Does the youth dependency ratio drive economic growth?

Jane Galt and Malcolm Gladwell have a tiff

Gladwell, citing David Bloom and David Canning, suggested that changes in the "youth dependency ratio," account for a big chunk of Irish economic growth.  The youth dependency ratio refers to how many young-uns require support, relative to the broader population.  The Irish legalized contraception in 1979, birth rates continued to fall, and later the economy boomed.  But is the connection a causal one?

Here is a basic argument and model that the youth dependency ratio can matter.

I can see three possible mechanisms.  1) Fewer babies mean that more women work.  2) Fewer babies mean that each baby gets more parental investment; in the long run those people are smarter.  3) Fewer babies raises the savings rate.

Which of these might have operated in Ireland? 

On Mechanism #1, Irish women still work much less than the OECD average, yet Ireland is wealthier than almost anywhere else in Europe.  If a theory of growth first postulates a big or dominant effect, and then predicts rates but fails when it comes to predicting levels, I worry.

If we look at "rates of growth" only, this estimate suggests that more Irish female labor accounts for 1.5 percent Irish growth a year.  That hardly covers the growth gap between Ireland and the rest of Europe.  One estimate of elasticities suggests that an extra kid lowers an Irish woman’s chance of working full-time by 11.3 percent, but raises her chance of part-time work by 7.7 percent.  How far does that get us?  Bloom is a renowned labor economist but his article is far from state of the art macroeconomics.  Do note that increases in "total factor productivity" — often driven by foreign investment — seem to be more important than "growth in labor inputs" by a three to two ratio.

It ought to be easy to show evidence that the Irish boom has been strongest in the sectors where women work the most, such as services and not manufacturing.  I can’t find that evidence, can my readers?

Mechanism #2 is for the long run and it cannot explain the Irish boom of recent times or the timing of its possible connection to contraception.  Higher skills are a big part of the Irish story, but the trend started in about 1967

Mechanism #3: In Ireland, since the mid 1970s, gross private savings rates have been falling, more or less.  More generally, time series models for a single country, including demographic ones, don’t predict savings rates very well.

These studies I am citing have their defects, but they do show that the overall question is not so simple.

Notes:

1. These graphs show that, for developing countries, the change in the youth dependency ratio has "eyeball power" for 1975-1990, but not for 1960-1975. 

2. This study of Asia suggests that the youth dependency ratio matters, often through the savings rate (not the Irish scenario); the entire story is conditioned by "institutional factors."

3. Latin America has had falling birth rates but has failed to cash in.  As Bloom stresses, favorable birth rates help only if the country has good policies for putting the new female workers into productive positions.  In this regard Galt and Gladwell may not be so far apart.

The bottom line: How much of the Irish boom is caused by the change in the youth dependency ratio?  I don’t know.  If I had to offer a "I’m just a poor lil’ ol’ blogger but I’ve read lots of real business cycles macroeconomics simulation papers" seat of the pants sort of estimate, I would opt for a maximum of 15 to 20 percent.  That’s certainly worth writing about, but it is not the major story either.  I’d like to see a sectoral decomposition analysis, and I suspect that would point our attention toward FDI, education, and a favorable tax regime as bigger factors.

Addendum: Malcolm adds more.

Comments

There's also the possibility of a reverse causal relationship. Country gets wealthier and has a greater variety of goods and services available as well as higher probability of children surviving to adult; as a result, parents decide to spend more on themselves and on fewer children.

Of course, that causal relationship is bigger in a transition from pre-industrial to industrial and information age economies, so it may not be appropriate to the case of Ireland.

But the rise in income per head has two components - the rise in the share of the population in employment and growth in the productivity of those employed. Demographics has had a powerful effect on the first of these - but there's no evidence that productivity growth in Ireland is fater now than it was in the 70s. Ireland's productivity performance has always been good - it was just being masked by crippling demographics.

