Thursday, October 08, 2015

Demographic Impacts: Renting vs. Owning, Labor Force Participation, GDP

by Bill McBride on 10/08/2015 04:44:00 PM

Here is a review of three key demographic points:

1) Demographics have been favorable for apartments, and will become more favorable for homeownership.

2) Demographics are the key reason the Labor Participation Rate has declined.

3) Demographics are a key reason real GDP growth is closer to 2% than 4%.

Renting vs. Owning

It was over five years ago that we started discussing the turnaround for apartments. Then, in January 2011, I attended the NMHC Apartment Strategies Conference in Palm Springs, and the atmosphere was very positive.

The drivers were 1) very low new supply, and 2) strong demand (favorable demographics, and people moving from owning to renting).

Demographics are still favorable, but my sense is the move "from owning to renting" has slowed. And more supply has been coming online.

On demographics, a large cohort has been moving into the 20 to 34 year old age group (a key age group for renters). Also, in 2015, based on Census Bureau projections, the two largest 5 year cohorts are 20 to 24 years old, and 25 to 29 years old (the largest cohorts are no longer be the "boomers").  Note: Household formation would be a better measure than population, but reliable data for households is released with a long lag.

Population 20 to 34 years old Click on graph for larger image.

This graph shows the population in the 20 to 34 year age group has been increasing.  This is actual data from the Census Bureau for 1985 through 2010, and current projections from the Census Bureau from 2015 through 2035.

The circled area shows the recent and projected increase for this group.

From 2020 to 2030, the population for this key rental age group is expected to remain mostly unchanged.

This favorable demographic is a key reason I've been positive on the apartment sector for the last five years - and I expect new apartment construction to stay relatively strong for a few more years.

And looking forward on demographics ...


Population 20 to 34 years oldThis graph shows the longer term trend for several key age groups: 20 to 29, 25 to 34, and 30 to 39 (the groups overlap).

This graph is from 1990 to 2060 (all data from BLS: current to 2060 is projected).

We can see the surge in the 20 to 29 age group (red).  Once this group exceeded the peak in earlier periods, there was an increase in apartment construction.  This age group will peak in 2018 (until the 2030s), and the 25 to 34 age group (orange, dashed) will peak in 2023.  This suggests demand for apartments will soften starting around 2020 +/-.

For buying, the 30 to 39 age group (blue) is important (note: see Demographics and Behavior for some reasons for changing behavior).  The population in this age group is increasing, and will increase significantly over the next 10+ years.  

This demographics is positive for home buying, and this is a key reason I expect single family housing starts to continue to increase in coming years.

Labor Force Participation

A significant decline in the participation rate was expected based on demographics (there is an ongoing debate about how much is due to demographics, and how much of the decline is cyclical - however, as I've pointed out many time, a careful analysis suggests most of the decline is due to demographics).

But what about the decline in the prime working age labor force participation rate?

Each month I post the following graph of the participation rate and employment-population rate for prime working age (25 to 54 years old) workers. The following graph is through the September report.

Employment Population Ratio, 25 to 54In the earlier period the participation rate for this group was trending up as women joined the labor force. Starting in the early '90s, the participation rate moved more sideways, with a downward drift starting around '00 - and with ups and downs related to the business cycle.

The 25 to 54 participation rate decreased in September to 80.6%, and the 25 to 54 employment population ratio was unchanged at 77.2%.

A couple of key points:

1) Analyzing and forecasting the labor force participation requires looking at a number of factors. Everyone is aware that there is a large cohort has moved into the 50 to 70 age group, and that that has pushing down the overall participation rate. Another large cohort has been moving into the 16 to 24 year old age group - and many in this cohort are staying in school (a long term trend that has accelerated recently) - and that is another key factor in the decline in the overall participation rate.

2) But there are other long term trends. One of these trends is for a decline in the participation rate for prime working age men (25 to 54 years old).   For some reasons, see: Possible Reasons for the Decline in Prime-Working Age Men Labor Force Participation and on demographics from researchers at the Atlanta Fed: "Reasons for the Decline in Prime-Age Labor Force Participation"

First, here is a graph of the participation rate by 5 year age groups for the years 2000, 2005, 2010, and 2015.

Labor Force Participation Rate, Age Groups, 2000 - 2015
1) the participation rate for the "prime working age" (25 to 54) is fairly flat (the six highest participation rates).

2) However, the lowest participation rate is for the 50 to 54 age group.

3) And notice that the participation rate for EACH prime age group was declining BEFORE the recession.  (Dark blue is January 2000, and light blue is January 2005).

