A working paper by Stone Center postdoctoral scholar Rafia Zafar shows that consumption expenditures can be used to accurately measure mobility in Indonesia, one of the largest lower-middle income countries in the world.
Research on intergenerational mobility within high-income countries has expanded greatly in recent decades. Yet there are far fewer studies on mobility within middle- and low-income countries than within high-income countries, despite the fact that such countries are home to 80 percent of the global population, and that rising inequality in these countries suggests that intergenerational mobility might be declining.
The primary problem is a lack of data. Measuring intergenerational mobility typically requires accurate income data linking one generation to the next, which are not available from most middle- and low-income countries. Even in middle- and low-income countries in which income data are available from tax records, the data likely present an incomplete picture, because more than 60 percent of the world’s population depends on income from the informal sector. Faced with a lack of income data, most researchers of intergenerational mobility in middle- and low-income countries have based their estimates on education, occupation, or assets. However, previous studies have shown that returns on education vary significantly across social groups and gender, and that occupation does not provide a good reflection of mobility. Also, assets in middle- and low-income countries tend to be self-reported, making their accuracy questionable.
But a new working paper by Rafia Zafar, a postdoctoral scholar at the Stone Center, shows that information on income can be combined with data on consumption expenditures to estimate intergenerational mobility. Consumption expenditures include all of a household’s spending on food, entertainment, education (including for children at home and away from home), utilities, and transportation. Consumption expenditures are already used to construct poverty lines in middle- and low-income countries, and previous research has shown that they are less vulnerable to measurement error than measurements of income. In addition, consumption expenditures provide a more accurate picture of the economic well-being of multigenerational households, which are common in middle- and low-income countries. (Living in a multigenerational household is, in fact, a potential source of mobility, because it enables multiple households to pool resources and share costs.)
Zafar’s paper focuses on Indonesia, the fourth largest country in the world by population, and the second largest lower-middle income country — according to World Bank classifications, which divide middle income countries into two groups — after India. Indonesia has experienced high economic growth for decades, but also increasing income inequality. In her study, Zafar constructed parent-child pairs using five rounds of the Indonesian Family Life Survey (IFLS), collected over 25 years. She then estimated relative mobility using both income and per capita consumption expenditures, drawing on data from multiple waves of the IFLS.
First, she estimated intergenerational mobility based on reported income data. This produced a baseline coefficient of 0.08. The baseline coefficient is a raw estimate of mobility between a parent and child; the closer its value is to 1, the lower the mobility. A coefficient of 0.08 therefore indicates very high mobility, suggesting that mobility in Indonesia is higher than in many high-income countries. Zafar then used per capita consumption expenditures as a measure of permanent income. This method produced a coefficient of 0.25, likely a far more realistic measure of intergenerational mobility in Indonesia.
Her study suggests that in middle- and low-income countries, intergenerational mobility estimates that rely on reported income may reflect a potentially large degree of measurement error, and that using consumption expenditures may lead to more accurate measurements. “There have been papers coming out in the past two and three years that have used consumption data for high-income countries,” Zafar says. “People have realized that consumption expenditures are a better measure of economic well-being or mobility over time. One of the main contributions of this paper is to show that instead of underestimating or overestimating mobility for middle- and low-income countries by using assets or occupation or education, we can just use the most direct data. For middle- and low-income countries, consumption expenditures are superior to any of those other measures of economic well-being in that regard.”
The applications of Zafar’s study extend far beyond Indonesia by paving the way for more research on intergenerational mobility in middle- and low-income countries. “People are becoming more aware of the growing inequality and low intergenerational mobility in middle- and low-income countries,” Zafar says. “Poverty is improving — it’s certainly still there — but it is improving, and we are seeing new issues coming to the fore because of rising inequality. We need to understand the impact policies centered around improving economic growth are having on inequality and intergenerational mobility in these countries.”
Read the Paper:
Intergenerational Mobility in Income and Consumption: Evidence from Indonesia