The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.
- BLS Handbook of Methods
- Bureau of Economic Analysis
- Bureau of Labor Statistics
- Congressional Budget Office
- Economic Data - FRED® II, St. Louis Fed
- Office of Management and Budget
- Statistics: Releases and Historical Data, Board of Governors
- U.S. Census Bureau Economic Programs
- White House Economic Statistics Briefing Room
April 11, 2016
The Rise of Shadow Banking in China
China's banking system has suffered significant losses over the past two years, which has raised concerns about the health of China's financial industry. Such losses are perhaps not all that surprising. Commercial banks have been increasing their risk-taking activities in the form of shadow lending. See, for example, here, here, and here for some discussion of the evolution of China's shadow banking system.
The increase in risk taking by banks has occurred despite a rapid decline in money growth since 2009 and the People's Bank of China's efforts to limit credit expansions to real estate and other industries that appear to be over capacity.
One area of expanded activity has been investment in asset-backed "securities" by China's large non-state banks. This investment has created potentially significant risks to the balance sheets of these institutions (see the charts below). Using the micro-transaction-based data on shadow entrusted loans, Chen, Ren, and Zha (2016) have provided theoretical and empirical insights into this important issue (see also this Vox article that summarizes the paper).
Recent regulatory reforms in China have taken a positive step to try to limit such risk-taking behavior, although the success of these efforts remains to be seen. An even more challenging task lies ahead for designing a comprehensive and sustainable macroprudential framework to support the healthy functioning of China's traditional and shadow banking industries.
April 04, 2016
Which Wage Growth Measure Best Indicates Slack in the Labor Market?
The unemployment rate is close to what most economists think is the level consistent with full employment over the longer run. According to the Federal Open Market Committee's latest Summary of Economic Projections, the unemployment rate is currently only 15 basis points above the natural rate. Yet, average hourly earnings (AHE) for production and nonsupervisory workers in the private sector increased a paltry 2.3 percent in March from a year earlier (as did the AHE of all private workers), and is barely above its average course of 2.1 percent since 2009.In contrast, the Atlanta Fed's Wage Growth Tracker (WGT) suggests that wage growth has been increasing. The February WGT reading was 3.2 percent (the March data will be available later in April), considerably higher than its post-2009 average of 2.3 percent.
Why is there such a large difference between these measures of wage growth? Besides differences in data sources, the primary reason is that they measure fundamentally different things. The WGT is an estimate of the wage growth of continuously employed workers—the same worker's wage is measured in the current month and a year earlier.
In contrast, the AHE measure is an estimate of the change in the typical wage of everyone employed this month relative to everyone employed a year earlier. Most of these workers are continuously employed, but some of those employed in the current month were not employed the prior year, and vice versa. These changes in the composition of employment can have a significant effect.
A recent study by Mary C. Daly, Bart Hobijn, and Benjamin Pyle at the San Francisco Fed shows that while growth in wages tends to be pushed higher by the wage gains of continuously employed workers, the net effect of entry and exit into employment tends to put a drag on the growth in wages. Moreover, the magnitude of the entry/exit drag can be relatively large, varies over time, and differs by the type of entry and exit.
For example, older workers who have retired and left the workforce tend to come from the higher end of the wage distribution, and their absence from the current period wage pool exerts downward pressure on the typical wage. The greater number of baby boomers starting to retire is having an even larger depressing effect on growth in wages than in the past. Because the WGT looks only at continuously employed workers, it is not influenced by these net entry/exit effects.
To the extent that firms adjust the pay for incumbent workers in response to labor market pressures to attract and retain workers, the WGT should reasonably capture changes in the tightness of the labor market.
Economists at the Conference Board modeled the relationship between different wage growth series and measures of labor market slack. One of the slack measures they use is the unemployment gap—the difference between an estimate of the natural rate of unemployment and the actual unemployment rate.To illustrate their findings, the following chart shows the WGT and AHE measures along with the unemployment gap lagged six months (using the Congressional Budget Office estimate of the natural rate).
The WGT appears to move more closely with the lagged unemployment gap than does the growth in AHE, and a comparison of the correlation coefficients confirms the stronger relationship with the WGT. The correlation between the lagged unemployment gap and the change in average hourly earnings is 0.75.
In contrast, the correlation with the wage growth tracker is higher at 0.93. Moreover, the unemployment gap-AHE relationship appears to be particularly weak since the Great Recession. The correlation since 2009 falls to just 0.08 for the AHE, whereas the WGT correlation is still 0.93.
Our colleagues at the San Francisco Fed concluded their analysis of the effect of flows into and out of the employment on wage growth by suggesting that:
"... wage growth measures that focus on the continuously full-time employed are likely to do a better job of gauging labor market strength, since they are constructed to more clearly capture the wage dynamics associated with improving labor market conditions. The Federal Reserve Bank of Atlanta's Wage Growth Tracker is an example."
That assessment is consistent with the Conference Board study, and suggests that labor markets may be tighter than is commonly believed based on sluggish growth in measures of average wages such as AHE.
