We use cookies on our website to give you the best online experience. Please know that if you continue to browse on our site, you agree to this use. You can always block or disable cookies using your browser settings. To find out more, please review our privacy policy.


Policy Hub: Macroblog provides concise commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues for a broad audience.

Authors for Policy Hub: Macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

Comment Standards:
Comments are moderated and will not appear until the moderator has approved them.

Please submit appropriate comments. Inappropriate comments include content that is abusive, harassing, or threatening; obscene, vulgar, or profane; an attack of a personal nature; or overtly political.

In addition, no off-topic remarks or spam is permitted.

November 24, 2014

And the Winner Is...Full-Time Jobs!

Each month, the U.S. Census Bureau for the U.S. Bureau of Labor Statistics (BLS) surveys about 60,000 households and asks people 15 years and older whether they are employed and, if so, if they are working full-time or part-time. The BLS defines full-time employment as working at least 35 hours per week. This survey, referred to as both the Current Population Survey and the Household Survey, is what produces the monthly unemployment rate, labor force participation rate, and other statistics related to activities and characteristics of the U.S. population.

For many months after the official end of the Great Recession in June 2009, the Household Survey produced less-than-happy news about the labor market. The unemployment rate didn't start to decline until October 2009, and nonfarm payroll job growth didn't emerge confidently from negative territory until October 2010. Now that the unemployment rate has fallen to 5.8 percent—much faster than most would have expected even a year ago—the attention has turned to the quality, rather than quantity, of jobs. This scrutiny is driven by a stubbornly high rate of people employed part-time "for economic reasons" (PTER). These are folks who are working part-time but would like a full-time job. Several of my colleagues here at the Atlanta Fed have looked at this phenomenon from many angles (here, here, here, here, and here).

The elevated share of PTER has left some to conclude that, yes, the economy is creating a significant number of jobs (an average of more than 228,000 nonfarm payroll jobs each month in 2014), but these are low-quality, part-time jobs. Several headlines have popped up over the past year or so claiming that "...most new jobs have been part-time since Obamacare became law," "Most 2013 job growth is in part-time work," "75 Percent Of Jobs Created This Year [2013] Were Part-Time," "Part-time jobs account for 97% of 2013 job growth," and as recently as July of this year, "...Jobs Report Is Great for Part-time Workers, Not So Much for Full-Time."

However, a more careful look at the postrecession data illustrates that since October 2010, with the exception of four months (November 2010 and May–July 2011), the growth in the number of people employed full-time has dominated growth in the number of people employed part-time. Of the additional 8.2 million people employed since October 2010, 7.8 million (95 percent) are employed full-time (see the charts).

The pair of charts illustrates the contribution of the growth in part-time and full-time jobs to the year-over-year change in total employment between January 2000 and October 2014. By zooming in, we can see the same thing from October 2010 (when payroll job growth entered consistently positive territory) to October 2014. Job growth from one month to the next, even using seasonally adjusted data, is very volatile.

To get a better idea of the underlying stable trends in the data, it is useful to compare outcomes in the same month from one year to the next, which is the comparison that the charts make. The black line depicts the change in the number of people employed each month compared to the number employed in the same month the previous year. The green bars show the change in the number of full-time employed, and the purple bars show the change in the number of part-time employed.

During the Great Recession (until about October 2010), the growth in part-time employment clearly exceeded growth in full-time employment, which was deep in negative territory. The current high level of PTER employment is likely to reflect this extended period of time in which growth in part-time employment exceeded that of full-time employment. But in every month since August 2011, the increase in the number of full-time employed from the year before has far exceeded the increase in the number of part-time employed. This phenomenon includes all of the months of 2013, in spite of what some of the headlines above would have you believe.

So, in the post-Great Recession era, the growth in full-employment is, without a doubt, way out ahead.

Author's note: The data used in this post, which are the same data used to generate the headlines linked above, reflect either full-time or part-time employment (total hours of work at least or less than 35 per week, respectively). They do not necessarily reflect employment in a single job.

July 21, 2014

GDP Growth: Will We Find a Higher Gear?

We are still more than a week away from receiving the advance report for U.S. gross domestic product (GDP) from April through June. Based on what we know to date, second-quarter growth will be a large improvement over the dismal performance seen during the first three months of this year. As of today, our GDPNow model is reading an annualized second-quarter growth rate at 2.7 percent. Given that the economy declined by 2.9 percent in the first quarter, the prospects for the anticipated near-3 percent growth for 2014 as a whole look pretty dim.

The first-quarter performance was dominated, of course, by unusual circumstances that we don't expect to repeat: bad weather, a large inventory adjustment, a decline in real exports, and (especially) an unexpected decline in health services expenditures. Though those factors may mean a disappointing growth performance for the year as a whole, we will likely be willing to write the first quarter off as just one of those things if we can maintain the hoped-for 3 percent pace for the balance of the year.

