Today's employment report for the month of February maybe took a bit of drama out of one key question going into the next meeting of the Federal Open Market Committee (FOMC): What will happen to the FOMC's policy language when the economy hits or passes the 6.5 percent unemployment rate threshold for considering policy-rate liftoff? With the unemployment rate for February checking it at 6.7 percent, a breach of the threshold clearly won't have happened when the Committee meets in a little less than two weeks.
I say "maybe took a bit of drama out" because I'm not sure there was much drama left. All you had to do was listen to the Fed talkers yesterday to know that. This is from the highlights summary of a speech yesterday by Charles Plosser, president of the Philadelphia Fed...
President Plosser believes the Federal Open Market Committee has to revamp its current forward guidance regarding the future federal funds rate path because the 6.5 percent unemployment threshold has become irrelevant.
... and this from a Wall Street Journal interview with William Dudley, president of the New York Fed:
Mr. Dudley, in a Wall Street Journal interview, also said the Fed's 6.5% unemployment rate threshold for considering increases in short-term interest rates is "obsolete" and he would advocate scrapping it at the Fed's next meeting March 18–19.
From our shop, Atlanta Fed president Dennis Lockhart echoed those sentiments in a speech at Georgetown University:
Given that measured unemployment is so close to 6.5 percent, the time is approaching for a refreshed explanation of how unemployment or broader employment conditions are to be factored into a liftoff decision.
That statement doesn't mean we in Atlanta are disregarding the unemployment rate altogether. We have for some time been describing the broader net we have cast in fishing for labor market clues. One important aspect of that broader perspective is captured in the so-called U-6 measure of unemployment, about which President Lockhart's speech gives a quick tutorial:
The data used to construct the unemployment rate come from a survey of households conducted by the Census Bureau for the Bureau of Labor Statistics. To be counted as a participant in the labor force, a respondent must give rather specific qualifying answers to questions in the survey...
Those who are available, have looked for work in the past year, but have not recently looked for work are labeled "marginally attached." They are not in the official labor force, so they are not officially unemployed. You might say they are a "shadow labor force"...
One measure that counts the marginally attached in the pool of the unemployed is U-6.
U-6 also includes working people who identify themselves as working "part time for economic reasons." These are people who want to work full time (defined as 35 hours or more) but are able only to get fewer than 35 hours of work.
The "shadow labor force" comment is based on these observations. First, in President Lockhart's words:
The makeup of the class of marginally attached workers is quite fluid. About 40 percent of the marginally attached in any given month join the official labor force in the subsequent month.
There is no new story there. The frequency with which people move from marginally attached to in the labor force has been stable for quite a while (see the chart):
President Lockhart's second observation regarding the marginally attached is more important:
But only about 10 percent of those who move into the labor force find a job right away. In effect, they went from unofficially unemployed to officially unemployed.
The chart below depicts this observation:
Relative to before the Great Recession, the frequency with which people transitioned from marginally attached to employment has fallen by about 5 percentage points.
That decline is related to this conclusion (again from President Lockhart):
Here's my point: what U-6 captures matters. Measures such as marginally attached and part time for economic reasons became elevated in the recession and have not come down materially. Said differently, broader measures of unemployment like U-6 suggest that a significant level of slack remains in our employment markets.
It is not that we have failed to see progress in the U-6 measure of labor market slack. In fact, since the end of the recession, the U-6 unemployment rate has declined about in tandem with the standard official unemployment rate (designated U-3 by the U.S. Bureau of Labor Statistics; see the chart):
What is the case is that we have failed to undo the outsized run-up in the marginally attached and people working part-time for economic reasons that occurred during the recession (see the chart):
One interpretation of these observations is that the relative increase in U-6 represents structural changes that cannot be fixed by policies aimed at stimulating spending. But we are drawn to the fact, described above, that the marginally attached are flowing into the labor market at the same pace as before the recession, but they are finding jobs at a much slower pace, making us hesitant to fully embrace a structural interpretation.
Or, as our boss said yesterday:
As a policymaker, I am concerned about the unemployed in the official labor force, but I am also concerned about the unemployed in the shadow labor force. To get close to full employment, as I think of it, would involve substantial absorption of this shadow labor force. I do not think we're near that point yet. This is one of the reasons I support continuing with a highly accommodative policy and deferring liftoff for a while longer.
But if you are looking for some good news, here it is: Though the official unemployment rate has been essentially flat for the past three months, the broader U-6 measure that we are monitoring closely has fallen by half a percentage point. More of that, and we will really be getting somewhere.
By Dave Altig, research director and executive vice president at the Atlanta Fed
A recent Wall Street Journal blog post caught our attention. In particular, the following claim:
It’s not size that matters—at least when it comes to job creation. The age of the company is a bigger factor.
The following chart shows the average job-creation rate of expanding firms and the average job-destruction rates of shrinking firms from 1987 to 2011, broken out by various age and size categories:
In the chart, the colors represent age categories, and the sizes of the dot represent size categories. So, for example, the biggest blue dot in the far northeast quadrant shows the average rate of job creation and destruction for firms that are very young and very large. The tiny blue dot in the far east region of the chart represents the average rate of job creation and destruction for firms that are very young and very small. If an age-size dot is above the 45-degree line, then average net job creation of that firm size-age combination is positive—that is, more jobs are created than destroyed at those firms. (Note that the chart excludes firms less than one year old because, by definition in the data, they can have only job creation.)
The chart shows two things. First, the rate of job creation and destruction tends to decline with firm age. Younger firms of all sizes tend to have higher job-creation (and job-destruction) rates than their older counterparts. That is, the blue dots tend to lie above the green dots, and the green dots tend to be above the orange dots.
The second feature is that the rate of job creation at larger firms of all ages tends to exceed the rate of job destruction, whereas small firms tend to destroy more jobs than they create, on net. That is, the larger dots tend to lie above the 45-degree line, but the smaller dots are below the 45-degree line.
As pointed out in the WSJ blog post and by others (see, for example, work by the Kauffman Foundation here and here), once you control for firm size, firm age is the more important factor when measuring the rate of job creation. However, young firms are more dynamic in general, with rapid net growth balanced against a very high failure rate. (See this paper by John Haltiwanger for more on this up-or-out dynamic.) Apart from new firms, it seems that the combination of youth (between one and ten years old) and size (more than 250 employees) has tended to yield the highest rate of net job creation.
By John Robertson, a vice president and senior economist in the Atlanta Fed’s research department, and
Ellyn Terry, a senior economic analyst in the Atlanta Fed's research department
Will the Federal Open Market Committee's (FOMC) current large-scale asset purchase program, so-called QE3, continue to melt away as spring arrives? The release of the minutes from the January meeting of the FOMC, along with commentary from various participants in that meeting (noted in rapid succession here, here, and here, for example) have left the distinct impression that the answer is most probably yes.
