Fertility: Just Mine the Data

People have been talking about fertility in Singapore. Raising issues like how the "Baby bonus has had little impact so far" and how the economy discourages working women to have children.

The fertility problem of may well be causally structural in an economic sense, and a solution like the baby bonus may or may not be the solution to the structural issues, depending on whether the incentive addresses fundamental underlying issues.

The Economist recently published an article titled "The best of all possible worlds?" (Economist, Aug 6th 2009). The article contained a section relating the Human Development Index (HDI) to total fertility rate (TFR). The HDI, taking values from 0 to 1, is a measure used by the United Nations. It has three components: life expectancy; average income per person; and level of education. The TFR is the expected number of births per woman over a lifetime.

Graphically, the data appears well fitted by a J-shaped curve that bottoms out in a trough of a less than 1.5 TFR at HDIs between about 0.85 to 0.92. (Based on 2006 data, Singapore had a HDI of 0.918. This corresponds to a total fertility rate of under 1.5 on the fitted curve. In 2006, Singapore had an actual total fertility rate of 1.06.) Let me just plop the image here.


The excellent fit suggests (and may be said to validate the statement) that structural issues strongly dictate fertility levels. This makes sense as the structural conditions of an economy have a major impact, in aggregate, on the decisions that individuals make.

In so far as encouraging parenthood is important to Singapore as a nation, it is necessary to understand what really drives decisions to or not to have children. While one may explain the strong trend as a shift in attitudes due to changes in the economic environment, details are needed in order to craft economic policy. This suggests the need for more detailed study. This would mean, at the very least, de-aggregating the HDI to pinpoint how each factor (education, income, life expectancy) affect fertility and going beyond that.

Public policy relating to parenthood is largely within the purview of MCYS. Unfortunately, it is not staffed sufficiently with economists or statisticians who have the necessary training to mine the data. I would suppose the government's Economist Service would be better placed to lend a hand.

This would at least put more quantitative substance from the real world into the policy discussion. While anecdotal evidence has the most heartstring-tugging appeal, it is statistics that we need, and soon. Alternatively, we could do whatever it takes to increase our HDI and hope for the best.