Do the data support a case for optimism? We have been tracking the six-month trends in four key series that we believe to be especially important for assessing the underlying momentum in the economy: consumer spending (real personal consumption expenditures, or real PCE) excluding medical services, payroll employment, manufacturing production, and real nondefense capital goods shipments excluding aircraft.

The following charts give some sense of how things are stacking up. We will save the details for those who are interested, but the idea is to place the recent performance of each series, given its average growth rate and variability since 1990, in the context of GDP growth and its variability over that same period.





What do we learn from the foregoing charts? Three out of four of these series appear to be consistent with an underlying growth rate in the range of 3 percent. Payroll employment growth, in fact, is beginning to send signals of an even stronger pace.

Unfortunately, the series that looks the weakest relates to consumer spending. If we put any stock in some pretty basic economic theory, spending by households is likely the most forward-looking of the four measures charted above. That, to us, means a cautious attitude is the still the appropriate one. Or, to quote from a higher Atlanta Fed power:

... it will likely be hard to confirm a shift to a persistent above-trend pace of GDP growth even if the second-quarter numbers look relatively good.

This experience suggests to me that we can misread the vital signs of the economy in real time. Notwithstanding the mostly positive and encouraging character of recent data, we policymakers need to be circumspect when tempted to drop the gavel and declare the case closed. In the current situation, I feel it's advisable to accrue evidence and gain perspective. It will take some time to validate an outlook that assumes above-trend growth and associated solid gains in employment and price stability.

Photo of Dave AltigBy Dave Altig, executive vice president and research director, and


Photo of Pat HigginsPat Higgins, a senior economist, both in the Atlanta Fed's research department


July 18, 2014

Part-Time for Economic Reasons: A Cross-Industry Comparison

With employment trends having turned solidly positive in recent months, attention has focused on the quality of the jobs created. See, for example, the different perspectives of Mortimer Zuckerman in the Wall Street Journal and Derek Thompson in the Atlantic. Zuckerman highlights the persistently elevated level of part-time employment—a legacy of the cutbacks firms made during the recession—whereas Thompson points out that most employment growth on net since the end of the recession has come in the form of full-time jobs.

In measuring labor market slack, the part-time issue boils down to how much of the elevated level of part-time employment represents underutilized labor resources. The U-6 measure of unemployment, produced by the U.S. Bureau of Labor Statistics, counts as unemployed people who say they want to and are able to work a full-time schedule but are working part-time because of slack work or business conditions, or because they could find only part-time work. These individuals are usually referred to as working part-time for economic reasons (PTER). Other part-time workers are classified as working part-time for non-economic reasons (PTNER). Policymakers have been talking a lot about U-6 recently. See for example, here and here.

The "lollipop" chart below sheds some light on the diversity of the share of employment that is PTER and PTNER across industries. The "lolly" end of the lollipop denotes the average mix of employment that is PTER and PTNER in 2013 within each industry, and the size of the lolly represents the size of the industry. The bottom of the "stem" of each lollipop is the average PTER/PTNER mix in 2007. The red square lollipop is the percent of all employment that is PTER and PTNER for the United States as a whole. (Note that the industry classification is based on the worker's main job. Part-time is defined as less than 35 hours a week.)

The primary takeaways from the chart are:

  1. The percent of the workforce that is part time varies greatly across industries (compare for example, durable goods manufacturing with restaurants).
  2. All industries have a greater share of PTNER workers than PTER workers (for example, the restaurant industry in 2013 had 32 percent of workers who said they were PTNER and about 13 percent who declared themselves as PTER).
  3. All industries had a greater share of PTER workers in 2013 than in 2007 (all the lollipops point upwards).
  4. Most industries have a lower share of PTNER workers than in the past (most of the lollipops lean to the left).
  5. Most industries have a greater share of part-time workers (PTER + PTNER) than in the past (the increase in PTER exceeds the decline in PTNER for most industries).

Another fact that is a bit harder to see from this chart is that in 2007, industries with the largest part-time workforces did not necessarily have the largest PTER workforces. In 2013, it was more common for a large part-time workforce to be associated with a large PTER workforce. In other words, the growth in part-time worker utilization in industries such as restaurants and some segments of retail has bought with it more people who are working part-time involuntarily.