The anticipated winding down of asset purchases almost inevitably invokes a habit of language concerning what it all means for the stance of monetary policy. From the New York Times, for example, we have this (emphasis added):
When Federal Reserve officials last met at the end of January, they were surprised by the strength of the economy, cheered by the optimism of consumers and convinced they should continue to dismantle the Fed's economic stimulus campaign, according to an account the Fed released Wednesday.
The sentiment expressed in that highlighted passage is front and center at the G-20 meetings, currently taking place in Australia (again, emphasis added):
Setting the scene for this weekend's Group of 20 meetings, Australian Treasurer Joe Hockey's main challenge was to avoid appearing partial in the escalating blame-game between the U.S. and developing countries over the recent exodus of capital from emerging markets….
Emerging market countries like India and Brazil have blamed the wide-scale selloff in local stocks, bonds and currencies on the Federal Reserve's plan to exit gradually from monetary-stimulus policies, which last year began sending investors into a panic.
Here's the thing. It is not at all clear that winding down asset purchases means an exit from or dismantling of monetary stimulus, gradual or otherwise. In Atlanta Fed President Dennis Lockhart's words yesterday:
In our public remarks over much of last year, my colleagues and I stressed a couple of very important messages. First, even with the phase-out of asset purchases, the basic stance of policy remains highly accommodative. To translate, the Committee intends to keep interest rates very low. The second message was that the QE program and the Fed's policy interest-rate target are two separate tools of policy. Consequently, we can wind down the asset purchases—a program that was meant to provide temporary, supplemental "oomph" to the low interest-rate policy—and preserve the accommodative positioning of policy appropriate for the reality of our economic situation.
But those are not just words. Several months back, Jim Hamilton publicized the work of Cynthia Wu and Dora Xia (former and current students of his), who have developed a method of using term structure data to infer the "shadow," or implicit, monetary policy rate. (Follow-up posts appeared thereafter at Econbrowser—here and here—and from the crew here at macroblog.)
Just recently, the Wu-Xia data has been updated, giving us a first glance at the post-taper shadow policy rate (see the chart):
Both Treasury yields and the shadow policy rate did in fact spike last June following then-Chairman Ben Bernanke's post-FOMC press conference, wherein he signaled that the asset-purchase taper was indeed on the table. But he also made the point that bringing down the QE pillar of the Fed's policy mix is decidedly not the same thing as bringing monetary stimulus to an end, a message that was subsequently emphasized by Fed officials many times, in many forums.
Though financial market participants may have been convinced that the taper meant tightening initially, it does appear that communications and forward guidance have done the trick of more than reversing that initial impression. At the very least, the Wu-Xia calculations are consistent with that interpretation.
By Dave Altig, executive vice president and research director at the Atlanta Fed
This analysis is a companion piece to my Atlanta Fed colleague John Robertson's recent macroblog post. John's blog highlighted some findings of a recent New York Fed study by Samuel Kapon and Joseph Tracy on the employment-to-population (E/P) ratio. Their work has received considerable attention in the media and blogosphere (for example, here, here, and here). Kapon and Tracy's final chart (reproduced below) has received particular scrutiny.
The blue line represents the authors' estimate of the demographically adjusted E/P ratio purged of business-cycle effects. This line can be thought of as "trend." The chart shows that as of November 2013, the E/P ratio was only –0.7 percentage point below trend. Was the "gap" between actual and trend E/P really this small?
Attempting to answer this question requires digging into the details of Kapon and Tracy's method for estimating trend. One key excerpt is the following:
To overlay our demographically adjusted E/P ratio with the actual E/P ratio, we need to adopt a normalization… [W]e adopt the normalization that over the thirty-one years in our data sample [1982–2013] any business-cycle deviations between the actual and the adjusted E/P ratios will average to zero.
This methodology seems reasonable since one might typically expect business cycle effects to average out over 30 years. However, the 1982–2013 sample period is somewhat unusual in that the unemployment rate was elevated at both the starting and ending points.
The chart below shows estimates of three labor market gaps derived from the Congressional Budget Office's (CBO) estimates—released on February 4, 2014—of the potential labor force and the long-term natural rate of unemployment. (This rate is often referred to as the nonaccelerating inflation rate of unemployment, or NAIRU, and refers to the level of unemployment below which inflation rises.)
On average, the trend E/P ratio is below the actual rate by 0.86 percentage point. If one were to normalize the Kapon and Tracy E/P trend so that its average value was equal to CBO's trend, then the November 2013 E/P gap is about 1.5 percentage points. Whether or not the CBO estimate is the right benchmark is a matter of taste. CBO's recent estimate of NAIRU in the fourth quarter of 2013—5.5 percent—is lower than the 6 percent median estimate from the Survey of Professional Forecasters in the third quarter of 2013.
A second, more subtle issue in the Kapon and Tracy analysis is their treatment of cohorts:
We divide these individuals into 280 different cohorts defined by each individual's decade of birth, sex, race/ethnicity, and educational attainment. We assume that individuals within a specific cohort have similar career employment rate profiles. We use the 10.2 million observations [of CPS microdata] to estimate these 280 career employment rate profiles.
A well-known 2006 Brookings paper by Stephanie Aaronson and other Fed economists modeled trend labor force participation rate(LFPR) using birth-year cohorts. With estimates of trend LFPR and NAIRU, we can back out a trend E/P ratio. The chart below, adapted from Aaronson et al., plots age-group LFPRs against birth year.
We see that successive birth-year cohorts born between 1925 and 1950 had steadily increasing labor force attachment. Attachment for more recently born cohorts has leveled off and even declined slightly. People born in the 1990s have very low labor force attachment by historical standards. The inclusion of the "1990s—decade of birth" dummy variable in the Kapon and Tracy research probably implies that their model is interpreting much of this decline as structural. However, an alternative interpretation is that the decline is cyclical, because persons born after 1990 have been in an environment of high unemployment for most of their short working lives.
To gauge the sensitivity of trend or structural LFPR to how the youngest cohorts are treated, I used a stripped-down version of a model similar to Aaronson et al. Monthly LFPRs are modeled as a function of age, sex, birth date, and the CBO's estimate of the output gap during the January 1981 to January 2014 period. Time series published by the U.S. Bureau of Labor Statistics for 30 different age-sex cells are used so that the regression has 11,550 observations. Structural LFPR is constructed with the fitted values of the regression with a value of 0 percent for the output gap at all points in time. The trend E/P ratio is then backed out with the CBO's estimate of NAIRU.
