The monthly employment report is among the most popular and controversial of the government's economic reports. The latest one released Tuesday was no exception, with its weaker-than-expected new jobs but a decline in the unemployment rate from 7.3% to 7.2%.
One of the reasons the monthly employment report is so controversial is that it's an incredible hodgepodge of data from bipolar sources: the Current Population Survey (CPS) of households and the Establishment Survey of businesses and government agencies. For example, the Nonfarm Employment number comes from the establishment data, but the unemployment rate is calculated from household data.
Additional complications are the substantial revisions to both the CPS and Establishment data, and the complexities of seasonal adjustment, which is available for many, but not all, of the data series.
Full-Time vs. Part-Time Employment in the 21st Century
Let's take a close look at some CPS numbers on Full and Part-Time Employment. Buried near the bottom of Table A-9 of the Household Data are the numbers for Full- and Part-Time Workers, with 35-or-more hours as the arbitrary divide between the two categories. The Labor Department has been collecting this since 1968, a time when only 13.5% of US employees were part-timers. That number peaked at 20% in January 2010. The latest month is only slightly lower at 19.0%.
Here is a visualization of the trend in the 21st century, with the percentage of full-time employed on the left axis and the part-time employed on the right. We see a conspicuous crossover during Great Recession.
The Impact of the Great Recession
Here is a closer look since 2007. The reversal began in 2008, but it accelerated in the Fall of that year following the September 15th bankruptcy of Lehmann Brothers. In this seasonally adjusted data the reversal peaked in early 2010. Over three-and-a-half years later the spread has narrowed a bit, but we're a long way from returning to the ratio before the Great Recession.
The two charts above are seasonally adjusted and include the entire workforce, which the CPS defines as age 16 and over. A problem inherent in using this broadest of cohorts is that it includes the population that adds substantial summertime volatility to the full-time/part-time ratio, namely, high school and college students. Also the 55-plus cohort includes a subset of employees that opt for part-time employment during the decade following the historical peak spending years (ages 45-54) and as a transition toward retirement.
The Core Workforce: Ages 25-54
The next chart reduces the summertime volatility problem by focusing on the 25-54 workforce. Note that the government's full-time/part-time data for this cohort is only available as non-seasonally adjusted. To help us recognize the summer seasonality that remains, I've used a lighter color for the summer-month markers, which are the most subject to temporary shifts from part-time to 35-plus hours of employment. I've also included 12-month moving averages for the two series to help us identify the slope of the trend over the past three years.
Like the 16-and-over version, this chart depicts a current situation that is considerably different from the era before the Financial Crisis, although the trend has been slowly improving since early 2010, as illustrated by the dotted-line moving averages. Since this is non-seasonally adjusted data, the summer months are obvious outliers, with full-time employment spiking during the June-August period. In July 2012 and again in August 2013 the data points nudged each other. This month, not surprisingly, the spread returned. But we can hope the dotted lines will continue to slope toward each other and eventually cross back to the old normal for this core population segment.
Meanwhile, we can speculate that the Affordable Care Act (aka Obamacare) is playing a role in employer decisions about full-time versus part-time employment. The $2,000 per employee penalty for employers who do not comply with regulations has influenced some employers to begin shifting their employment policies. In July the government pushed the start of the penalty from January 2014 to January 2015. But the anticipation of the penalty, even though delayed a year, will probably continue to influence the interim decisions of private employers.
Another Look at the Core Trend
I'll close with a final chart of the 25-54 population data, this time showing the cumulative change in two series since January 2007. I've again included 12-month moving averages to highlight the trend underlying the seasonal volatility of this non-seasonally adjusted data, which is more evident in the part-time series. The two callouts are for the latest moving averages, which at the point are virtually equidistant from the flatline.
As I mentioned at the outset, the monthly employment report is a hodgepodge of data. The full-time/part-time ratio is but a tiny piece of the whole, and the magic 35-hour threshold is arguably arbitrary. But this is, nevertheless, a metric that bears watching in the months ahead as we try to understand where the economy is headed. The big question is whether the ratio shift is the result of structural change initially triggered by the Great Recession but now driven a combination of several factors, not least of which are workplace demographics and increased productivity driven by technology.
My data source for the charts and numbers in the commentary is the BLS's Databases & Tools page -- specifically the Labor Force Statistics here.