(ii) to 1(iii). Proof: It follows from the assumptions over the accumulation process of the landlords and workers (assumption 3). Note that, because of the assumption made about the accumulation processes of di erent groups, the landlords are the only who will accumulate enough capital to become eventually capitalist. Therefore in our model the main engine of political changes is given by the transformations which take place inside the landlords elite. Some of them, the richest, will progressively change their interests and this continuous process lead to dramatic change in the equilibrium of the society. We can now analyze the situation characterized by a bias towards the rural sector. After a st phase of supply excess on the labor market (case 2(i)), and as more and more landlords get access to the entrepreneurial technology, we move directly to a new equilibrium in which there is an excess of labor demand (case 2(ii)). This is the case because, contrary to the previous situation, the advantages of the rural technology prevent landlords to choose being employed as workers. These information are recorded in the following lemma, Lemma 5 There exists a unique 1 such that an economy characterized by a labor equilibrium 2(i) will move to a new equilibrium 2(ii). Proof: Same reasoning of Lemma Political Structural Change Given the above described economic structural change let us now analyze the dynamics in the political arena. Consider a society which at time 0 is characterized by a dominant class (the landlords). This 16

17 was mostly the case on historical grounds: land ownership was the prerequisite for having political rights. As a consequence, the landholding elite will choose an oligarchic constitution and no public good will be produced. But this situation does not represent a steady state; in fact, as time goes, the richest landlords will accumulate enough capital to become capitalists and eventually the elite loose the dominant position. De ne this transition period as et. Proposition 6 We have two di erent scenarios depending on (Yc RI) 2 >< q: Case 1: If et t 1, then all the landless and the landlords with capital form a coalition supporting democracy. If t 1 < et t 2, then all the landlords without capital form a coalition with the landless without capital supporting democracy. Case 2: If et 1, then all the landless and the landlords with capital form a coalition supporting democracy. If 1 < et there is a transition to democracy supported by a coalition between landowners and capitalist without land Proof: It follows from Lemma 3 and 4, and from Proposition 1. 6 Conclusions This paper studies the forces which lead to a change of a political system characterized by a dominant social group towards more democratic institutions. In particular, the focus of our analysis is on the interaction among social groups characterized by con icting interests which can form coalitions in order to change the status quo. In order to analyze the structural change of the economy we employ an occupational choice model, and we show how the equilibrium on the labor market changes as more and more capital is accumulated. In particular, an economy which begins with labor abundance ends out in a new equilibrium were the raw labor is in fact scarce. This change, which in turn leads to a progressive change in the interests of capitalists and workers, may lead to the formation of new political coalitions. The other main force which shapes the dynamic evolution of our economy is the progressive shift of the richest landlords into the capitalistic group; indeed the capitalistic class get bigger as time goes as opposed to the traditional 17

18 land aristocracy. Clearly, such a dynamics heavily a ects the balance of power among groups. The interaction between these two forces is analyzed and a full characterization of the di erent transitions towards democracy is presented. 7 References Acemoglu, D., and J. Robinson (2000): Why Did the West Extend the Franchise? Democracy, Inequality, and Growth in Historical Perspective, Quarterly Journal of Economics, (4), Acemoglu, D., and J. Robinson(2001): A Theory of Political Transitions, American Economic Review, 91(4), Acemoglu, D., and J. Robinson(2003): A Theory of Political Development, mimeo, MIT. Acemoglu, D., and J. Robinson(2004): Economic Origins of Dictatorship and Democracy. unpublished book manuscript, available at Banarjee, A. and A. Newman (1994): Occupational Choice and the Process of Development, Journal of Political Economy. Bourguignon, F. and T. Verdier: Oligarchy, Democracy, Inequality and Growth, Journal of Development Economics. Cervellati, M., P. Fortunato, and U. Sunde (2004): From Hobbes to Rousseau: Inequality, Development and Democratization, mimeo, IZA, Bonn and Universit a di Bologna. Dahl, R. A. (1971): Polyarchy: Participation and Opposition. Yale University Press, New Haven. Galor, O. and J. Zeira (1993): Income Distribution and Macroeconomics, Review of Economic Studies. Ghatak, M. and N. Jiang (2000): A Simple Model of Inequality, Occupational Choice and Development, Journal of Development Economics. Oxoby, R., and H. Llavador (2004): Partisan Competition, Growth and the Franchise, Universitat Pompeu Fabra Discussion Paper, 730. Ticchi, D., and A. Vindigni (2003): Endogenous Constitutions, mimeo, University of Stockholm. 18