So the increase in PTER since 2007 is widespread. But is that a secular trend? If it is, then the increase in the PTER share would be evident since the recession as well. The next lollipop chart presents evidence by comparing 2013 with 2012:

This chart shows a recent general improvement. In fact, 25 of the 36 industries pictured in the chart above have experienced a decline in the share of PTER, and 21 of the 36 have a smaller portion working part-time in total. Exceptions are concentrated in retail, an industry that represents a large share of employment. In total, 20 percent of people are employed in industries that experienced an increase in PTER from 2012 to 2013. So while overall there has been a fairly widespread (but modest) recent improvement in the situation, the percent of the workforce working part-time for economic reasons remains elevated compared with 2007 for all industries. Further, many people are employed in industries that are still experiencing gains in the share that is PTER.

Why has the PTER share continued to increase for some industries? Are people who normally work full-time jobs still grasping those part-time retail jobs until something else becomes available, has there been a shift in the use of part-time workers in those industries, or is there a greater demand for full-time jobs than before the recession? We'll keep digging.

Photo of John RobertsonBy John Robertson, a vice president and senior economist, and


Photo of Ellyn TerryEllyn Terry, a senior economic analyst, both of the Atlanta Fed's research department


July 10, 2014

Introducing the Atlanta Fed's GDPNow Forecasting Model

The June 18 statement from the Federal Open Market Committee opened with this (emphasis mine):

Information received since the Federal Open Market Committee met in April indicates that growth in economic activity has rebounded in recent months.... Household spending appears to be rising moderately and business fixed investment resumed its advance, while the recovery in the housing sector remained slow. Fiscal policy is restraining economic growth, although the extent of restraint is diminishing.

I highlighted the business fixed investment (BFI) part of that passage because it contracted at an annual rate of 1.2 percent in the first quarter of 2014. Any substantial turnaround in growth in gross domestic product (GDP) from its dismal first-quarter pace would seem to require that BFI did in fact resume its advance through the second quarter.

We won't get an official read on BFI—or on real GDP growth and all of its other components—until July 30, when the U.S. Bureau of Economic Analysis (BEA) releases its advance (or first) GDP estimates for the second quarter of 2014. But that doesn't mean we are completely in the dark on what is happening in real time. We have enough data in hand to make an informed statistical guess on what that July 30 number might tell us.

The BEA's data-construction machinery for estimating GDP is laid out in considerable detail in its NIPA Handbook. Roughly 70 percent of the advance GDP release is based on source data from government agencies and other data providers that are available prior to the BEA official release. This information provides the basis for what have become known as "nowcasts" of GDP and its major subcomponents—essentially, real-time forecasts of the official numbers the BEA is likely to deliver.

Many nowcast variants are available to the public: the Wall Street Journal Economic Forecasting Survey, the Philadelphia Fed Survey of Professional Forecasters, and the CNBC Rapid Update, for example. In addition, a variety of proprietary nowcasts are available to subscribers, including Aspen Publishers' Blue Chip Publications, Macroeconomic Advisers GDP Tracking, and Moody's Analytics high-frequency model.

With this macroblog post, we introduce the Federal Reserve Bank of Atlanta's own nowcasting model, which we call GDPNow.

GDPNow will provide nowcasts of GDP and its subcomponents on a regularly updated basis. These nowcasts will be available on the pages of the Atlanta Fed's Center for Quantitative Economic Research (CQER).

A few important notes about GDPNow:

  • The GDPNow model forecasts are nonjudgmental, meaning that the forecasts are taken directly from the underlying statistical model. (These are not official forecasts of either the Atlanta Fed or its president, Dennis Lockhart.)
  • Because nowcasts are often based on both modeling and judgment, there is no reason to expect that GDPNow will agree with alternative forecasts. And we do not intend to present GDPNow as superior to those alternatives. Different approaches have their pluses and minuses. An advantage of our approach is that, because it is nonjudgmental, our methodology is easily replicable. But it is always wise to avoid reliance on a single model or source of information.
  • GDPNow forecasts are subject to error, sometimes substantial. Internally, we've regularly produced nowcasts from the GDPNow model since introducing an earlier version of it in an October 2011 macroblog post. A real-time track record for the model nowcasts just before the BEA's advance GDP release is available on the CQER GDPNow webpage, and will be updated on a regular basis to help users make informed decisions about the use of this tool.

So, with that in hand, does it appear that BFI in fact "resumed its advance" last quarter? The table below shows the current GDPNow forecasts:

We will update the nowcast five to six times each month following the releases of certain key economic indicators listed in the frequently asked questions. Look for the next GDPNow update on July 15, with the release of the retail trade and business inventory reports.

If you want to dig deeper, the GDPNow page includes downloadable charts and tables as well as numerical details including the model's nowcasts for GDP, its subcomponents, and how the subcomponent nowcasts are built up from both the underlying source data and the model parameters. This working paper supplies the model's technical documentation. We hope economy watchers find GDPNow to be a useful addition to their information sets.

Photo of Pat HigginsBy Pat Higgins, a senior economist in the Atlanta Fed's research department