March 15, 2016
Collateral Requirements and Nonbank Online Lenders: Evidence from the 2015 Small Business Credit Survey
Businesses can secure a bank loan by offering collateral—typically a business asset such as equipment or real estate. However, the recently released 2015 Small Business Credit Survey (SBCS) Report on Employer Firms,conducted by seven regional Reserve Banks, found that 63 percent of business owners who had borrowed also used their personal assets or guarantee to secure financing. Surprisingly, the use of personal collateral was common not only among startups. Older and relatively larger small firms (see the following chart) also relied heavily on personal assets.
Source: 2015 Small Business Credit Survey
Note: "Unsure", "None", and "Other" were also options but are not shown on the chart.
Alternative lending options also exercised
Not every small business owner has sufficient hard assets, such as real estate or equipment, that can be used as collateral to secure a traditional bank loan or line of credit. For these circumstances, there are options such as credit cards and products offered by nonbank lenders (mostly operating online) that have less stringent underwriting requirements than banks. Many online nonbank lenders advertise unsecured loans or require only a general lien on business assets, without valuing those business assets.
In the 2015 SBCS, 20 percent of small firms seeking loans or lines of credit applied at nonbank online lenders. These lenders have a good reputation for quick application turnaround, and the collateral requirements can be looser than those applied by traditional lenders. But when borrowers were asked about their overall experience, only a net 15 percent of businesses approved at nonbank online lenders were satisfied (40.6 percent were satisfied and 25.3 percent were dissatisfied). In contrast, small banks received a relatively high net satisfaction score of 75 percent (see the chart).
Source: 2015 Small Business Credit Survey Report on Employer Firms
1 Satisfaction score is the share satisfied with lender minus the share dissatisfied.
2 "Online lenders" are defined as alternative and marketplace lenders, including Lending Club, OnDeck, CAN Capital, and PayPal Working Capital.
3 "Other" includes government loan funds and community development financial institutions.
The survey also showed that high interest rates were the primary reason for dissatisfaction at nonbank online lenders (see the chart).
Source: 2015 Small Business Credit Survey Report on Employer Firms
Note: Respondents could select multiple options. Select responses shown due to low observation count.
Merchant cash advances make advances
Most applicants to nonbank online lenders were seeking loans and lines of credit, but some were seeking a product that tends to be particularly expensive relative to other finance options: merchant cash advances (MCA). MCAs have been around for decades, but their popularity has risen in the wake of the financial crisis. Typically a lump-sum payment in exchange for a portion of future credit card sales, the terms of MCAs can be enticing because repayment seems easier than paying off a structured business loan that requires a fixed monthly payment. Instead, the lender is paid back as the business generates revenue, in theory making cash flow easier to manage.
One potential challenge for users of MCA products is interpreting the repayment terms. Instead of displaying an annual percentage rate (APR), MCAs are usually advertised with a "buy rate" (typically 1.2 to 1.4). For example, a buy rate of 1.3 on $100,000 would require the borrower to pay back $130,000. However, a percentage of the principal is not the same as an APR. The table below compares total interest payments made on a 1.3 MCA versus a 30 percent APR business loan repaid over 12 months and over six months. With a 12-month business loan, a 30 percent APR would equal total interest payments of roughly $17,000. With a six-month business loan, repayment would include about $9,000 in interest.
Because an MCA is structured as a commercial transaction instead of a loan, it is regulated by the Uniform Commercial Code in each state instead of by banking laws such as the Truth in Lending Act. Consequently, the provider does not have to follow all of the regulations and documentation requirements (such as displaying an APR) associated with making loans.
Converting a buy rate into an APR is not straightforward for many potential users, as was made clear in a recent online lending focus group study with small business owners conducted by the Cleveland Fed. When asked what the APR was on a $40,000 MCA that required a repayment of $52,000 (the same as a 1.3 buy rate), their answers were the following: (Product A is the MCA type of product; see the study for exactly how it was presented to respondents.)
Source: Federal Reserve Bank of Cleveland
The correct answer is that "it depends on how long it takes to pay back." For example, if the debt is repaid over six months, the APR would be 110 percent (as this calculator shows).
Nonbank online lenders can fill gaps in the borrowing needs of small business. But there may also be a role for greater clarity to ensure borrowers understand the terms they are signing up for. In a September 2015 speech, Federal Reserve Governor Lael Brainard highlights one self-policing movement already well under way:
Some have raised concerns about the high APRs associated with some online alternative lending products. Others have raised concerns about the risk that some small business borrowers may have difficulty fully understanding the terms of the various loan products or the risk of becoming trapped in layered debt that poses risks to the survival of their businesses. Some industry participants have recently proposed that online lenders follow a voluntary set of guidelines designed to standardize best practices and mitigate these risks. It is too soon to determine whether such efforts of industry participants to self-police will be sufficient. Even with these efforts, some have suggested a need for regulators to take a more active role in defining and enforcing standards that apply more broadly in this sector.
Many, but not all, nonbank online lenders have already signed the Small Business Borrower Bill of Rights. Results from the 2015 Small Business Credit Survey Report on Employer Firms can be found on our website.
February 17, 2016
Are Paychecks Picking Up the Pace?
