The reaction of private consumption to government expenditure in Estonia has gained little interest in previous literature. However, some facts might have been overlooked. First, the consolidated government budget has been out of balance. Second, Estonian consumers, it is often argued, have strong liquidity constraints – are less capable of taking loans against future income. This might have been true at the beginning of the transition period, but does not correspond to the situation where the public is worried about a lending boom. The objective of this paper is to estimate the extent of Ricardian equivalence in the Estonian economy. To that end the liquidity constraints and finite planning horizons are analysed and empirical estimations carried out.
The existence of Ricardian equivalence has been much debated in previous literature. However, there are no conclusive answers. The literature in favour stresses the importance of consumption smoothing and government bonds as not part of net wealth (Barro 1974). One of the most popular arguments for rejecting Ricardian equivalence is that it assumes perfect capital markets. Seater (1993) counts 12 arguments for why Ricardian equivalence might not hold. Many authors have taken a moderate position, where they see some of the effects of government consumption, but do not see the evidence giving full support to the hypothesis.
Robert Barro’s article, published in 1974, introduced a modern discussion of Ricardian equivalence. Testing has been conducted using micro and macro data. This paper follows the macro approach. In the approach taken by Haque and Montiel (1989), the hypothesis is that people are divided into two groups: liquidity constrained and unconstrained. The consumption of the constrained depends on their present and previous income. The unconstrained smooth their consumption and the amount consumed depends only on the last period of consumption, which acts as a proxy for an optimal consumption path. In the Aschauer (1985) approach, private and government consumption are estimated in a simultaneous system. All parameters are unconstrained in the first system. In the second, constrained system, the government consumption parameters are plugged into the private consumption function and the system is reestimated. The parameters of the two systems are then compared.
This paper consists of four sections. Section I includes two models that are used in empirical estimations. Estonian government and private income and expenditures data are presented in Section II. In section III tests for liquidity constraints, finite planning horizons and Ricardian equivalence using the full information maximum likelihood and instrumental variable techniques are carried out. The final section presents concluding remarks.
Contents
Introduction
1. Consumption Theory, Liquidity Constraints and Ricardian Equivalence
- 1.1. Consumption and Ricardian Equivalence
1.2. Liquidity Constraints and Finite Planning Horizons
2. Government and Private Consumption in Estonia
- 2.1. Government Deficit and Debt
2.2. Income, Consumption and Liquidity Constraints
3. Empirical Results
- 3.1. Tests for Liquidity Constraints
3.2. Rational Expectations Approach
Conclusions
References
Appendices
- Appendix 1. Haque and Montiel (1989) Model Constraints
Appendix 2. Description of Variables
Download
Liquidity Constraints and Ricardian Equivalence in Estonia