The model is run with two different assumptions: First, following the approach of Aaronson et al., people born after 1986 have the same birth-year cohort effects as those born in December 1986. Second, no constraints are placed on birth-year cohort effects. Trend values of LFPR and E/P (taking on board the CBO's NAIRU) are plotted in the two charts below:
The January 2014 E/P gap with unconstrained cohort effects, as in Kapon and Tracy, is –1.0 percent, well below the –1.7 percent gap in the model with constrained cohort effects. Ultimately, both models are still very consistent with Kapon and Tracy's bottom line:
It is important to control for changing demographic factors when looking at the behavior of the E/P ratio over time. This step is particularly important today when these demographic factors are exerting downward pressure on the actual E/P rate, suggesting that the recent lack of improvement in the E/P ratio does not imply a lack of progress in the labor market. The adjusted E/P rate corroborates the basic picture from the unemployment rate that the labor market has been recovering over the past few years, but that it still has a ways to go to reach a full recovery.
By Pat Higgins, senior economist in the Atlanta Fed's research department
Trying to interpret changes in labor utilization measures such as the employment-to-population ratio is complicated by the fact that they do not refer to the same set of people over time. The age composition of the population is changing, and behavior can vary across and within age cohorts.
This issue is illustrated in a recent New York Fed study of the employment-to-population ratio by Samuel Kapon and Joseph Tracy. This ratio nosedived during the recent recession by about 4 percentage points and has barely budged since.
This measure of labor utilization is the clear laggard on any labor market recovery dashboard. But the authors show that it is not so clear that the employment-to-population ratio is really so far from where it should be, once you control for the fact the employment rates tend to be lower for younger and older people and that the age composition within the population has shifted over time. This idea is similar to the one used to estimate the trend labor force participation rate in this Chicago Fed study by Daniel Aaronson, Jonathan Davis, and Luojia Hu. The issue of controlling for dominant demographic trends is one of the reasons we at the Atlanta Fed decided not to feature either the overall employment-to-population ratio or the overall labor force participation rate in our Labor Market Spider Chart.
A simple, and admittedly crude, alternative to computing the demographically adjusted employment-to-population ratio trend is to look at a segment of the population that is on a relatively flat part of the employment (or participation) rate curve. A common standard for this is the so-called prime-aged population (people aged 25 to 54). These individuals are less likely to be making retirement decisions than older individuals and are less likely to be making schooling decisions than younger people. Of course, this approach doesn't control for within-cohort factors like educational differences.
So what do we find? The prime-aged employment-to-population ratio declined almost 5 percentage points between the end of 2007 and 2009 (versus 4 percentage points overall) and since then has recovered about 25 percent of that decline. Using the end of 2007 as reference, the Kapon and Tracy trend estimate has declined about 1.7 percentage points, which implies the overall employment-to-population ratio, by not continuing to decline, has improved by about 40 percent.
Then what does the analysis say about labor utilization in the wake of the recession? Once demographic factors are controlled for, both aforementioned measures indicate that labor-resource utilization has improved relative to trend. In fact, as Kapon and Tracy note, the relative improvement would be even greater if you believed that employment was above trend before the recession.
By John Robertson, a vice president and senior economist in the Atlanta Fed's research department
Last month, we at the Atlanta Fed had the great pleasure of hosting Sergio Rebelo for a couple of days. While he was here, we asked Sergio to share his thoughts on a wide range of current economic topics. Here is a snippet of a Q&A we had with him about the state of the euro-area economy:
Sergio, what would you say was the genesis of the problems the euro area has faced in recent years?
The contours of the euro area’s problems are fairly well known. The advent of the euro gave peripheral countries—Ireland, Spain, Portugal, and Greece—the ability to borrow at rates that were similar to Germany's. This convergence of borrowing costs was encouraged through regulation that allowed banks to treat all euro-area sovereign bonds as risk free.
The capital inflows into the peripheral countries were not, for the most part, directed to the tradable sector. Instead, they financed increases in private consumption, large housing booms in Ireland and Spain, and increases in government spending in Greece and Portugal. The credit-driven economic boom led to a rise in labor costs and a loss of competitiveness in the tradable sector.
Was there a connection between the financial crisis in the United States and the sovereign debt crisis in the euro area?
Simply put, after Lehman Brothers went bankrupt, we had a sudden stop of capital flows into the periphery, similar to that experienced in the past by many Latin American countries. The periphery boom quickly turned into a bust.
What do you see as the role for euro area monetary policy in that context?
It seems clear that more expansionary monetary policy would have been helpful. First, it would have reduced real labor costs in the peripheral countries. In those countries, the presence of high unemployment rates moderates nominal wage increases, so higher inflation would have reduced real wages. Second, inflation would have reduced the real value of the debts of governments, banks, households, and firms. There might have been some loss of credibility on the part of the ECB [European Central Bank], resulting in a small inflation premium on euro bonds for some time. But this potential cost would have been worth paying in return for the benefits.
And did this happen?
In my view, the ECB did not follow a sufficiently expansionary monetary policy. In fact, the euro-area inflation rate has been consistently below 2 percent and the euro is relatively strong when compared to a purchasing-power-parity benchmark. The euro area turned to contractionary fiscal policy as a panacea. There are good theoretical reasons to believe that—when the interest rate remains constant that so the central bank does not cushion the fall in government spending—the multiplier effect of government spending cuts can be very large. See, for example, Gauti Eggertsson and Michael Woodford, “The Zero Interest-rate Bound and Optimal Monetary Policy,” and Lawrence Christiano, Martin Eichenbaum, and Sergio Rebelo, "When Is the Government Spending Multiplier Large?”
Theory aside, the results of the austerity policies implemented in the euro area are clear. All of the countries that underwent this treatment are now much less solvent than in the beginning of the adjustment programs managed by the European Commission, the International Monetary Fund, and the ECB.
Bank stress testing has become a cornerstone of macroprudential financial oversight. Do you think they helped stabilize the situation in the euro area during the height of the crisis in 2010 and 2011?
No. Quite the opposite. I think the euro-area problems were compounded by the weak stress tests conducted by the European Banking Association in 2011. Almost no banks failed, and almost no capital was raised. Banks largely increased their capital-to-asset ratios by reducing assets, which resulted in a credit crunch that added to the woes of the peripheral countries.
But we’re past the worst now, right? Is the outlook for the euro-area economy improving?
After hitting the bottom, a very modest recovery is under way in Europe. But the risk that a Japanese-style malaise will afflict Europe is very real. One useful step on the horizon is the creation of a banking union. This measure could potentially alleviate the severe credit crunch afflicting the periphery countries.
Thanks, Sergio, for this pretty sobering assessment.
By John Robertson, a vice president and senior economist in the Atlanta Fed’s research department
Editor’s note: Sergio Rebelo is the Tokai Bank Distinguished Professor of International Finance at Northwestern University’s Kellogg School of Management. He is a fellow of the Econometric Society, the National Bureau of Economic Research, and the Center for Economic Policy Research.
Despite the addition of only 74,000 jobs to the economy in December, the unemployment rate dropped significantly—from 7 percent to 6.7 percent. The decline came mostly from a decrease in the labor force.
Since the recession began, the labor force participation rate (LFPR) has dropped from 66 percent to 63 percent. Many people have left the labor force because they are discouraged from applying (U.S. Bureau of Labor Statistics data indicate that a little under 1 million people fall into this category). But the primary drivers appear to be an increase in the number of people who are either retired, disabled/ill, or in school.
