Consider Figure 9-10, which plots the gross domestic product (GDP) growth, in percent, against the ratio of investment/GDP, in percent, for several countries for 1974 to 1985.28 The various countries are divided into three groups- those that experienced positive real (i.e., inflation-adjusted) interest rates, those that experienced moderately negative real interest rates, and those that experienced strongly negative interest rates.
a. Develop a suitable model to explain the percent GDP growth rate in relation to percent investment/GDP rate.
b. From Figure 9-10, do you see any evidence of heteroscedasticity in the data? How would you test its presence formally?
c. If heteroscedasticity is suspected, how would you transform your regression to eliminate it?
d. Suppose you were to extend your model to take into account the qualitative differences in the three groups of countries by representing them with
FIGURE 9-10
Real interest rates, investment, productivity, and growth in 33 developing countries from 1974 to 1985

dummy variables. Write the equation for this model. If you had the data and could estimate this expanded model, would you expect heteroscedasticity in the extended model? Why or why not?

Transcribed Image Text:

e Positive real interest rates 9 Moderately negative real interest rates (0 to * Strongly negative real interest rates 5 percent) * Republic of Korea Algeria Singapore Pakistan o Indonesia。. Thailand Malaysia Sri Lanka IndiaEcuador o Tunisia Brazil o Turkey o Mexico Portugal Tanzania Malawi 3 Yugoslavia Chile Philippines Sierra Leone Cote d'Ivoire o Venezuela Uruguay Argentina Zambia 15 25 30 35 40 Average productivity Nigeria Zaire -1 of investmenta ▲ Ghana Investment/GDP (percent) " Line represents sample average.