From the minutes of the January 26–27 meeting of the Federal Open Market Committee, it's clear that many participants saw tightening labor market conditions during 2015:
In their comments on labor market conditions, participants cited strong employment gains, low levels of unemployment in their Districts, reports of shortages of workers in various industries, or firming in wage increases.
Based on the Atlanta Fed's Wage Growth Tracker (WGT), the median annual growth in hourly wage and salary earnings of continuously employed workers in 2015 was 3.1 percent—up from 2.5 percent in 2014 and 2.2 percent in 2013. That is, the typical wage growth of workers employed for at least 12 months appears to be trending higher.
However, wage growth by job type varies considerably. For example, the WGT for part-time workers has been unusually low since 2010. The following chart displays the WGT for workers currently employed in part-time and full-time jobs. For those in part-time jobs, the WGT was 1.9 percent in 2015, versus 3.3 percent for those in full-time jobs. The part-time/full-time wage growth gap has closed somewhat in the last couple of years but is still large relative to its size before the Great Recession. Note that full-time WGT is similar to the overall WGT because most workers captured in the WGT data work full-time (81 percent in 2015).
In addition to hours worked, median wage growth also tends to vary across occupation. The following chart plots the WGT for workers in low-skill jobs, versus those in mid- and high-skill jobs. (We define low-skill jobs as those in occupations related to food preparation and serving; building and grounds cleaning; and maintenance, protection, and personal care services.)
Notably, after lagging during most of the recovery, median wage growth in low-skill occupations increased 2.8 percent in 2015, versus 2.0 percent in 2014 and compared to 3.2 and 2.7 percent for other occupations in 2015 and 2014, respectively.
The improvement in wage growth for low-skill occupations seems mostly attributable to full-time workers; wage growth for people in low-skill jobs working part-time was about half that (1.6 percent versus 3.0 percent) of those working full-time (see the chart).
This pickup in low-skill wage growth fits with some anecdotal reports we've been hearing. Some of our contacts in the Southeast have reported increasing wage pressure for workers in lower-skill occupations within their businesses. One can also see evidence of growing tightness in the market for low-skill jobs in the help-wanted data. As the following chart shows, the ratio of unemployed to online job postings for low-skill jobs is always higher than for middle- and high-skill occupations. But the ratio for low-skill jobs is now well below its prerecession level, and the tightness has increased during the last two years.
The take-away? Wage growth for continuously employed workers appears to have picked up some steam in 2015, and the recent trend in wage growth is positive across a variety of job characteristics. Wage growth for people in lower-skill jobs has increased during the last couple of years, consistent with evidence of increasing tightness in the market for those types of jobs. The largest discrepancy in wage growth appears to be among part-time workers, whose median gain in hourly wages in 2015 still fell well short of those in full-time jobs.
February 05, 2016
Introducing the Refined Labor Market Spider Chart
In January 2013, Atlanta Fed research director Dave Altig introduced the Atlanta Fed's labor market spider chart in a macroblog post.
In a follow-up post that June, Atlanta Fed colleague Melinda Pitts and I introduced a dedicated page for the spider chart located at the Center for Human Capital Studies (CHCS) webpage. It shows the distribution of 13 labor market indicators relative to their readings just before the 2007–09 recession (December 2007) and the trough of the labor market following that recession (December 2009). The substantial improvement in the labor market during the past three years is quite evident in the spider chart below.
As of December 2012, none of the indicators had yet reached their prerecession levels, and some had a long way to go. Now, many of these indicators are near their prerecession values—and some have blown by them.
To make the spider chart more relevant in an environment with considerably less labor market slack than three years ago, we are introducing a modified version, which you can see here. Below is an example of a chart I created using the menu-bars on the spider chart's web page:
In this chart, I plot the May 2004 and November 2015 percentile ranks of labor market indicators relative to their distributions since March 1994. As with the previous spider chart, indicators such as the unemployment rate, where larger values indicate more labor market slack, have been multiplied by –1. The innermost and outermost rings represent the minimum and maximum values of the variables from March 1994 to January 2016. The three dashed gray rings in between are the 25th, 50th, and 75th percentiles of the distributions. For example, the November 2015 value of 12-month average hourly earnings growth (2.26 percent) is the 23rd percentile of its distribution. This means that 23 percent of the other monthly observations on hourly earnings growth since March 1994 are lower than it is.
I chose May 2004 and November 2015 because they had the last employment situation reports before "liftoffs" of the federal funds rate. November 2015 appears to be stronger than May 2004 for some indicators (job openings, unemployment rate, and initial claims) and weaker for others (hires rate, work part-time for economic reasons, and the 12-month growth rate of the Employment Cost Index).
The average percentile ranks of the variables for these two months are similar, as the chart below depicts:
Also shown in the chart is the Kansas City Fed's Level of Activity Labor Market Conditions Indicator. It is a sum of 24 not equally weighted labor market indicators, standardized over the period from 1992 to the present. In spite of its methodological and source-data differences with the average percentile rank measure plotted above, it tracks quite closely, especially since 2004. However, as shown in the spider chart that I referred to above, there is quite a bit of variation within the indicators that may provide additional information to our analysis of the average trends.