Certainly, the aging of the population accounts for much of the increase in the retired and disabled/ill categories. Still, there has been a lot of movement over the past few years in the reasons people cite for not participating in the labor force within age groups. Knowing the reasons why people have left (or delayed entering) the labor force can help us understand how much of the decline will likely halt once the economy picks back up and how much is permanent. (For more on this topic, see here, here, and here.)
The chart below shows the distribution of reasons in the fourth quarter of 2013. (Of the people not in the labor force, 1.6 percent indicate they want a job and give a reason for not being in the labor force. They are categorized here as "want a job" only.) Young people are not in the labor force mostly because they are in school. Individuals 25 to 50 years old who are not in the labor force are mostly taking care of their family or house. After age 50, disability or illness becomes the primary reason people do not want to work—until around age 60, when retirement begins to dominate.
How has this distribution changed over the past seven years? For simplicity, I've grouped people by age to show changes over time in the reasons people give for not being in the labor force. However, you can also see an interactive version of the same data without age buckets—and download the data—here.
Of the 12.6 million increase in individuals not in the labor force, about 2.3 million come from people ages 16 to 24, and of that subset, about 1.9 million can be attributed to an increase in school attendance (see the chart below). In particular, young people aged 19 to 24 are more likely to be in school now than before the recession. Among college-age people, those absent from the labor force because they are in school rose from 57 percent to 60 percent. Among people of high school age, the share not in the labor force because they are in school rose from 87 percent to 88 percent.
The number of middle-aged workers not in the labor force rose by 1.8 million (or 11 percent), with four main factors driving the increase.* "Wants a Job" increased 546,000 (34 percent). The "In School" category increased 438,000 (a 38 percent rise). "Disability/Illness" rose 393,000 (an 8 percent rise), and 302,000 more people said they were retired (a 43 percent rise; see the chart below).
Among individuals aged 51 to 60, those not in the labor force increased by 1.6 million (or 16 percent). This increase came almost entirely from the number of people who are disabled or ill, which rose by 1.3 million (a 33 percent increase). Interestingly, the number of retired individuals actually fell by 305,000 between the fourth quarter of 2007 and the fourth quarter of 2010. Since then, the number of retired people within this age group has risen 183,000 but remains 122,000 lower than fourth-quarter 2007 levels. So it seems more people in this age group were delaying retirement instead of leaving early (see the chart below).
About 6.8 million of the 12.6 million increase in those not in the labor force came from the 61-and-over category. An additional 5.3 million (a 17 percent increase) are retired, and 1 million more (a 34 percent increase) are not in the labor force because they are disabled or ill. The other categories were little changed (see the chart below).
In total, the number of people not in the labor force rose by 12.6 million (16 percent) from the fourth quarter of 2007 to the fourth quarter of 2013. About 5.5 million more people (a 16 percent increase) are retired, 2.9 million (a 23 percent increase) are disabled or ill, and 2.5 million (a 19 percent increase) are in school. An additional 161,000 are taking care of their family or house, and an additional 99,000 are not in the labor force for other reasons. The fraction who say they want a job has risen the most (32 percent) but has contributed only 11 percent to the total change. The chart below shows the overall contributions by reason to the changes in labor force participation for all age groups since the onset of the recession.
What further changes can we anticipate? It's hard to say, as many moving parts are at play. Most people currently in school will be approaching the labor market upon graduation. But increased college and graduate school enrollment could augur a permanent shift in the portion of the population who are in school instead of the labor force. We can also expect continued downward pressure on the LFPR from retiring baby boomers as well as boomers who exit the labor force because of disability or illness.
Last, the portion of people who want a job has increased the most since the recession began, and is currently 1.4 million above its prerecession level. People in this category tend to have greater labor force attachment, making them more likely to shift into the labor force. In fact, the number of people in this category has already started to decrease—and is down 709,000 from the fourth quarter 2012.
My Atlanta Fed colleagues Julie Hotchkiss and Fernando Rios-Avila in their 2013 paper "Identifying Factors behind the Decline in the U.S. Labor Force Participation Rate," looked at a range of LFPR projections for 2015–17 based on different labor market assumptions. Depending on the future strength of the U.S. labor market, the projections are highly varying—ranging between a decline of 2.4 percentage points and an increase of 2 percentage points from the 2010–12 average of 64.1 percent. So far, more factors are pulling down the LFPR than pushing it up; the latest reading for December 2013 is already 1.3 percentage points below the 2010–12 average. At that pace, the Hotchkiss et al. lower-bound estimate will be reached before the end of 2014, unless the dynamics change as the economy further improves.
By Ellyn Terry, an economic policy analysis specialist in the research department of the Atlanta Fed
* I've chosen to break the "middle-age" grouping at age 50 instead of 54 because the probability of retiring has changed in different ways over the past few years for the 25- to 50-year-old group and the 51- to 60-year-old group. See the chart mentioned earlier for more detail.
The December meeting of the Federal Open Market Committee (FOMC), as summarized in the minutes published last week, debated the context for tapering the quantitative easing (QE) program of asset purchases and adjusting the FOMC’s forward guidance on the federal funds rate. One of the issues debated was postrecession progress in the labor market. For example, participants struggled with the reasons for the large drop in labor force participation in recent years:
Some participants cited research that found that demographic and other structural factors, particularly rising retirements by older workers, accounted for much of the recent decline in participation. However, several others continued to see important elements of cyclical weakness in the low labor force participation rate and cited other indicators of considerable slack in the labor market, including the still-high levels of long-duration unemployment and of workers employed part time for economic reasons and the still-depressed ratio of employment to population for workers ages 25 to 54. In addition, although a couple of participants had heard reports of labor shortages, particularly for workers with specialized skills, most measures of wages had not accelerated. A few participants noted the risk that the persistent weakness in labor force participation and low rates of productivity growth might indicate lasting structural economic damage from the financial crisis and ensuing recession.
In a speech on Monday, Atlanta Fed President Dennis Lockhart emphasized similar concerns. He posed the question of whether the improvement in the unemployment rate since the end of the recession, now having recovered about 65 percent of its 2007–09 increase, is overstating the actual progress in the utilization of the nation’s labor resources. President Lockhart observes:
But the unemployment rate is influenced by labor force participation, and there has been a sizable decline in the share of the population in the labor force since 2009. This explains how you could get a big drop in the unemployment rate with anemic job gains, as occurred in December.
The labor force participation rate has fallen from 65.8 percent of the population at the end of 2008 to 62.8 percent in December 2013. On this, President Lockhart notes:
Some of the decline in labor force participation since 2009 is due to the baby boomers retiring, but even among prime-age workers—those aged 25 to 54—the participation rate is down significantly [2.1 percentage points]. This suggests that other factors, such as low prospects of finding a job, are playing a role.
