APPENDIX II.1 Model Estimation and Data Sources
Barro, R. and J. W., 2000, “International Data on Educational Attainment, Updates and Implications,” NBER Working Paper 7911, Cambridge, MA.
Cohen, D. and M. Soto, 2001, “Growth and Human Capital: Good Data, Good Results,” Technical Papers 179, Organization for Economic Cooperation and Development.
IMF, 2003, “How Can Economic Growth in the Middle East and North Africa Region Be Accelerated?”, World Economic Outlook, September 2003.
Nehru, V. and A. Dhareshwar, 1993, “A New Database on Physical Capital Stock: Sources, Methodology and Results,” Revista de Análisis Económico 8(1): 37–59.
Prepared by Anna Ivanova. The author would like to thank staff from the Ministry of Planning and from the Central Agency for Public Mobilization and Statistics (CAPMAS) for their assistance with the data used in this chapter.
Data reported in the Penn World Tables and the WDI suggest that, measured in PPP terms, Egypt’s real GDP per capita caught up with the average level of other non–oil producers in the MENA region during the 1990s.
The rapid growth in employment and labor force during the 1980s mirrored the demographic boom of the 1960s and 1970s, which continued through the 1980s, with population growing on average at 2.3 percent per annum. More recently, population growth has stabilized at about 2 percent per annum.
A Vector Error Correction Model (VECM) was used to estimate a production function of the form yHLt = αt + αkHLt, where yHLt = ln(Yt/HtLt) is the logarithm of output per workforce adjusted for human capital quality; kHLt = ln(Kt/HtLt) is the logarithm of capital per workforce adjusted for human capital quality; and at (the logarithm of TFP) was assumed to follow: αt = α0 + αt + ϵt. The estimated coefficient of α (the capital share) was 0.25, and the estimated coefficient on the time trend (a) was 0.01. See Appendix II.1 for details.
The study covered 84 countries, representing 95 percent of the world’s GDP and 85 percent of the world’s population. Egypt was one of the countries in the sample.
The results reported are those obtained from applying the Hodrick–Prescott filter. The main results were not affected when a band–pass filter was used or when the trend in output per worker was calculated as the fitted values from the VECM.
Similarly, in order for overall output growth to rise to 4.8 percent per year (the average growth from 1981-2000) on a sustained basis, real gross fixed investment would need to increase by 12.4 percent annually.
The number of lags included in the VECM was chosen based on two information criteria: Schwartz Bayesian and Hannan and Quinn. One lag was chosen for the underlying VAR in levels, which implied that no lags had to be included in the estimation in first differences.