We made a number of other changes to the spider chart to ensure it reflects current labor market issues. These changes are documented in the FAQs and "Indicators" sections of the new spider chart page. Of particular note, users can choose not only the years for which they wish to track information, but also the period of reference that provides the basis of the spider chart. The payroll employment variable is now the three-month average change rather than a level. Temporary help services employment has been dropped, and two measures of 12-month compensation growth and the employment-population ratio (EPOP) for "prime-age workers" (25 to 54 years) have been added.
Some care should be taken when comparing recent labor market data values with those 10 or more years ago as structural changes in the labor market might imply that a "normal" value today is different than a "normal" value in, say, 2004. The variable choices for the refined spider chart were made to mitigate this problem to some extent. For example, we use the prime-age EPOP as a crude adjustment for population aging, putting downward pressure on the labor force participation rate and EPOP over the past 10 years (roughly 2 percentage points). This doesn't entirely resolve the comparability issue since, within the prime-age population, the self-reporting rate of illness or disability as a reason for not wanting a job has increased about 1.5 percentage points since 1998 (see the macroblog posts here and here and the CHCS Labor Force Participation Dynamics webpage). If this increase in disability reporting is partly structural—and a Brookings study by Fed economist Stephanie Aaronson and others concludes it is—some of the decline in the prime-age EPOP since the late 1990s may not be a result of a weaker labor market per se.
Other variables in the spider chart may have had structural changes as well. For example, a study by San Francisco Fed economists Rob Valleta and Catherine van der List concludes that structural factors explain just under half of the rise in the share of workers employed part-time for economic reasons over the 2006 to 2013 period.
To partially account for structural changes in trends, we allow the user to select one of 11 time periods over which the distributions are calculated. The default period is March 1994 to present, which is what was used in the example above, but users can choose a window as short as five years where, presumably, structural changes are less important. A trade-off with using a short window is that a "normal" value may not produce a result close to the median. For example, the median unemployment rate is 5.6 percent since March 1994 and 7.3 percent since February 2011. The latter value is much farther away from the most recent estimates of the natural rate of unemployment from the Congressional Budget Office and the Survey of Professional Forecasters (both 5.0 percent).
In our June 2013 macroblog post introducing the spider chart, we wrote that we would reevaluate our tools and determine a more appropriate way to monitor the labor market when "the labor market has turned a corner into expansion." The new spider chart is our response to the stronger labor market. We hope users find the tool useful.
January 29, 2016
Shrinking Labor Market Opportunities for the Disabled?
The labor force participation rate (LFPR) among prime-age (25–54 years old) people averaged 80.8 percent in 2015, down 1.8 percentage points (2.6 million people) from 2009, according to the U.S. Bureau of Labor Statistics. According to our calculations from the Current Population Survey (CPS), a drop in LFPR among individuals with disabilities accounts for about a fifth of that decline.
Many people with disabilities are active in the labor force, working or looking for work. But the disabled LFPR has fallen a lot in recent years—it's down from 39.3 percent in 2009 to 34.5 percent in 2015. In other words, for some reason, more prime-age individuals with disabilities have opted out of the labor market.
A rising share of the prime-age population with a disability is not the culprit. In fact, the 2015 average disability rate was 6.4 percent, the same as in 2009. It is possible that the severity rather than incidence of disabilities has increased in recent years; labor market attachment does vary with type of disability, as this report shows.
But we suspect that the relatively large decline in disabled labor market attachment probably has also to do with shifts in employment opportunities for those with a disability. Some insight into this issue can be seen by looking at the change in employment shares in occupations that tend to have relatively low pay.
Workers with a disability tend to make less than nondisabled workers. We estimate the median wage of a worker with a disability in 2009 to have been 76 percent that of a nondisabled worker. In 2015, the relative median wage had declined to 74 percent. This drop is partly related to a relative increase in the share of employment for workers with a disability in low-paying occupations (which we define as jobs in personal care, food services, janitorial services, etc.), as the chart shows. The employment news has not been all bad for workers with a disability, however. There has been a rise in the share of employment of people with disabilities in higher-paying occupations since 2009, although they do tend to earn less than other workers in those types of occupations.
For some workers with disabilities, the financial return to employment versus nonemployment may have become somewhat less attractive in recent years. One factor related to the decision to engage in the labor market is the ability to collect Social Security Disability Insurance (SSDI). SSDI claims rose notably when the unemployment rate was high, which is consistent with the idea that the expected return to labor market activity for some individuals with a disability declined.
Job seekers with a disability have also struggled to find jobs offering the hours they desire. For example, the share of unemployed people finding full-time or voluntary part-time employment within a month, or "the hours-finding rate," is much lower for the prime-age disabled than for the nondisabled, and this share has improved relatively less over the recovery. Between 2009 and 2015, the average disabled hours-finding rate improved 4.7 percentage points, from 9.5 to 14.2 percent. During the same period, the nondisabled hours-finding rate increased 6.3 percentage points, from 13.0 to 19.3 percent.
The incidence of disability among prime-age individuals has not increased in recent years. But the labor market attachment of the disabled has declined, and this decline accounts for about one-fifth of the 1.8 percentage point fall in prime-age labor force participation between 2009 and 2015. Those with disabilities already have a harder time finding well-paying jobs, but that difficulty appears to have increased in that time span.