To examine this possibility, we can look at the sum of marginally attached workers. These are people who say they are willing to work and have looked for work recently but are not currently looking.
The marginally attached are not counted in the official labor force statistic. During the recession, the number of marginally attached swelled (from around 1.4 million at the end of 2007 to 2.4 million at the end of 2009). Since the end of 2009, the marginally attached rate (as a share of the labor force including marginally attached) has retraced only 12 percent of the recessionary increase. From this, President Lockhart concludes:
It’s accurate to say the country has a large number of people in the so-called “shadow labor force.”
Because the sharp decline in labor force participation is not fully understood, and because the unemployment rate decline conflates declines in participation with employment gains, President Lockhart suggests it is useful to also look at the share of the prime-age population that is employed. Between the end of 2007 and 2009 the employment-to-population rate for this group declined from 79.7 to 74.8 percent. Since 2009, employment gains for the core of the workforce have advanced only 27 percent toward the prerecession peak (for the entire population over age 16, the recovery is essentially zero). Variations on this theme can be seen here and here.
Usually, the employment to population rate and the unemployment rate move in lock step (because labor force movements are very gradual). But that has not been the case during this recovery.
In addition to unemployment, President Lockhart highlights the issue of underemployment:
Many Americans are working fewer hours than they would prefer because their employers are offering them only part-time work. The share of workers who are involuntarily working part-time doubled during the recession and has moved only about 30 percent lower since the recovery began.
So, on the question of whether the unemployment rate decline has overstated actual progress in labor utilization, Lockhart says yes:
To sum up, these comparisons of employment data suggest that the labor market is not as healthy as the improved unemployment rate might suggest. The unemployment rate drop may overstate progress achieved.
The Atlanta Fed has been featuring the labor market spider chart tool on its website as a way to track relative progress in a number of labor market indicators since the end of the recession. For the purposes of President Lockhart’s speech, the relative improvement in various indicators of the rate of labor utilization was presented graphically in the form of yardage gains from the goal-line of a football field. The changes can be seen here (the data are from the U.S. Bureau of Labor Statistics and Atlanta Fed calculations). The idea is that the labor utilization “team” was driven back to its own goal line from the end of 2007 through the end of 2009, and the graphic shows how many yards (percent) the team has recovered as of the January 10 labor report. (The use of a football field image is perhaps appropriate, given that the recent BCS championship game featured two teams from the Sixth District.)
President Lockhart also suggests a link between labor market slack and the weak pricing trends we have experienced in recent years:
It’s worth noting that wage and salary income growth remains weak. I hear very little from business contacts about upward wage pressures except in a few specialized job categories. Wage pressures usually accompany growing demand and rising inflation but, although demand appears to be growing, inflation is very soft.
In fact, looking at the recent disinflation apparent in virtually all consumer price statistics relative to the FOMC’s longer-run objective, President Lockhart acknowledges the risk of an inflation “safety”:
...I think inflation will stabilize and begin to move back in the direction of the FOMC’s 2 percent objective as the economy gathers momentum. So I’m interpreting the soft inflation numbers as a risk signal. Through the lens of prices, the economy could be weaker than we currently believe.
By John Robertson, a vice president and senior economist in the Atlanta Fed’s research department
Roughly a year ago, the Federal Open Market committee (FOMC) switched from date-based forward guidance on the federal funds rate path to guidance based on economic conditionality. The idea, as Chairman Bernanke put it in his post-FOMC press conference, is that "[b]y tying future monetary policy more explicitly to economic conditions, this formulation of our policy guidance should also make monetary policy more transparent and predictable to the public."
Now, on the one hand, you can't be any more clear than to say that the policy interest rate will remain near zero until such-and-such a date. But if you really want to know the "reaction function" that guides monetary policy decisions, date-based guidance isn't going to speak very clearly to this question. Rather, you would probably rather know the economic conditions that would warrant the FOMC's decision to adjust the policy rate.
Let me suggest that clear communication is one of the foundations of good monetary policy because it's one of the foundational characteristics of good money.
A textbook description of money is usually just a recitation of its functions—it acts as a store of value, a medium of exchange, and a unit of account. This definition of money is a rather hollow one (as Minneapolis Fed President Narayana Kocherlakota noted back in his academic days) because it tells us only what money does but doesn't speak to the core issue—what is the problem that money solves?
The "unit of account" function, in particular, gets little development in the textbooks and has generally not carried much weight in the academic literature on the theory of money. (There are a few exceptions, like this NBER working paper by Matthias Doepke and Martin Schneider.) But if people are going to communicate with one another about value, those communications are going to be most effective if done using some standardized metric—and that's where money comes in. As a "unit of account," our money is how we communicate about value. It can be a physical thing, like a particular commodity, or it can be an abstract concept, like the broad purchasing power of a medium of exchange.
But this isn't to imply that all things are equally up to the job of being a good unit of account. Many economists, beginning with Adam Smith, have been critical of commodity-based monetary systems in this regard. In Congressional testimony in 1922 about stabilizing the purchasing power of our money, famed economist Irving Fisher argued that while gold may have been chosen as our money because it was a good medium of exchange, it had proven to be a poor choice as a unit of account on which contracts could be negotiated. Indeed, he argued for a system where the value of money was fixed in terms of a statistical index of its broad purchasing power, a system certainly similar in spirit to the one the Federal Reserve pursues today:
Is it not absurd to have a dollar also a unit in weight, when it is not intended to measure weight, but is intended to measure purchasing power. It is used in commerce in buying and selling, by debtor and creditor for lending and repaying; and we propose that the repayment shall be just. What does that mean? It does not mean that you shall return a given weight of gold or a given weight of anything; it means that you shall return to the lender something that is a just equivalent. Value is involved in there, and value is statistically increased by an index number average purchasing power.
In other words, it's essential that the unit of account conveys value so that the units expressed in trade, contracts, and financial accounts are both meaningful and durable. We recently produced a simple four-minute video on the subject. Give it a view and let us know what you think. We're big on getting our communications right.
By Mike Bryan, vice president and senior economist in the Atlanta Fed's research department
A few posts back my Atlanta Fed colleagues Tim Dunne and Ellie Terry offered up our latest contribution to the ongoing head-scratching over the rather spectacular decline in U.S. labor force participation (LFP) since the onset of the Great Recession in December 2007. “Rather spectacular” in this case means a fall in the participation rate from 66 percent (of the working age population either working or actively seeking work) to the 63 percent level reported for November. In people terms, that 3 percentage point decline represents a reduction of about 1.4 million participants in the U.S. labor market.
Like many other analysts, Dunne and Terry find that the drop in labor force participation appears to come from a combination of demographic factors—mainly the aging of the population—and other causes not specifically identified but generally interpreted to be associated with the weak economy in one way or another.
Two developing stories suggest the LFP may not be leaving the spotlight just yet. The first is this one, from USA Today:
Some 1.3 million Americans are set to lose their unemployment benefits Saturday...
