Inter-regional redistribution through infrastructure investment: tactical or programmatic?

Abstract

In this paper we study the political economy of the inter-regional allocation of investment in infrastructure, in an effort to disentangle tactical and programmatic motives, where tactical politics refers to the discretionary allocation of investment to districts with more ‘political clout’ and programmatic politics refers to the allocation of investment according to certain ‘objective’ criteria (e.g., income level). We use a panel of data from the Spanish electoral districts for the period 1964–2004 to estimate an equation in which investment depends both on economic and political variables. The results show that tactical politics do matter since the districts receiving the most funds are those in which: (i) the incumbent’s vote margin is low, (ii) few votes are needed to win an additional seat in the legislature, (iii) the central and regional governments are controlled by the same party, and (iv) regional parties play a pivotal role in the legislature. However, the results also show that programmatic politics matter, since inter-regional redistribution is shown to increase: (i) with the arrival of democracy and European Union funds, (ii) under leftist governments, and (iii) the weaker the correlation is between ‘political clout’ and regional income.

Appendix

A model of tactical redistribution

In this case, the incumbents’ problem can be written as:

(A.1)

where $\overline{i}$ is the (exogenous) amount of resources available, i _j is the per capita investment allocated to region j, and n _j and e _j are the population and the number of representatives of that region j in the legislature. The First Order Condition (FOC) of this problem are

(A.2)

where v _j =n _j F _j (θ _j u _j ) are the number of votes obtained in j, F _j being an incumbent’s vote distribution function with density f _j ; u _j is the utility of a resident in j, depending on per capita resident income y _j , which is in turn influenced by the per resident infrastructure capital stock of the region, c _j ; θ _j is an indicator of the productivity of the efforts of the incumbent party in district j; and λ is the Lagrange multiplier. Now, we assume, firstly, that F _j is uniform with support |−η _j ,η _j |. Secondly, we assume that $u_{j}=k+((1+\varepsilon)y_{j}^{1+\varepsilon} )$, with ε≤0 to reflect diminishing marginal utility of income. Finally, we also assume that $y_{j} = a_{j}c_{j}^{\alpha _{j}}$, with a _j being a district-specific productivity shifter and α _j ≥0 the coefficient of infrastructure capital in the production function. Using all this information, knowing that ∂c _j /∂i _j =1, and defining μ _j =∂e _j /∂v _j , we can rewrite (B.2):

(A.3)

Now, after taking logs and rearranging, we can provide the following expression for the amount of infrastructure capital desired by the incumbent for district j:

(A.4)

where β _j =lnα _j −lnλ, σ=1+ε and lnψ _j =lnμ _j +lnθ _j +lnη _j .

A model of ‘constitutional’ programmatic redistribution

The incumbent allocates investment across districts to maximize the following social welfare function:

(A.5)

where n _j is the population of district j. This is a Constant Elasticity of Substitution (CES) social welfare function that allows varying degrees of relative regional inequality aversion. The ϕ parameter quantifies the aversion to regional output inequality, and ranges from −∞ to one. As ϕ becomes more negative, inequality aversion increases. When ϕ→−∞, the function approaches pure equity. In the intermediate Cobb-Douglas case ϕ is zero. And if ϕ is equal to one, then the government is exclusively worried about efficiency. In this case, w equals the country’s aggregate utility, which is reduced to aggregate output when using the above utility function with ε=0. Note that we assume that the ϕ parameter depends on a vector of variables (ς) which pick up what we have termed as ‘constitutional’ influences on aversion to regional inequality. The FOC of (B.5) are:

(A.6)

Using the same functional forms as above for the utility and production functions, and taking logs, the expression for the desired amount of infrastructure capital for district j is:

(A.7)

A model of ‘electoral’ programmatic redistribution

The constitution constrains investment allocation using a formulation such as $i_{j} = \pi y_{j}^{\phi}$, where π is a parameter which depends on the overall amount of investment funds available (which we consider fixed) and ϕ is the ‘taste-for-redistribution’ parameter, which now is the decision variable. Taking into account the production function above, after various calculations, the effect of ϕ on y _j is:

(A.8)

If we assume, for the sake of simplicity, equal productivity parameters across districts (α _j =α) and instantaneous depreciation (meaning that c _j =i _j ), expression (B.8) reduces to παy _j lny _j , indicating that an increase in the efficiency orientation of public investment benefits rich regions. The incumbent’s problem can now be stated as:

(A.9)

The FOC is now

(A.10)

Setting ε=0 in the utility function, and defining the political weights Ψ _j as in previous sections, the FOC in (B.10) can now be expressed as:

(A.11)

where λ ^∗=λπ/α is a constant that can be assumed to be equal to one, for ease of exposition. Expression (B.11) says that, in equilibrium, ϕ should be set in order to ensure that the resulting weights applied to the different income classes (i.e., $1/y_{j}^{1 - \phi}$) match the income pattern of political weights (i.e., the correlation between Ψ _j and y _j ). Note that as ϕ moves towards zero the weights approach inverse income (i.e., 1/y _j ) while when ϕ=1 all income classes receive the same unit weight. So, if rich regions become more powerful (i.e., the correlation between Ψ _j and y _j increases) ϕ should rise in order to make the weights less correlated with inverse income. Following this result, the equation relating the levels of capital stock and income can be written as:

(A.12)

where ρ=corr(ψ _j ,y _j ) is the correlation between ‘political clout’ and per capita income.