By John Robertson, a senior policy adviser in the Atlanta Fed's research department, and
January 15, 2016
Are Long-Term Inflation Expectations Declining? Not So Fast, Says Atlanta Fed
"Convincing evidence that longer-term inflation expectations have moved lower would be a concern because declines in consumer and business expectations about inflation could put downward pressure on actual inflation, making the attainment of our 2 percent inflation goal more difficult."
—Fed Chair Janet Yellen, in a December 2, 2015, speech to the Economic Club of Washington
To be sure, Chair Yellen's claim is not controversial. Modern macroeconomics gives inflation expectations a central role in the evolution of actual inflation, and the stability of those expectations is crucial to the Fed's ability to achieve its price stability mandate.
The real question on everyone's mind is, of course, what might constitute "convincing evidence" of changes in inflation expectations. Recently, several economists, including former Treasury Secretary Larry Summers and St. Louis Fed President James Bullard, have weighed in on this issue. Yesterday, President Bullard cited downward movements in the five-year/five-year forward breakeven rates from the five- and 10-year nominal and inflation-protected Treasury bond yields. In November, Summers appealed to measures based on inflation swap contracts. The view that inflation expectations are declining has also been echoed by the New York Fed President William Dudley and former Minneapolis Fed President Narayana Kocherlakota.
Broadly speaking, there seems to be a growing view that market-based long-run inflation expectations are declining and drifting significantly away from the Fed's 2 percent target and that this decline is troublingly correlated with oil prices.
A problem with this line of argument is that the breakeven and swap rates are not necessarily clean measures of inflation expectations. They are really better referred to as measures of inflation compensation because, in addition to inflation expectations, these measures also include factors related to liquidity conditions in the markets for these securities, technical features of the inflation protection in each security, and inflation risk premia. Here at the Atlanta Fed, we've built a model to separate these different components and isolate a better measure of true inflation expectations (IE).
In technical terms, we estimate an affine term structure model—similar to that of D'Amico, Kim and Wei (2014)—that incorporates information from the markets for U.S. Treasuries, Treasury Inflation-Protected Securities (TIPS), inflation swaps, and inflation options (caps and floors). Details are provided in "Forecasts of Inflation and Interest Rates in No-Arbitrage Affine Models," a forthcoming Atlanta Fed working paper by Nikolay Gospodinov and Bin Wei. (You can also see Gospodinov and Wei (2015) for further analysis.) Essentially, we ask: what level of inflation expectations is consistent with this entire set of financial market data? And we then follow this measure over time.
As chart 1 illustrates, we draw a very different conclusion about the behavior of long-term inflation expectations. The chart plots the five-year/five-year forward TIPS breakeven inflation (BEI) and the model-implied inflation expectations (IE) for the period January 1999–November 2015 at a weekly frequency. Unlike the raw BEI, our measure is quite smooth, suggesting that long-term inflation expectations have been, and still are, well anchored.
After making an adjustment for the inflation risk premium, we term the difference between BEI and IEs a "liquidity premium," but it really includes a variety of other factors. Our more careful look at the liquidity premium reveals that it is partly made up of factors specific to the structure of inflation-indexed TIPS bonds. For example, since TIPS are based on the non-seasonally adjusted consumer price index (CPI) of all items, TIPS yields incorporate a large positive seasonal carry yield in the first half of the year and a large negative seasonal carry yield in the second half. Chart 2 illustrates this point by plotting CPI seasonality (computed as the accumulated difference between non-seasonally adjusted and seasonally adjusted CPI) and the five-year breakeven inflation.
Redemptions, reallocations, and hedging in the TIPS market after oil price drops and global financial market turbulence can further exacerbate this seasonal pattern. Taken together, these factors are the source of correlation between the BEI measures and oil prices. To confirm this, chart 3 plots (the negative of) our liquidity premium estimate and the log oil price (proxied by the nearest futures price).
Our measure of long-term inflation expectations is also consistent with long-term measures from surveys. Chart 4 presents the median along with the 10th and 90th percentiles of the five-year/five-year forward CPI inflation expectations from the Philadelphia Fed's Survey of Professional Forecasters (SPF) at quarterly frequency. This measure can be compared directly with our IE measure. Both the level and the dynamics of the median SPF inflation expectation are remarkably close to that for our market-based IE. It is also interesting to observe that the level of inflation "disagreement" (measured as the difference between the 10th and 90th percentiles) is at a level similar to the level seen before the financial crisis.
Finally, we note that TIPS and SPF are based on CPI rather than the Fed's preferred personal consumption expenditure (PCE) measure. CPI inflation has historically run above PCE inflation by about 30 basis points. Accounting for this difference brings our measure of the level of long-term inflation expectations close to the Fed's 2 percent target.
To summarize, our analysis suggests that (1) long-run inflation expectations remain stable and anchored, (2) the seemingly large correlation of market-implied inflation compensation with oil prices arises mainly from the dynamics of the TIPS liquidity premium, and (3) long-run market- and survey-based inflation expectations are remarkably close in terms of level and dynamics over time. Of course, further softness in the global economy and commodity markets may eventually drag down long-term expectations. We will continue to monitor the pure measure of inflation expectations for such developments.