Federal emergency benefits will end when funds run out for a program created during the recession to supplement the benefits that states provide. The cutoff will initially affect 1.3 million people, but 1.9 million more will lose benefits by mid-2014 when their 26 weeks of state paychecks run out, according to the National Employment Law Project.
What will those 1.3 million Americans do when their benefits run dry? According to a recent study by Princeton University’s Henry Farber and the San Francisco Fed’s Robert Valletta—also presented at a conference hosted here at the Atlanta Fed in October—on balance, the affected individuals are likely to leave the labor force:
We examined the impact of the unprecedented extensions of UI [unemployment insurance] benefits in the United States over the past few years on unemployment dynamics and duration and compared their effects with the extension of UI benefits in the milder recession of the early 2000s. We found small but statistically significant reductions in unemployment exits and small increases in unemployment durations arising from both sets of UI extensions. The magnitude of these overall effects is similar across the two episodes...
We find that the effect on exit from unemployment occurs primarily through a reduction in labor force exits rather than through exit to employment (job finding). This is important because it implies that extended benefits do not delay the time to re-employment substantially and so do not have first-order efficiency effects. The major effect of extended benefits is redistributive, providing income to job losers who would have exited the labor force otherwise (consistent with Card et al. 2007). [link mine]
In other words, if a significant decline in unemployment benefits comes to pass, we may well see another bump downward in the labor force participation rate. Although a decline in LFP associated with the expiration of extended UI benefits would fall in Dunne and Terry’s nondemographic category, the Farber and Valletta results suggest that we should interpret any such decline as structural. And structural in this case means not directly amenable to correction by policies aimed at stimulating spending.
The other important piece of recent news, however, is this one, which you probably heard about:
According to the Bureau of Economic Analysis, real gross domestic product—output produced in the United States—actually grew at a rate of 4.1% in the third quarter, up from BEA’s previous estimate of a 3.6% growth rate. The final results are also a gain over the second quarter’s 2.5% GDP growth.
Furthermore, as noted at Calculated Risk, the good news doesn’t stop there:
A little Christmas cheer...
Macroeconomic Advisers...[raised] its estimate for fourth-quarter growth. It now forecasts gross domestic product to expand at an annualized rate of 2.6% in the final three months of the year, up three-tenths of a percentage point from an earlier estimate.
And Goldman Sachs has increased their Q4 GDP tracking to 2.4% annualized growth.
That all adds up to pretty decent growth in the second half of the year. If it persists, and the long-awaited acceleration in the economic expansion finally arrives, better labor market conditions should follow. And if the six-year fall in LFP has in large measure been driven by weak economic conditions, we should at least see a pause in participation declines as economic activity picks up. Actually, we should probably see an outright increase.
The next several quarters, then, may well provide some clarity as to the persistent question of whether or not the large recent exodus of Americans from the labor force has been the result of a lackluster economy. In this period, we may get some clarity as to whether efforts to stem that exodus were justified by a correct diagnosis of the underlying cause.
By Dave Altig, executive vice president and research director at the Atlanta Fed
By pure coincidence, two interviews with Pennsylvania State University professor Neil Wallace have been published in recent weeks. One is in the December issue of the Federal Reserve Bank of Minneapolis’ excellent Region magazine. The other, conducted by Chicago Fed economist Ed Nosal and yours truly, is slated for the journal Macroeconomic Dynamics and is now available as a Federal Reserve Bank of Chicago working paper.
If you have any interest at all in the history of monetary theory over the past 40 years or so, I highly recommend to you these conversations. As Ed and I note of Professor Wallace in our introductory comments, very few people have such a coherent view of their own intellectual history, and fewer still have lived that history in such a remarkably consequential period for their chosen field.
Perhaps my favorite part of our interview was the following, where Professor Wallace reveals how he thinks about teaching economics, and macroeconomics specifically (link added):
If we were to construct an economics curriculum, independent of where we’ve come from, then what would it look like? The first physics I ever saw was in high school... I can vaguely remember something about frictionless inclined planes, and stuff like that. So that is what a first physics course is; it is Newtonian mechanics. So what do we have in economics that is the analogue of Newtonian mechanics? I would say it is the Arrow-Debreu general competitive model. So that might be a starting point. At the undergraduate level, do we ever actually teach that model?
[Interviewers] That means that you would not talk about money in your first course.
That is right. Suppose we taught the Arrow-Debreu model. Then at the end we’d have to say that this model has certain shortcomings. First of all, the equilibrium concept is a little hokey. It’s not a game, which is to say there are no outcomes associated with other than equilibrium choices. And second, where do the prices come from? You’d want to point out that the prices in the Arrow-Debreu model are not the prices you see in the supermarket because there’s no one in the model writing down the prices. That might take you to strategic models of trade. You would also want to point out that there are a lot of serious things in the world that we think we see that aren’t in the model: unemployment, money, and [an interesting notion of] firms aren’t in the Arrow-Debreu model. What else? Investing in innovation, which is critical to growth, isn’t in that model. Neither is asymmetric information. The curriculum, after this grounding in the analogue of Newtonian mechanics, which is the Arrow-Debreu model, would go into these other things. It would talk about departures from that theory to deal with such things; and it would describe unsolved problems.
So that’s a vision of a curriculum. Where would macro be? One way to think about macro is in terms of substantive issues. From that point of view, most of us would say macro is about business cycles and growth. Viewed in terms of the curriculum I outlined, business cycles and growth would be among the areas that are not in the Arrow-Debreu model. You can talk about attempts to shove them in the model, and why they fall short, and what else you can do.
Of the many things that I have learned from Professor Wallace, this one comes back to me again and again: Talk about how to get the things in the model that are essential to dealing with the unsolved problems, honestly assess why they fall short, and explore what else you can do. To me, this is not only a message of good science. It is one of intellectual generosity, the currency of good citizenship.
I was recently asked whether I align with “freshwater” or “saltwater” economics (roughly, I guess, whether I think of myself as an Arrow-Debreu type or a New Keynesian type). There are many similar questions that come up. Are you a policy “hawk” or a policy “dove”? Do you believe in old monetarism (willing to write papers with reduced-form models of money demand) or new monetarism (requiring, for example, some explicit statement about the frictions, or deviations from Arrow-Debreu, that give rise to money’s existence)?
What I appreciate about the Wallace formulation is that it asks us to avoid thinking in these terms. There are problems to solve. The models that we bring to those problems are not true or false. They are all false, and we—in the academic world and in the policy world—are on a common journey to figure out what we are missing and what else we can do.
It is deeply misguided to treat models as if they are immutable truths. All good economists appreciate this intellectually. And yet there is an awful lot of energy wasted, especially in the blogosphere, on casting aspersions at those who are perceived to be seeking answers within other theoretical tribes.