Everyone is aware that there large cohorts moving into retirement - and a large cohort in the 20 to 24 age group - but there has also been in a shift in the prime working age groups.

Labor Force Participation Rate, Percent Prime by Age GroupThis graph shows the percent of the prime working age population in each 5 year age group.  Note: Jumps are due to changes in population control.

Since the lowest prime participation is for the 50 to 54 age group (the second lowest is for the 25 to 29 age group), lets focus on those two groups. The 50 to 54 age group is the red line (now the largest percentage of the prime working age) and the 25 to 29 age group is the blue line (now the second largest percentage).  Just these shifts in prime demographics would lead to a somewhat lower prime working age participation rate.    Overall, the impact of this shift is small compared to long term trends.

Lets focus on just one age group and just for men to look at the long term trend.

Labor Force Participation Rate, Men, 40 to 44 Click on graph for larger image.

This fourth graph shows the 40 to 44 year old men participation rate since 1976 (note the scale doesn't start at zero to better show the change).

There is a clear downward trend, and a researcher looking at this trend in the year 2000 might have predicted the 40 to 44 year old men participation rate would about the level as today (see trend line).

Clearly there are other factors than "economic weakness" causing this downward trend.   I listed some reasons a few months ago, and new research from Pew Research suggests stay-at-home dads is one of the reasons: Growing Number of Dads Home with the Kids

Labor Force Participation Rate, Men, Prime Age GroupsThis graph shows the trends for each prime working age men 5-year age group.

Note: This is a rolling 12 month average to remove noise (data is NSA), and the scale doesn't start at zero to show the change.

Clearly there is a downward trend for all 5 year age groups. When arguing about the decline in the prime participation rate, we need to take these long term trends into account.

Labor Force Participation Rate, Men, Prime Age GroupsAnd this graph shows the same data but with the full scale (0% to 100%).  The trend is still apparent, but the decline has been gradual.

The bottom line is that the participation rate was declining for prime working age workers before the recession, there the key is understand and adjusting for the long term trend..

Here is a look at the participation rate of women in the prime working age groups over time.

Labor Force Participation Rate, Women, Prime Age GroupsThe graph shows the trends for each prime working age women 5-year age group.

Note: This is a rolling 12 month average to remove noise (data is NSA), and the scale doesn't start at zero to show the change.

For women, the participation rate increased significantly until the late 90s, and then started declining slowly.  This is a more complicated story than for men, and that is why I used prime working age men to show the gradual downward decline in participation that has been happening for decades (and is not just recent economic weakness).

Labor Force Participation Rate, Women, Prime Age GroupsThe second graph shows the same data for women but with the full scale (0% to 100%).  The upward participation until the late 80s is very clear, and the decline since then has been gradual.

The bottom line is that the participation rate was declining for prime working age workers before the recession, there are several reasons for this decline (not just recent "economic weakness") and the prime working age participation rate is probably close to expected without the recession.

GDP: 2% is the New 4%

And a third key point: We should have been expecting slower growth this decade due to demographics - even without the housing bubble-bust and financial crisis.

One simple way to look at the change in GDP is as the change in the labor force, times the change in productivity. If the labor force is growing quickly, GDP will be higher with the same gains in productivity. And the opposite is true.

So here is a graph of the year-over-year change in the labor force since 1950 (data from the BLS).

Year-over-year Change Labor ForceClick on graph for larger image

The data is noisy - because of changes in population controls and the business cycle - but the pattern is clear as indicated by the dashed red trend line. The labor force has been growing slowly recently but the rate of increase has been declining for some time.

We could also look at just the prime working age population - I've pointed out before the that prime working age population has just started growing again after declining for a few years (see See: Prime Working-Age Population Growing Again)

Now here is a look at real GDP for the same period.

Year-over-year Change GDPThe GDP data (year-over-year quarterly) is also noisy, and the dashed blue line shows the trend.

GDP was high in the early 50s - and early-to-mid 60s because of government spending (Korean and Vietnam wars).  As in example, in 1951, national defense added added 6.5 percentage points to GDP.  Of course we don't want another war ...

Now lets put the two graphs together.

Year-over-year Change Labor Force and GDPOther than the early period with a boost from government spending, the growth in GDP has been tracking the growth in the labor force pretty well.  The difference in growth between the dashed blue and red lines is due to gains in productivity.

The good news is that will change going forward (prime working age population will grow faster next decade).  But right now, due to demographics, 2% real GDP growth is the new 4%.