By Nikolay Gospodinov, financial economist and policy adviser; Paula Tkac, vice president and senior economist; and Bin Wei, financial economist and associate policy adviser, all of the Atlanta Fed's research department
January 07, 2016
What Occupational Projections Say about Entry-Level Skill Demand
On December 8, 2015, the U.S. Bureau of Labor Statistics (BLS) released its latest projections of labor force needs facing the U.S. economy from now until 2024.
Every two years, the BLS undertakes an extensive assessment of worker demand based on a number of factors: projected growth in the overall economy, dynamics of economic growth (such as which industries are growing fastest), labor force demographics (for example, the aging of the labor force), and expected changes in the labor force participation rate. Total worker demand includes both the number of workers needed to meet economic growth as well as the number of workers needed to replace current workers expected to retire.
A number of observations about these projections have already been identified. For example: overall employment growth will be slower, health care jobs will continue growing, and computer programmer jobs will lose ground.
In addition to the number of workers that will be in demand in different occupations in the U.S. economy, the BLS reports the skills that are needed for entry into those occupations—skills pertaining to both education levels and on-the-job training. As I perused this report, I was surprised at how much attention the press pays to the growth in high-skilled jobs at the expense of attention paid to those occupations requiring less skill but actually employ the greatest number of workers.
To be clear, the BLS does not project the educational requirements that will be needed for entering each occupation in 2024. It merely reports the most common education, training, and experience requirements needed to enter each occupation in the base year (in this case, 2014). Also it's important to note that these estimates of education needed to enter an occupation do not necessarily (and almost surely do not) match the average education of workers in that occupation at any given time, as those averages will reflect workers of many different ages and experience. The BLS gives a detailed description of how it identifies the entry-level educational and training requirements for each occupation. With those caveats in mind, let's take a look at the current distribution of jobs across the most common educational requirement for entering occupations.
The chart below tells us that, together, the typical entry-level requirement of a high school degree or less corresponds to nearly 64 percent of all jobs in the U.S. economy in 2014, while those typically requiring at least a bachelor's degree for entry represent 25.6 percent of jobs. The projected distribution of jobs in 2024 looks nearly identical: entry-level requirements (based on 2014 assessments) for 63 percent of all jobs requiring a high school degree or less and 26.2 percent requiring a bachelor's degree or more.
To be sure, the growth in higher-skill jobs far outpaces that for low- or middle-skill jobs. The number of jobs requiring a bachelor's degree or more for entry (in 2014) is expected to grow by 34 percent, whereas the number of jobs requiring less than a bachelor's degree is expected to grow by only 6 percent. This difference in growth rates reflects, in part, an expected continuation of the phenomenon of declining middle-skill jobs that my Atlanta Fed colleagues (and others) have discussed previously. Although labeled "middle-skill," entry into these occupations (such as office support and many manufacturing occupations) is not likely to require more than a high school degree.
However, even though the growth in low- and middle-skill jobs is expected to be slower than in higher-skill jobs, the total number of job openings based on predicted growth and replacement needs between 2014 and 2024 is expected to be nearly 32 million for jobs requiring less than a bachelor's degree for entry (based on 2014 assessments), with 30 million of those requiring only a high school degree or less. The total number of job openings requiring at least a bachelor's degree is expected to be about 12 million. In other words, the number of jobs requiring a high school degree or less in order to enter is twice as large as the number of jobs that require a college degree to enter.
The other side of this story, however, is that those jobs typically requiring less education at entry don't pay nearly as much as jobs requiring higher levels of education. The dollar figures in parentheses on the chart reflect the median annual salary of jobs with the different entry-level educational requirements. What we see is that while the majority of U.S. jobs require a high school degree or less at entry, those jobs pay less than half of what a job requiring at least a college degree pays.
So let's say a worker wants good job prospects (with a large number of job openings over the next decade), doesn't want to go to college, and wants to optimize chances for the highest salary possible. What is this worker to do? Fortunately, some of my colleagues at the Atlanta Fed, Cleveland Fed, and Philadelphia Fed have produced a report identifying what they call "opportunity occupations," which they define as those paying salaries higher than the geographically-adjusted national median for at least 70 percent of adults who have less than a college education. Some jobs among their top opportunity occupations are nurses, bookkeepers, first-line supervisors of retail workers, truck drivers, computer user support specialists, police officers, and electricians and workers in several other construction trades. Their report also identifies the U.S. metropolitan areas possessing a high share of opportunity occupations.Even though the share of jobs in the U.S. economy requiring less than a college degree at entry is getting smaller (very slowly), the largest number of jobs in the economy is, by far, jobs requiring less than a college degree at entry, and those jobs offer a wide range of options that pay above the national median wage.
November 05, 2015
A Closer Look at Changes in the Labor Market
The Atlanta Fed's Center for Human Capital Studies hosted its annual employment conference on October 1–2, 2015, organized once again by Richard Rogerson (Princeton University), Robert Shimer (University of Chicago), and the Atlanta Fed's Melinda Pitts. This macroblog post provides a summary of the papers presented at the conference.