Some problems are well-suited to Newtonian mechanics, some are not. Some amendments to Arrow-Debreu are useful; some are not. And what is well-suited or useful in some circumstances may well be ill-suited or even harmful in others. Perhaps if we all acknowledge that none of us knows which is which 100 percent of the time, we can make just a little more progress on all those unsolved problems in the coming year. At a minimum, we would air our disagreements with a lot more civility.
By Dave Altig, executive vice president and research director at the Atlanta Fed
In an earlier macroblog post, our colleague Julie Hotchkiss examined the decline in labor force participation from the onset of the Great Recession into early 2012, concluding that cyclical factors likely accounted for most of the drop. In this post, we examine how labor force participation has changed since the start of 2012 (and admittedly, we’re much less ambitious in our analysis than Julie). Motivating our analysis, in part, is the observation that much of the recent decline in the labor force participation rate (LFPR) is related to rising retirements (see the November 19 Research Rap by Shigeru Fujita). This is not surprising, as the percentage of individuals aged 65 and older in the population has been increasing sharply over the last half decade. That said, our approach indicates that the LFPR of prime-age workers (ages 25–54) continues to fall, and this is an important source of the overall decline in LFPR in the recent data. Such declines in LFPR in these age categories should be less related to retirement decisions, keeping on the table the possibility that a weak overall labor market remains a key drag on labor force participation.
A straightforward decomposition illustrates that the decline in LFPR among prime-age workers is a major contributor to the overall decline in LFPR. To see this, we separate the change in LFPR into three components: one that measures the change due to shifts in the LFPR within age groups—the within effect; one that measures changes due to population shifts across age groups—the between effect; and one that allows for correlation across the two effects—a covariance term. It works out the covariance term is always very close to zero, so we will omit discussion of that term here. The analysis breaks the data down into five age groups: 16–24, 25–34, 35–44, 45–54, and 55+.
The chart presents the decomposition from Q1 2012 to Q3 2013. Over this period, the overall LFPR declined by half a percentage point, from 63.8 percent to 63.3 percent. The blue areas represent the change due to within-age-group effects, and the green areas represent the change due to between-age-group effects. The sum of the bars is equal to the overall change in labor force participation.
Three key results emerge. First, increases in labor force participation for the youngest age group boosted overall labor force participation by 0.075 percentage points. Second, the growing population share of the 55+ age group reduced LFPRs over the period by 0.21 percentage points, accounting for roughly 40 percent of the overall decline. Third, labor force participation for prime-age workers continued to fall. The combined within effect for the prime-age individuals (25–34, 35–44, and 45–54) reduced the participation rate by 0.28 percentage points—or a little over half of the overall decline in labor force participation. Additional declines in labor force participation were associated with the reduction in population shares of prime age workers.
From an accounting standpoint, the analysis shows that the fall in the LFPR for prime-age workers is a main contributing factor to the recent decline in labor force participation. Indeed, the LFPR of prime-age workers fell from 81.6 to 81.0 from Q1 2012 to Q3 2013, with similar declines for both men and women. Given that prime-age workers make up more than half of the population, it is not surprising that the drop in the LFPR for these age groups accounts for a substantial fraction of the overall decline.
To put this in perspective, we present the same decomposition from Q1 2010 to Q4 2011, where the decline in the LFPR is 0.8 percentage point. While the magnitude of the overall change is different, the decomposition results are quite similar. The decline in participation rates for prime-age workers accounts for a little over 60 percent of the overall decline, with a substantial drag from the rise in the share of older workers (accounting for a third of the drop). In short, the changes in participation due to within and between effects over the first two years look quite similar to that of the second two years of the labor market recovery.
A corollary to this analysis is that these sources of decline in labor force participation have allowed the unemployment rate to decline more sharply than expected, given the moderate employment growth observed. We will not take a stand on whether these are “wrong” or “right” reasons for unemployment rate declines. Rather, we note that the patterns observed early in the recovery are still in place (more or less) in the recent data.
By Timothy Dunne, a research economist and policy adviser,
and Ellie Terry, an economic policy analysis specialist, both in the research department of the Atlanta Fed
The financial system cannot be made completely safe because it exists to allocate funds to inherently risky projects in the real economy. Thus, an important question for policymakers is how best to structure the financial system to absorb these losses while minimizing the risk that financial sector failures will impair the real economy.
Standard theories would predict that one good way of reducing financial sector risk is diversification. For example, the financial system could be structured to facilitate the development of large banks, a point often made by advocates for big banks such as Steve Bartlett. Another, not mutually exclusive, way of enhancing diversification is to create a system that shares risks across banks. An example is the Dodd-Frank Act mandate requiring formerly over-the-counter derivatives transactions to be centrally cleared.
However, do these conclusions based on individual bank stability necessarily imply that risk sharing will make the financial system safer? Is it even relevant to the principal risks facing the financial system? Some of the papers presented at the recent Atlanta Fed conference, "Indices of Riskiness: Management and Regulatory Implications," broadly addressed these questions and others. Other papers discuss the impact of bank distress on local economies, methods of predicting bank failure, and various aspects of incentive compensation paid to bankers (which I discuss in a recent Notes from the Vault).
The stability implications of greater risk sharing across banks are explored in "Systemic Risk and Stability in Financial Networks" by Daron Acemoglu, Asuman Ozdaglar, and Alireza Tahbaz-Salehi. They develop a theoretical model of risk sharing in networks of banks. The most relevant comparison they draw is between what they call a "complete financial network" (maximum possible diversification) and a "weakly connected" network in which there is substantial risk sharing between pairs of banks but very little risk sharing outside the individual pairs. Consistent with the standard view of diversification, the complete networks experience few, if any, failures when individual banks are subject to small shocks, but some pairs of banks do fail in the weakly connected networks. However, at some point the losses become so large that the complete network undergoes a phase transition, spreading the losses in a way that causes the failure of more banks than would have occurred with less risk sharing.
Extrapolating from this paper, one could imagine that risk sharing could induce a false sense of security that would ultimately make a financial system substantially less stable. At first a more interconnected system shrugs off smaller shocks with seemingly no adverse impact. This leads bankers and policymakers to believe that the system can handle even more risk because it has become more stable. However, at some point the increased risk taking leads to losses sufficiently large to trigger a phase transition, and the system proves to be even less stable than it was with weaker interconnections.
While interconnections between financial firms are a theoretically important determinant of contagion, how important are these connections in practice? "Financial Firm Bankruptcy and Contagion," by Jean Helwege and Gaiyan Zhang, analyzes the spillovers from distressed and failing financial firms from 1980 to 2010. Looking at the financial firms that failed, they find that counterparty risk exposure (the interconnections) tend to be small, with no single exposure above $2 billion and the average a mere $53.4 million. They note that these small exposures are consistent with regulations that limit banks' exposure to any single counterparty. They then look at information contagion, in which the disclosure of distress at one financial firm may signal adverse information about the quality of a rival's assets. They find that the effect of these signals is comparable to that found for direct credit exposure.
