Thursday 14 April 2016

How elders' savings alter support ratio

Ask: NUS Economists
By Tilak Abeysinghe, Published The Straits Times, 13 Apr 2016

Q Is the old-age dependency ratio really increasing?

A The old-age dependency ratio (ODR) is one of the simplest measures commonly used to assess the burden of the elderly.

It is usually computed as the ratio of the population aged 65 and over to the population aged 15-64 or 20-64. This is usually expressed as a percentage to indicate the number of elderly dependants per 100 working-age people.

Does an increase in this demographic ratio amount to increasing economic dependence? If a retiree has saved enough for his remaining life, should he be counted as a dependant economically?

Although there are some alternative measures, they are mostly for total dependency, both of the youth and the old. They are based on the weighted sum of age-specific population groups.

One measure is the ratio of non-workers to workers weighted by the labour force participation rates. As a measure of old-age dependency, this assumes that anyone not in the labour force is a dependant.

Another measure is the ratio of consumers to producers based on consumption and income weights. This measure is not ideal for comparing countries with and without generous government transfers. Moreover, these methods require strong assumptions about the movement of the weights when projecting dependency into the future.

We need alternative measures to reflect the economic dependency of the elderly. I have developed one such measure that is easy to compute and easy to project. This is a savings-adjusted ODR that requires an adjustment to the conventional demographic ratio to account for savings available to the elderly.

Savings need to be re-scaled by dividing by a basic consumption expenditure level that is required to sustain a socially acceptable minimum standard of living. This expenditure level is the same for every elderly person, but it can vary over the years depending on how the cost of living changes.

If this savings/expenditure ratio is unity for every elderly person, then the savings-adjusted ODR works out to be the same as the conventional ODR.

Elderly folk who have this ratio greater than unity are less economically dependent. Therefore, very rich elderly folk with large bank balances will not be counted as economic dependants under this adjustment.

Without having to go to micro-level data, the adjustment can be done using the average data available in household expenditure surveys. The Singapore household expenditure survey provides the average monthly household income and expenditure by age group. By combining data from a number of previous surveys, we can construct the average income, expenditure and savings profiles over the working age for the same representative household.

Accumulated savings at age 65 is what is available for retirement. By projecting the expenditure profile over life expectancy, we can work out the expected expenditure at age 65. The average savings/ expenditure ratio at age 65 is the adjustment factor needed.

By dividing the conventional ODR by the average savings/ expenditure ratio, we get the savings-adjusted ODR.

(Technical details can be found in an article available at

The table shows the conventional and savings-adjusted dependency and support ratios. The support ratio, which is the reciprocal of the dependency ratio without the percentage conversion, indicates the number of working-age people available to support one elderly person.

It is interesting to note that the savings/expenditure ratio of the elderly lies well below unity for the cohorts born before 1955. These belong to the pioneer generations of Singapore who have been receiving a substantial amount of government support since 2014.

The estimates show that the conventional ODR trends upwards as expected. What is noteworthy is the mild downward trend of the savings-adjusted ODR.

Even between 2010 and last year, with mostly observed data, the conventional ODR has increased by 4.2 percentage points, whereas the adjusted ODR indicates a 0.8 percentage point drop.

Obviously, for a long-term drop in adjusted ODR, the productive capacity of the economy must continue to increase or at least must not decline. Nevertheless, it is clear that in a growing economy, the economic dependency measured by the conventional dependency and support ratios could be highly misleading.

The table shows that for the cohort born in 1970, who will turn 65 in 2035, there will be only three working-age people supporting one elderly person. But when savings are taken into account, the number goes up to a more reassuring 6.2.

The writer is the director of the Singapore Centre for Applied and Policy Economics at the Department of Economics, NUS.

No comments:

Post a Comment