Many measures of labor market performance remain at relatively low levels compared with levels seen before the Great Recession. A key question for policymakers and academic researchers is the extent to which these changes reflect a slow recovery from a large cyclical shock—or do they simply represent the "new normal"? This conference brought together researchers studying several dimensions of these changes in labor-market outcomes. A common theme is that current labor market outcomes largely reflect the ongoing effect of secular trends that predated the Great Recession.
Recent empirical work has highlighted that the U.S. economy, and in particular the labor market, has seen a pronounced downward trend in several measures of "dynamism." Prominent among these measures are decreases in job and worker flows as well as in the entry rate of new establishments. A key challenge is to uncover the driving forces behind these trends and determine whether they reflect a worsening of U.S. economic performance.
Three papers addressed these changes. In "Changing in Business Dynamism: Volatility of vs. Responsiveness to Shocks?," Decker, Haltiwanger, Jarmin, and Miranda pose a key question for assessing whether these declines might reflect positive versus negative forces. Specifically, if lower volatility in firm-level outcomes reflects a change in the volatility in the economic environment in which firms operate, then it might well be a positive development. On the other hand, if the decreased volatility in firm-level measures reflects less responsiveness to changes in the economic environment, then the changes may constitute a negative development. The paper notes that elements of each may be present in different sectors of the economy, but their analysis suggests that lower responsiveness to shocks is an important factor.
In a second paper on the topic, "Dynamism Diminished: The Role of Credit Conditions," Davis and Haltiwanger focus on the decline in the business entry rate and consider one particular driving force: the role of housing wealth in facilitating start-up entrepreneurship. They ask whether cities that had the largest drops in housing wealth also had larger drops in entrepreneurial activity, holding other factors constant. Their analysis finds a strong correlation between the two, suggesting that the loss of housing wealth from the Great Recession has had a significant negative effect on the rate of business startups.
A third paper on the theme of diminished dynamics offered a somewhat different perspective. In their paper "Understanding the Thirty Year Decline in the Start-Up Rate: A General Equilibrium Approach," Karahan, Pugsley, and Sahin offer a more innocuous interpretation of the trend decline in the entry rate. They note that the growth of the U.S. labor force has slowed in the last 30 years, because of the aging of the baby boomers as well as the slowdown in the growth rate of women in the labor force. Standard models of industry equilibrium imply that this will require a slowdown in the rate of growth of firms, achieved through a decrease in the rate of entry. They also note that standard models imply that substantial differences in cohort dynamics in response to such a change will not be evident, and they depict this in the data.
Secular changes in inequality have received much attention in recent years. Two papers examined the nature of these changes. In "Firming Up Inequality," Bloom, Guvenen, Price, Song, and von Wachter use tax return data from the Social Security Administration to examine the underlying sources of increased income inequality since 1978. A key feature of this analysis is that it is based on tax return data for the universe of individuals, making it much more extensive and reliable than estimates based on smaller samples and self-reported measures of income. The authors find that the rise in income inequality is dominated by an increase in income dispersion across firms rather than within firms, which seems to result from an increase in the extent of sorting of workers across firms. The authors suggest that this increase reflects a change in the way firms are organized. The authors also show that executive pay plays essentially no role in the overall rise of inequality.
Lochner and Shin also examine the dynamics of inequality in "Understanding Earnings Dynamics: Identifying and Estimating the Changing Roles of Unobserved Ability, Permanent and Temporary Shocks." This paper focuses on changes in labor earnings among males from 1970 to 2008. Unlike the previous paper that focused on dispersion between and within firms, this paper focuses on permanent versus transitory components of inequality and the extent to which changes in inequality reflect changes in the price of unobserved skill. The paper provides a detailed decomposition of the evolution of these various components over a 40-year period. The decomposition between permanent and transitory components is of central concern since higher transitory variance averages out over time at the individual level. One key finding is that since 1990, the dispersion of permanent shocks has increased, especially for low-income workers.
Hall and Schulhofer-Wohl analyze changes in match efficiency in the U.S. economy since 2001 in their paper "Measuring Job Finding Rates and Matching Efficiency with Heterogeneous Job Seekers." Standard estimates based on an aggregate matching function that treats all workers as identical imply that matching efficiency has deteriorated dramatically during the Great Recession and its aftermath. The authors show that if one takes into account heterogeneity in matching rates for workers with different observable characteristics, a very different picture emerges. In particular, although a decrease is still evident in the aftermath of the Great Recession, this decrease reflects a continuation of an existing downward trend. The key implication is that lower matching rates currently found in the data reflect a secular trend.
In "The Great Reversal in the Demand for Skill and Cognitive Tasks," Beaudry, Green, and Sand offer a new perspective on secular trends in the labor market. Key to their explanation is that the boom prior to 2000 is associated with investment in the new general-purpose technology associated with information technology. Their theory holds that this technology is put in place during a period of high investment demand and high demand for skilled labor. But once the new technology is in place, it requires much less high-skilled labor to maintain or operate it. In this "de-skilling" phase, high-skilled individuals will move to jobs that are lower in the skill spectrum, thereby displacing individuals with lower skill levels to either move farther down ladder or even out of the labor force. The authors argue that this de-skilling phase began sometime around 2000 and was somewhat obscured prior to the Great Recession. The paper presents a stylized model of this process and presents several pieces of empirical evidence consistent with this dynamic. The key implication is that recent developments in the labor market indicate secular trends.
