Population aging and labour's share of income

There aren't many statistical regularities in macroeconomics, and very few of them are as important as the apparent long-term stability of the shares of income that go to capital and labour. This 'stylized fact' was first noted by Nicholas Kaldor back in 1957, and his list of stylized facts shows up in every macroeconomics sequence. Or at least, it should. Among other things, it's the reason why the Cobb-Douglas functional form is so widely used for production functions.

One thing that occurred to me as I was putting together my recent posts on population aging is that if this stylized fact continues to be as robust in the future as it has been in the past, then demographic change is going to bring up some important distributional questions.

Although Kaldor based his observations on pre-1957 US and UK data, labour's share of income has stayed roughly constant in Canada over the past 40 years. The income measure I'm using is Net National Income (NNI), for reasons that I read somewhere (can't recall where) but which make sense to me. Net incomes (as opposed to gross incomes) don't include resources that must be used to replace depreciated capital. National income includes revenues generated abroad, less income generated in Canada that is sent to foreigners. Another way of making the distinction is that domestic income is produced in Canada, while national income is what Canadians receive. Using other measures doesn't seem to change the basic story.

 

It should be pointed out that the claimed stability is in the long term. Since profits are more volatile than wages, capital shares are procyclical - and labour share are countercyclical. The decline in the labour share during the 2000s was the subject of some discussion, although it wasn't - and isn't - clear that the decline wasn't just an artifact of an unusually strong expansion. This IMF study tried to say something about the effect of globalisation and labour shares of income, but my inner time series econometrician is sceptical about attempts to pick out a change in a long-run trend in 25 years of data.

On the other hand, if the labour share of income isn't falling, then we should be preparing for some really hard policy problems in the near future. Something like  8%-12% of the population is about to age out of the work force and into retirement, and if retirees incomes are funded at least in part by their savings, then the share of total income generated by capital should increase.

If the capital share doesn't increase - that is, if there's a decline in average assets per retiree and/or if the rate of return on those assets falls - then those retirement incomes are going to have to come from the shrinking number of people who will still be of working age. The politics of such a large, intergenerational transfer of income could be difficult to sustain.

Comments

[Nick Rowe]

Oh dear. More bad news for our pension plans! Trying to think of some reason to hope you are wrong.

Ummm. The Cobb Douglas production function would explain factor shares of NDP, but not NNP (NNI?), if there is international borrowing and lending. But since the demographic change is roughly global, this won't matter much.

[Jon]

Agree. Definitely agree. My long term prediction for equity returns is low. Retirement budgeting is based on gradual liquidation of assets. There is going to be very strong downward pressure on asset values as a result unless government old age program do a singapore and mandate investment accounts rather than transfer programs.

[Jim Rootham]

What happens to that graph if you took out compensation to senior brass?

[Lord]

It is irrelevant whether investment accounts or transfers are involved, both are intergenerational. Not needing to grow beyond technological change means most earnings will be distributed as dividends and assets will eventually be consumed as they won't need replacing. Emerging markets will continue to grow for some time even as developed markets flatten or shrink. The UK has had a stable population for 30 years with gdp per capita growth better than the US. But the world will slow over time.

[Joseph]

"There is going to be very strong downward pressure on asset values as a result unless government old age program do a singapore and mandate investment accounts rather than transfer programs."

Why do accounts help things compared with transfer programs? It does force at least some purchase of these assets (supporting the price at least a little bit) but that can only help so much given the number of retirees selling assets. In any case, there would be some of this support from us generation Xers buying assets for our own RRSPs.

In some senses transfer programs seem to make the transfer more transparent which increases the odds of an open debate on the right compromise between preventing poverty in old age and not over-taxing current workers.

[RSJ]

But we do not have a uniform distribution of wealth -- the wealthiest will keep accumulating assets, living off of the interest, until they die and will not become net sellers of assets. Wealth inequality will moderate any downward pressure on asset prices.

In general, it's "bad practice" to view financial assets at the macro level as being a store of wealth for retirement. That may be how an individual looks at it, but in aggregate this is not the case. You can buy assets on credit -- the non-financial sector doesn't do that a whole lot (except in housing) but the financial sector does. Because of that, you can't argue that there isn't enough money to buy assets that others are trying to sell, so that the assets must fall to a lower price.

If you look at assets as the present value of future profits and if there is a large population of retirees that are spending money but not receiving wages, then whether or not this increases profits or decreases profits will be a function of how the transfers are funded. If the transfers are paid for with corporate profit taxes, then this will put downward pressure on asset prices, but if the transfers are paid for by taxes on labor income, then there is no reason to believe that the market value of assets will decline, and it may well increase.

[Rick B]

Stephen,

A clarification, I'm too lazy to check this out myself

Does NNI include public sector incomes and treat the government as a corporation with tax revenues taken as profits?

[JCD]

"If the capital share doesn't increase - that is, if there's a decline in average assets per retiree and/or if the rate of return on those assets falls - then those retirement incomes are going to have to come from the shrinking number of people who will still be of working age. The politics of such a large, intergenerational transfer of income could be difficult to sustain."

I'm not an economist, but it seems to me that an increase in capital's share of income would effectively, in many ways, act as the same kind of intergenerational transfer. However, as there is obviously not a direct correspondence between age and income type, the effects of each type of intergenerational transfer would not be identical. This seems a very important point, as increases in the income share of capital would accrue most not to those with the most need, but rather to those with the most capital. This can be contrasted with intergenerational transfers of income, which could be targeted based on need (or even just distributed equally). Thus, although such intergenerational transfers may be politically difficult, this may just be a messaging failure.

[Stephen Gordon]

Yes, to the extent that a shift from labour to capital would benefit those who are older, that would be an intergenerational transfer. But if that doesn't happen, it would have to be more of brute-force tax-the-young-and-give-to-the-old thing.