Principal investigator: Viviane Pires Ribeiro
Article title: Economic Growth and Income Distribution and Education Investment Policies in Brazilian States: Theory and Econometric Analysis
Article authors: Joilson Dias and Maria Helena Ambrósio Dias
Location of the intervention: Brazilian states
Sample size: 130 observations
Main theme: Economic Policy and Governance
Type of Intervention: Income distribution and investment in education
Primary variable of interest: Economic growth
Evaluation method: Experimental Evaluation (RTC)
Evaluation Context
Economic policies aimed at improving human capital and income distribution seek long-term productivity growth and are evaluated at the country level. Dias and Dias (2007) state that the factors that hinder analysis at the country level are the differences between their physical and organizational, social and cultural structures, and, mainly, their economic policies. This set of differences implies ambiguous or difficult-to-interpret explanatory results for the effects of inequality on economic growth. In the study carried out by the authors, by focusing on Brazilian states, the effects of differences in the results are minimized, allowing quantitative aspects of their specific economic policies on productivity growth to be analyzed.
In this way, Dias and Dias (2007) contribute to the literature by addressing differentiated aspects of investment in education, income distribution, and investment in technology, at both theoretical and empirical levels, for the Brazilian states.
Intervention Details
The econometric evaluation of the economic growth of Brazilian states, considering data on the variable of technology imports, resulted in limiting the panel period to between 1992 and 1996 due to data availability. Therefore, the panel used by the researchers consists of 26 states (n=26) for a period of 5 years (T=5), totaling 130 observations (N=nT). The data used are sourced from Ipeadata and the National Household Sample Survey (PNAD).
The dependent variable used in the model's estimation process is the growth rate of worker productivity in the Brazilian states (as the control variable). This variable results from dividing the Gross Domestic Product of the states by the number of workers. The percentage of unemployed (control variable) is included to minimize variations in productivity resulting solely from the relative influence of economic cycles in each state and to help control for differences in the allocation of human capital among the states. The other control variable is a composite variable, representing the level of capital and technology in the state's economy. The other independent variables are: the ratio between the average education level of workers and employers; the level of taxation in the economy used as investment in education; income distribution in the economy; and the level of exogenous technology in the economy, represented by spending on technology imports in relation to each state's GDP.
Exclusive
The model developed by Dias and Dias (2007) follows that of Person and Tabellini (1994) regarding the incorporation of income distribution as an important element in determining a taxation policy in economies. However, the definition of income distribution has been modified and now includes only two classes: employees and employers.
In empirical tests, one of the important variables is the ratio of education levels between employees and employers. In addition to this, a measure of technology and the following income inequality indices are used: i) income share of the poorest 40%; ii) income share of the richest 10%; iii) income share of the middle class; and iv) the Gini coefficient, and two other control variables: v) the percentage of self-employed workers; and vi) the unemployment rate. The panel data econometrics employed consists of unit root tests to verify data stationarity, heterogeneity, autocorrelation, cross-sectional dependence, and estimates of fixed, random, and dynamic effects.
Result
The econometric results obtained by Dias and Dias (2007) demonstrate that the productivity growth rate of Brazilian states is positively affected by: i) the increase in the ratio of the educational level of employees vis-à-vis employers; ii) income redistribution policies that favor the poorest 40% and the middle class; and iii) imported technology. However, investments in education by Brazilian states have negative effects on their productivity growth rate, acting as a reallocation tax of inputs from the goods production sector to the education sector in the short term.
Lessons in Public Policy
The negative effects of income distribution on the rate of economic productivity growth were confirmed in econometric tests conducted by Dias and Dias (2007), especially when the authors considered the Gini index and the income concentration variable of the richest 10%. However, it was found that the income proportion of the poorest 40% and that of the broad middle class indicate the benefits of income redistribution policies in favor of these classes on the rate of productivity growth. Regarding the implementation method, according to the theoretical model, for a positive association between these classes and productivity growth to occur, they must be allowed greater access to education and technology.
The negative effects of short-term investments in education on the productivity growth rate found by the authors may support the thesis that investment in education causes an immediate reallocation of resources from the goods and services production sector to the human capital accumulation sector, potentially being one of the inhibiting factors for greater investments in education in Brazilian states. Furthermore, the degree of interaction with foreign technologies is an important factor for economic growth, but at the same time, a higher level of capital reduces the long-term growth rate, which may indicate convergence among the states when considering the other effects of the variables as constant.
Econometric estimates indicate that the educational levels of employees and employers are important factors in the productivity growth rate of states. However, income inequality has the effect of generating higher tax levels; even if these are fully invested in education, there will be negative implications for economic growth in the short term. Thus, awareness of this fact contributes to ensuring that investments in education are continuous and lasting in order to obtain the expected human capital, and in the future, these will generate long-term productivity growth, offsetting this short-term social cost.
References
Dias, J., & Dias, MHA (2007). Economic growth and income distribution and education investment policies in Brazilian states: theory and econometric analysis. Economic Studies (São Paulo), 37
