Reopening the Convergence Debate: A New Look at Cross-Country Growth Empirics
read more
Citations
Financial Intermediation and Growth: Causality and Causes
Finance and the Sources of Growth
The New Growth Evidence
Determinants of Long-Term Growth: A Bayesian Averaging of Classical Estimates (Bace) Approach
A Reassessment of the Relationship between Inequality and Growth
References
Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.
A Contribution to the Theory of Economic Growth
A Contribution to the Empirics of Economic Growth
Large sample properties of generalized method of moments estimators
Economic Growth in a Cross Section of Countries
Related Papers (5)
A Contribution to the Empirics of Economic Growth
Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.
Frequently Asked Questions (15)
Q2. What future works have the authors mentioned in the paper "Reopening the convergence debate: a new look at cross-country growth empirics" ?
In particular, regional data sets on the United States, Europe and Japan have been used to study questions similar to those addressed with the Sum- mers and Heston data ( see, e. g., Barro and Sala-i-Martin ( 1992, 1995, ch. 11 ) ). The authors have information from period 0 to period T0τ for the dependent variable, and from period 0 to T τ for the explanatory variables ( T = T0 − 1 ). Arguably, to the extent that such regions share similar technologies and tastes, the incorrect treatment of the individual effect may be less seri- ous a source of bias. 0. Hence, Hansens ’ s ( 1982 ) results imply that a consistent estimate of θ can be obtained as θj = argmin ( 1 N NX i=1 Z 0iνi ) 0Aj ( 1 N NX i=1 Z 0iνi ) ( 14 )
Q3. What is the key assumptionunderlying the consistency of their estimates?
In the present context, lack of first-order serialcorrelation in the error terms of equations (8) and (9) is the key assumptionunderlying the consistency of their estimates.
Q4. Why do some authors argue that cross-country differences in the aggregate production function should be ruled?
Some authors arguethat cross-country differences in the aggregate production function should beruled out a priori, because knowledge of how capital and labor can be mostefficiently combined to generate output flows freely across borders.
Q5. How many first-difference equations can be used to construct the m2 test?
24By its construction, the m2 test can only be performed when there are at least 3 first-difference equations for each unit in the sample.
Q6. What are the main assumptions that have been suggested to eliminate the individual effects?
In the panel data literature, several transformations, other than first-differencing, have been suggested to eliminate the individual effects.
Q7. How much is the share of capital required to be consistent with the new estimate?
In order to be consistent with their new estimate of a 10% con-vergence rate the formula requires a share of capital on the order of 30%,the standard figure associated with non-augmented versions of the growthmodel.
Q8. What is the main implication of a high rate of conditional convergence?
The main implication of a high rate of conditional convergence is that economies spend most of their time in a neighbourhood of their steady state.
Q9. What is the problem with the -matrix method as applied to cross-country?
A second problem with the “Π-matrix” method as applied to cross-country growth is that its consistency relies on an assumption of homoscedasticity in the error terms.
Q10. What is the implication of the -matrix method?
The implication is the following: if strict exogeneity is violated, the Π-matrix estimates are inconsistent, while their GMM regressions are consistent.
Q11. Why is it difficult to appeal to transitional dynamics to explain why some countries grow much faster than?
Because their high rate of convergence implies that countries are close to thesteady state, it is difficult to appeal to transitional dynamics to explain whysome countries grow much faster than others.
Q12. Why do the authors use a mixed panel-cross-section technique?
Because they can only obtain an implied estimate of λ when using a time-invariant measure of investment in human capital, the authors employ a mixed panel-cross-section technique whose interpretation in terms of the Solow modelomitted variable bias and endogeneity in explaining this difference is approx-imately unchanged.
Q13. What is the relationship between GDP levels and growth rates in the open economy model?
In the two-sector, open economy model analyzed by Ventura (1995) a strong negative association between GDP levels and growth rates in a cross-section can be associated with permanent differences in international growth rates.
Q14. What is the reason why the authors believe their results are immune from bias generated by business cycle phenomena?
It is clear from equation (6) that one possible test of the Solow model is a testof the restriction that the coefficients on ln(s) and ln(n+ g + d) (say δ1 and19An additional reason why the authors believe their results may be immune from bias generated by business cycle phenomena is that the authors estimate a regression in deviations from period means.
Q15. How many percent of the growth rate is reflected in the coefficienton lagged output?
In particular, interpreting the coefficienton lagged output as reflecting the speed of conditional convergence indicatesthat this parameter is about ten percent per year for a wide range of specifi-cations.