How Responsible Have Early Retirements Been in Explaining Downward Breaks in US Employment and GDP?

 

Someone asked me how responsible reduced job search effort could be in explaining results in my new book Bad Breaks in Real GDP and Employment:  Exploring the Persistence of Aggregate Demand Shocks in the United States from Palgrave Macmillan in terms of a loss of real GDP, perhaps compared with previous trends.  That is, how important is it that people in the US may be substituting leisure for work (for example, early retirement) in explaining likely downward breaks in US real Gross Domestic Product (GDP) as found in Bad Breaks in Real GDP and Employment (short title)?  At least part of the thinking is that as production has increased over the decades, people may be able to live in at least some comfort without having jobs.

To begin to answer, let me point out what has been happening to living standards for millions of people in the US for decades.  I think that I heard someone on TV within the last few months say that we now have a country where we have $100 million dollar houses but a lot of our infrastructure is crumbling.  Consider how many people seem to be having a hard time getting by, let alone getting ahead.  We could have even more stuff to distribute with higher per capita real GDP.  To the extent that early retirement explains a loss of US real GDP, are many people retiring early due to weak aggregate demand growth in the US?  Does the explanation relying on early retirement for slower economic growth assume that if people decided to continue working, then they would be gainfully employed at wages or salaries that they would willingly accept, without taking any jobs away from anyone else?  Early retirements or other withdrawals from the labor market could create job opportunities for others with sufficient aggregate demand.

It seems possible that a wave of early retirements around a recession could lead to a downward break in the growth trajectories of nonfarm payroll employment and real GDP around the time of the recession.  However, are the early retirements due to weak derived demand for labor?  The demand for labor is contingent on the demand for the goods and services produced by the labor, with weak growth in wages and salaries due at least in part to weak growth in aggregate demand.  The early retirement explanation seems at least to me to be more of a supply focused explanation with fewer workers constraining supply capabilities.  But, would at least some come back into the labor market at least temporarily for the right offers assuming sufficient aggregate demand?  Also, many downward breaks in US real GDP occurred around times of falling inflation and perhaps short-term decreases in price levels.  This would suggest decreasing aggregate demand.  Additionally, note that until about 2-to-4 years ago, US inflation had been relatively contained for about 40 years.  This is despite at times very expansionary monetary and fiscal policies, reinforcing the possibility if not the likelihood of otherwise weak growth in aggregate demand.

I acknowledge that it may not be possible for real GDP to return to some pre-recession trends estimated in the book.  But even in that case or in those cases, given that the US economy now exceeds $20 trillion of real GDP, a sustained increase in real GDP of just a few percent amounts to a relatively large amount of goods and services that people could enjoy.

Maybe I should point out that to this point, I may have found possible evidence of a greater number of downward breaks in U.S. real GDP than the number of downward breaks in US nonfarm payroll employment.  Although there may have been downward breaks in US. nonfarm employment around the times of the dot com bubble recession (2001) and the Great Recession (2008-2009), if the finding holds true that there are more frequent downward breaks in real GDP than employment, then I am having difficulty seeing how early retirement can explain all of the downward breaks in real GDP. 

One thing that I may look into is whether the use of monthly data for nonfarm payroll employment may have led to the finding of fewer breaks than the number found for real GDP, which is measured no more frequently than on a quarterly basis.  Could it be that monthly measurements rather than quarterly measurements make it less likely for hypothesis tests to find evidence of breaks?

Another factor may also be relevant.  The US economy requires not only the ability to produce the goods and services that count as part of GDP in a quarter for those goods and services to be part of GDP in that quarter.  Also required is that someone or some group must pay for most of the items counted in US real GDP.  Imputed rent could be an exception because the housing services that homeowners receive from living in their own homes is not directly paid by the homeowners.  But again, that is an exception. 

The reason for this discussion is that if more people retire early, then if the early retirees could somehow increase their real spending on real GDP items by a large enough amount despite early retirement and if the economy would have the ability to produce those additional goods and services perhaps at least in part due to an improvement in technology, then real GDP could continue to grow sufficiently to avoid a downward break despite the early retirements.

 In conclusion, substituting leisure for work to some degree could easily result in less growth in real GDP in the US.  However, I question how responsible it is for ‘bad breaks.’