Business World, 30 September 2013


Good amounts of spittle and bile have been expended analyzing the country’s unemployment statistics. That’s understandable. It was the government itself, after all, which made the question of “inclusive growth” turn on the question of job-creation. So the rate of employment — or its converse, the unemployment rate — seems to be the natural measure of success. Critics of the government, on the other hand, pooh-pooh high GNI (Gross National Income) growth rates as an instance of “jobless” growth — using the same largely stagnant unemployment rate as the proof of failure.

But hang on. Just how good is unemployment as a measure of welfare in a developing country? If the principal concern of the government and its critics is to relieve the poverty of the vast majority and to pull them into the middle class, then the unemployment rate is actually a bad statistic to focus on.

The reason — as most development economists (should) know — is that poverty is only very imperfectly related to unemployment in the absence of social insurance and in the presence of a large informal sector. In reality, most of the poor are not unemployed, and most of the unemployed are not poor.

You will see this if you look hard enough. The Institute for the Study of Inequality and Poverty (ISIP, based in the UP School of Economics) used the merged labor-force and family income surveys of 2009 to estimate the incidence of poverty among various sections of the labor force. As it turns out, poverty incidence is worse among the employed than among the unemployed (23% versus 17%)

unemployment and poverty

If poverty incidence is any metric of welfare, therefore, one must recognize that the unemployed as a group are in fact in a better position than those who are employed. This conjecture is supported by other data, including the fact that the average unemployed person has completed high school, while the average employed person has not.

A sharper picture emerges when we divide the employed into the fully employed and the underemployed. A surprising statistic that emerges is that poverty among the fully employed (19%) is almost the same as among the unemployed (indeed slightly worse). It is among the underemployed that poverty incidence is the worst.The 36% poverty incidence in that group is more than double that of the unemployed and is the most severe among groups of the labor force.

Here’s a final twist: in terms of number, most poor people are actually found among the employed. Indeed, almost half of the poor (46%) are already fully employed, while 22% of them are underemployed. The unemployed account for only 4% of the poor. The deficit in employment opportunities is understated, therefore, if one assumes it can be measured by unemployment or underemployment. For even unemployment and underemployment lumped together comprises only one-quarter of the number of poor.

Such apparent paradoxes are not hard to explain. To be counted among the “unemployed,” one needs not to have worked even a single hour in the reference week and to have been actively seeking work. In rich countries, once a person loses a job, unemployment insurance and welfare payments kick in, which allow her to devote herself full time to her job search. The level of such benefits will typically be well below income from regular employment, so that the roster of the unemployed will largely coincide with the list of the poor.

No such luck in a developing country, however. Without social and unemployment insurance, poor people have no other means of subsistence to fall back on and must find work no matter how low the pay and productivity. These are mostly informal sector jobs in agriculture and in small-scale retail services (e.g., hawking and vending). Entering those jobs is easy since the pay is low, as are skill-requirements. Some economists have actually termed this “disguised unemployment,” since it is statistically indiscernible. As the old development saw says, “The poor cannot afford to be unemployed; in the extreme, they must employ themselves.”

By contrast, it is people with savings, who come from better-off families, and who have better qualifications and prospects (e.g., college graduates) that have both the incentive and the means to support themselves through a spell of full-time job-hunting. This is why an unemployed person in the Philippines is more than 80% likely not to be poor. Unemployment is largely a middle-class phenomenon.

For government and its critics to obsess about unemployment, therefore, is to completely miss the mark and underestimate the task at hand.

It is the quality of jobs that matters for the poor, a fact to be addressed only by shifting the country’s employment structure away from low- to high-productivity employment. Indeed one might imagine a situation where unemployment barely changed, but where poverty was drastically reduced. That would occur if low-productivity workers managed to shift from marginal jobs to high-productivity sectors such as manufacturing and high-value services. Or if massive private investment increased the productivity of farm employment. Or if labor’s skills were enhanced through training and education.

Don’t count jobs, silly; think structural shift instead.