Autor, Figlio, Karbownik, Roth and Wasserman examine a different trend in U.S. labor market outcomes. In "Family Disadvantage and the Gender Gap in Behavioral and Education Outcomes," the authors examine the growing gap between male and female educational attainment. This gap is particularly large for children from disadvantaged backgrounds. The authors evaluate the hypothesis that the gap reflects differences in the sensitivity of boys and girls to adverse environments. They use data from Florida that allow them to study brother-sister pairs, allowing them to control for family environment. The key finding is that their study supports this hypothesis, though they are unable to identify which specific factors might be at work.
Full papers for most of these presentations are available on the Atlanta Fed's Center for Human Capital Studies website.
October 19, 2015
Should We Be Concerned about Declines in Labor Force Growth?
For the second month in a row, the October jobs report from the U.S. Bureau of Labor Statistics (BLS) has revealed a decline in the labor force. From August to September, the labor force lost a seasonally adjusted 350,000 participants. And the August number of participants was a seasonally adjusted 41,000 below July's level. Although two months don't necessarily make a trend, observers have noticed the declines in the labor force (here and here, for example), and they deserve some attention.
Economists might be concerned about these labor force declines for two reasons. First, these losses might indicate that the current unemployment rate doesn't accurately reflect a strong labor market. Second, our economy needs labor to make things, perform services, and continue to grow. Let's take a look at the evidence supporting these two concerns.
Concerns about a shadow weak labor market
Two pieces of evidence suggest that the declines in the labor force don't indicate a weak labor market: employment growth and the reasons people cite for being out of the labor force. Employment growth is robust. According to the Atlanta Fed's Jobs Calculator, the labor market needs to create an average of only 112,000 jobs per month to maintain its relatively low unemployment rate of 5.1 percent. During 2015, the economy has created, on average, 198,000 jobs per month.
But we might be concerned if the workers leaving the labor market were entering into the no-man's land of the marginally attached, a term describing those who want a job, are available to work, have looked for work in the previous year, but recently have stopped looking. Some of these people have stopped looking explicitly because they think jobs prospects are poor (called "discouraged workers"). Others have stopped looking for other reasons such as attending school or taking care of family members. If these categories of nonparticipants were absorbing a large share of those leaving the labor force, we could be concerned that they would, at any moment, reenter the labor market and push that unemployment rate right back up again. The chart below tells us that this possibility is unlikely.
The chart decomposes the year-over-year changes in the total number of labor force participants into changes in the population and the negative changes among reasons given for nonparticipation in the labor force. (I use year-over-year changes because the reasons given for not being in the labor force are not seasonally adjusted.) Year-over-year changes in the population have been consistent in their contributions to changes in the labor force, propping it up. The growth in the contribution of those not wanting a job (pulling down labor force growth) has been fairly striking.
The share of people giving other reasons for not being in the labor force (discouraged, not available, etc.)—in addition to making relatively small contributions to changes to the labor force—has mostly been shrinking since April, meaning that they cannot explain the recent slowing of labor force growth. In other words, only a very small part of the growth in nonparticipants has come from those marginal workers who are most likely to reenter the labor force. So the first fear—that this declining labor force growth is producing a false sense of security in a relatively strong labor market—appears unfounded.
Threats to economic growth
Labor is an important component in the production process. Short of dramatic technological advancements, both the manufacturing and service sectors need a consistent source of labor to fuel output. Even though the economy appears to be on the right track with respect to job creation, ongoing declines in labor force growth could pose a challenge to economic growth. Additionally, as employers compete for fewer workers, we would expect wages to be bid up. Keep an eye on the Atlanta Fed's wage tracker to see how slowing labor force growth plays out in wages.
- Was May's Drop in Labor Force Participation All Bad News?
- Wage Growth for Job Stayers and Switchers Added to the Atlanta Fed's Wage Growth Tracker
- Experts Debate Policy Options for China's Transition
- It’s Not Just Millennials Who Aren't Buying Homes
- After the Conference, Another Look at Liquidity
- Moving On Up
- Putting the Wage Growth Tracker to Work
- Can Two Wrongs Make a Right?
- Are People in Middle-Wage Jobs Getting Bigger Raises?
- GDPNow and Then
- June 2016
- May 2016
- April 2016
- March 2016
- February 2016
- January 2016
- November 2015
- October 2015
- September 2015
- August 2015
- Business Cycles
- Business Inflation Expectations
- Capital and Investment
- Capital Markets
- Data Releases
- Economic conditions
- Economic Growth and Development
- Exchange Rates and the Dollar
- Fed Funds Futures
- Federal Debt and Deficits
- Federal Reserve and Monetary Policy
- Financial System
- Fiscal Policy
- Health Care
- Inflation Expectations
- Interest Rates
- Labor Markets
- Latin America/South America
- Monetary Policy
- Money Markets
- Real Estate
- Saving, Capital, and Investment
- Small Business
- Social Security
- This, That, and the Other
- Trade Deficit
- Wage Growth