Exports and economic growth in Central and East European countries during transition

Abstract

This paper quantifies the contribution of exports to economic growth in Central and East European countries (CEECs) during transition. Two theoretical models are examined: the first is based on an aggregate production function which includes exports as an additional ‘input’; while the second is based on a two-sector (exports and non-exports) model where exports provide positive externalities in non-export production. Each model is estimated with both fixed and random effects using panel data. Results show that the random effects model is preferred and that exports have a significant impact on economic growth.

Notes

For a comprehensive review of the export-growth literature, see Giles and Williams (2000a, 2000b).

Export externalities exist which improve efficiency through better management, labour training and so on.

The two factor fixed effects model also includes time-specific dummies which model technical change. However, with annual data for 1994–1999, technical change is not expected to be a significant determinant of economic growth.

The two factor random effects model also includes a time-specific error component to model technical change.

Robust errors are not presented for the random effects model because of estimation problems (Greene, 2000, pp. 580–1).

The corresponding test for the model in Equation 5 is

 = 88.70.

Both Equations 2 and 5 were also estimated using two factor fixed and random effects models. Using fixed effects, the significance of the time dummies was tested using Wald tests and results show that

 = 0.96 for Equation 2, and

 = 0.96 for Equation 5; in each case the null that the coefficients associated with the time dummies are zero is not rejected. Similarly using random effects, estimates of the variances of the random, country and time errors are 16.160, 9.048 and 0.165 for Equation 2, and 17.340, 8.412 and 0.139 for Equation 5 both of which imply that time effects are relatively unimportant. Thus, technical change appears absent in CEECs between 1994–1999, as expected.

In comparison, Ram's (1985) estimates of the

-coefficient for 73 less developed countries range from 0.094–0.148 using data for 1960–1977.

In comparison, Feder's (1983) estimates of the social marginal products of investment in exports for 32 semi-industrialized less developed countries range from 0.226–0.289 using data for 1964–1973.