You seem to ignore the simple mechanism in Bloom -- Williamson (1998) where a demographic dividend leads to an increase in the ratio of workers to population (temporarily), and so leads to a rise in per-capita output even if per-worker output is unchanged. Since TFP is measured by comparing outputs to inputs, this leads to a measured rise in TFP simply as a transition effect.

I think that a far more interesting question for M. Cowen to inveigh on (pretty please?) is Gladwell's original article, which was about a different sort of dependency. I couldn't quite get my mind around the question of whether having the cost of pensions fall disproportionately on firms that have survived long enough to have large pension obligations (but which would be extremely profitable if they were to start operating tomorrow) could conceivably be less efficient than having the cost fall equally on all firms.

Martin is right on. I too was looking for the effect of the EU transfers. They have had such a strong effect that accurately measuring other effects that have had a much smaller impact becomes too difficult.

The Irish are inherently superior. Only centuries of fratricidal chaos held back Irish growth. The current boom took off because Irish women forced a truce. Note: in a few decades I will write the same thing about Iraqis.

In the comments on Jane's reply to Gladwell, world class commenter "Tino" (who did such good work attempting to educate Marginal Revolution last spring about immigration), does the math:

"In 1987 Ireland had a GDP per capita that was 49% of the US. If this year they suddenly jumped to their 2004 demographic, with the same productivity, their GDP per capita would become 55% of the US. In fact last year it was 90% of the US.

"Demography is not much of the Celtic Tigar’s story, the dependency ration explains roughly 15% of the catch-up Ireland experienced. Even if we use the even lower 1970 figure for share of Irish population aged 16-65 demography still only accounts for 20% of the narrowing of the gap to the US."

I go for explanation number 2, with the twist that Irish preferences shifting from large families to fewer, more skilled children had as its first effect smaller families, and as its second effect a shift in public policy. Irish fertility peaked around 1970, even if it was another decade before contraception was legal.

The higher levels of religious belief in Ireland could help drive the Irish work ethic.

"For instance, 71% of the U.S. population believe in hell and the country boasts the world's highest per capita income, according to the 2003 United Nations Human Development Report and 1990-1993 World Values Survey.

Ireland, not far behind the United States in terms of income, likewise has a healthy fear of a netherworld with 53% of the population acknowledging hell's existence."

From http://www.usatoday.com/money/economy/fed/2004-07-27-fed-hell_x.htm?csp=26&RM_Exclude=Juno

The really big differences in per capita income between countries are in the productivity of the average workers, not in the number of dependents each worker supports. Tunisia and Ukraine have a higher % of their population in the 15-64 age cohort than Ireland, but their workers are just as a fraction as productive as Ireland's.

By the way, Ireland isn't really as affluent as its GDP per capita figures suggest (35% higher than the UK). Multinationals play lots of accounting games to take their profits in low tax Ireland. Still, Ireland has improved rapidly in real terms.

Ireland was able to reduce its taxes due to the EU transfers. I don't believe Portugal did.

I'm single girl and i search man my mail: ludalove2006@gmail,com
I'm 27y.old. If you search woman write to me and i can send to you my new pics
and tell more about myself. I post this message from this forum because
i don't have credit card and can't use dating site.
If you want find a friend please write to me i am very lonely woman.
Ludmila

This might prove to be true in terms of some countries. But this turns to be short sighted when we consider a country like China. As late as 25 years ago, China was concerned it had too many children to support. Today, however, China faces the opposite problem: as a result of the success of its "one-child" policy, the country faces the prospect of having too few children to support a rapidly aging population. (http://www.prb.org/Template.cfm?Section=PRB&template=/ContentManagement/ContentDisplay.cfm&ContentID=13825) .This also has lead to the country’s aged population investing a lot of money in their health care. But what would happen at this rate, where bulk of the population would age and there would not be too many youngsters in the working population?
On the contrary, even though India is a major competitor to China in terms of population, trends and predictions show that in India the number of people in the working age is greater.
This factor of the working population could be one of the important factors to predict a country’s economic growth in coming years. So, it might not always prove to be true that – when you have lesser young ones (who would be dependants) the economy is likely to grow in a faster rate.

Comments for this post are closed