In the 1970s and 1980s, the public sectors of many Latin American economies accumulated large external debts, at the same time as their private sectors accumulated large external assets. This phenomenon partly explains the very high real interest rates faced by the Latin American public sectors, while declining domestic investment meant that severe adjustments were needed to generate foreign exchange for debt service. In Discussion Paper No. 253, Research Fellows Alberto Alesina and Guido Tabellini seek to explain this behaviour as the result of political polarization and instability.
The authors consider a two-period general equilibrium model of a small open economy in which political parties representing `workers' and `capitalists' randomly alternate in office. At the end of period one the (rational) voters vote for the party which is expected to deliver the highest utility for themselves in the second period. The party in office redistributes income in favour of its own constituency through taxes and transfers.
If the capitalist party holds office in the first period, it will borrow abroad as much as it can in order to transfer resources to the capitalists, constrained only by the need to repay the debt at the cost of curtailing the transfers to the capitalists in the second period. The capitalists, in turn, use these transfers to increase consumption, acquire foreign assets sheltered from fiscal expropriation, and increase domestic investment. Thus in equilibrium, the government borrows from abroad while the private sector acquires foreign assets although they face the same world interest rate, owing to uncertainty about the identity of future governments.
But why do the governments not attempt to prevent capital flight, through the imposition of capital controls? Alesina and Tabellini find that the desirability of capital controls depends on the political nature of the government. The capitalist government never finds it optimal to impose capital controls. Capital controls are more ambiguous for the workers' government: workers are worse off because they cannot freely increase their savings to offset government dissaving; on the other hand, capital controls force the capitalists to invest more domestically, so increasing domestic investment and widening the tax base in the second period. The authors find that the workers' government will always impose capital controls, but that their level will depend on the workers' degree of risk-aversion. The greater workers' risk-aversion, the more costly it is to restrict their ability to smooth consumption intertemporally in exchange for a higher domestic capital stock. Alesina and Tabellini demonstrate that in several Latin American countries, left-wing governments have indeed traditionally been more inclined to impose capital controls.
Political uncertainty also explains over-accumulation of government external debt. Alesina and Tabellini consider the case in which both governments already tax their opponents at the maximum feasible level in order to redistribute to their own constituency. Debt incurred in the first period can therefore only be serviced at the expense of transfers to the constituency of the second-period government. There is thus a probability that the debt will be serviced at the expense of the opposing constituency. Since these costs are not internalized by the first-period government, over-borrowing will occur. These results hold even if each government has the option of repudiating the debt inherited from its predecessor

External Debt, Capital Flight and Political Risk
Alberto Alesina and Guido Tabellini

Discussion Paper No. 253, August 1988 (IM)