Helwege and Zhang's results suggest that we should be at least as concerned about separate banks' exposure to an adverse shock that hits all of their assets as we should be about losses that are shared through bank networks. One possible common shock is the likely increase in the level and slope of the term structure as the Federal Reserve begins tapering its asset purchases and starts a process ultimately leading to the normalization of short-term interest rate setting. Although historical data cannot directly address banks' current exposure to such shocks, such data can provide evidence on banks' past exposure. William B. English, Skander J. Van den Heuvel, and Egon Zakrajšek presented evidence on this exposure in the paper "Interest Rate Risk and Bank Equity Valuations." They find a significant decrease in bank stock prices in response to an unexpected increase in the level or slope of the term structure. The response to slope increases (likely the primary effect of tapering) is somewhat attenuated at banks with large maturity gaps. One explanation for this finding is that these banks may partially recover their current losses with gains they will accrue when booking new assets (funded by shorter-term liabilities).
Overall, the papers presented in this part of the conference suggest that more risk sharing among financial institutions is not necessarily always better. Even though it may provide the appearance of increased stability in response to small shocks, the system is becoming less robust to larger shocks. However, it also suggests that shared exposures to a common risk are likely to present at least as an important a threat to financial stability as interconnections among financial firms, especially as the term structure and the overall economy respond to the eventual return to normal monetary policy. Along these lines, I recently offered some thoughts on how to reduce the risk of large widespread losses due to exposures to a common (credit) risk factor.
By Larry Wall, director of the Atlanta Fed's Center for Financial Innovation and Stability
Note: The conference "Indices of Riskiness: Management and Regulatory Implications" was organized by Glenn Harrison (Georgia State University's Center for the Economic Analysis of Risk), Jean-Charles Rochet, (University of Zurich), Markus Sticker, Dirk Tasche (Bank of England, Prudential Regulatory Authority), and Larry Wall (the Atlanta Fed's Center for Financial Innovation and Stability).
In March 2012, the Federal Reserve Bank of Atlanta launched its Jobs Calculator, an application that illustrates the relationship between the unemployment rate, growth in payroll employment, the labor force participation rate, and a few other variables to boot. Most notably, it tells us how many jobs need to be created to achieve a specific unemployment rate within a given period of time. This tool has turned out to be a useful one for anchoring discussions about national employment growth and unemployment among policy makers and the media.
However, the national employment situation masks significant differences in state labor markets. For example, at the trough of the business cycle (June 2009), the national unemployment rate was 9.5 percent, but it ranged from 4.2 percent in North Dakota to 15.2 percent in Michigan. State policy makers, in managing the dynamics of their own employment situation, need to know the data on a state level.
We are pleased to announce that the Atlanta Fed recently unveiled the state-level Jobs Calculator. The same tool that has been used for national discussions is now available for state-level analyses (see the figure below).
Not only does this state tab allow a quick overview of the historical employment growth in each state (see, for example, Alabama's historical employment growth in the figure below), but it also has the same functionality as the national Jobs Calculator. (Because of the recent partial government shutdown, the data are updated only through August; state-level employment data for October will be available November 22.)
Like the national Jobs Calculator, the state-level version allows the user to input a target unemployment rate, choose the number of months desired to hit the target rate, and find out how many new jobs are required per month to get there. But the calculator is flexible enough to allow other interesting experiments as well.
Consider the case of Florida. During the recession, Florida experienced a significant decline in its population growth. It has gone from a high of about 0.2 percent growth per month (roughly 2.4 percent per year) to its current 0.115 percent growth per month (about 1.38 percent per year; see the figure below). Suppose policy makers in Florida want to know how a return to prerecession population growth might affect the number of jobs needed to maintain its current unemployment rate over the next 12 months. (Note that as of August, the unemployment rate in Florida was 7 percent.)
The calculator's default settings always answer the question, “How many jobs per month does it take to maintain today's unemployment rate over the next 12 months?” To answer our hypothetical policy makers' question, all they would have to do is enter a prerecession monthly population growth rate of 0.2 percent into Florida's state Jobs Calculator, leaving everything else the same. Given the current data in hand, we would discover that Florida would need to generate about 6,000 more jobs per month at the higher population growth than at the current—and lower—population growth to stabilize the unemployment rate at 7 percent.
The data behind the state-level Jobs Calculator come from the U.S. Census Bureau's Establishment Survey, the same data used for the national Jobs Calculator, combined with the Local Area Unemployment Statistics (LAUS) programs run by each state. The LAUS contain the regional and state employment statistics that are consistent with data from the Census Bureau's Current Population Survey. State-level population estimates are provided by the U.S. Census Bureau (and are described in more detail here). You'll note that the LAUS data, especially for very small states, look more erratic than national or larger states' numbers—the unfortunate consequence of small sample sizes.
LAUS data are generally issued about the third Friday of each month following the reference month, which means that the state-level Jobs Calculator statistics will be updated about two weeks after the national Jobs Calculator. The schedule of release dates is available from the U.S. Bureau of Labor Statistics.
By Julie Hotchkiss, a research economist and policy adviser in the Atlanta Fed's research department
The fed funds rate has been at the zero lower bound (ZLB) since the end of 2008. To provide a further boost to the economy, the Federal Open Market Committee (FOMC) has embarked on unconventional forms of monetary policy (a mix of forward guidance and large-scale asset purchases). This situation has created a bit of an issue for economic forecasters, who use models that attempt to summarize historical patterns and relationships.
The fed funds rate, which usually varies with economic conditions, has now been stuck at near zero for 20 quarters, damping its historical correlation with economic variables like real gross domestic product (GDP), the unemployment rate, and inflation. As a result, forecasts that stem from these models may not be useful or meaningful even after policy has normalized.
A related issue for forecasters of the ZLB period is how to characterize unconventional monetary policy in a meaningful way inside their models. Attempts to summarize current policy have led some forecasters to create a "virtual" fed funds rate, as originally proposed by Chung et al. and incorporated by us in this macroblog post. This approach uses a conversion factor to translate changes in the Fed's balance sheet into fed funds rate equivalents. However, it admits no role for forward guidance, which is one of the primary tools the FOMC is currently using.
So what's a forecaster to do? Thankfully, Jim Hamilton over at Econbrowser has pointed to a potential patch. However, this solution carries with it a nefarious-sounding moniker—the shadow rate—which calls to mind a treacherous journey deep within the hinterlands of financial economics, fraught with pitfalls and danger.
The shadow rate can be negative at the ZLB; it is estimated using Treasury forward rates out to a 10-year horizon. Fortunately we don't need to take a jaunt into the hinterlands, because the paper's authors, Cynthia Wu and Dora Xia, have made their shadow rate publicly available. In fact, they write that all researchers have to do is "...update their favorite [statistical model] using the shadow rate for the ZLB period."
That's just what we did. We took five of our favorite models (Bayesian vector autoregressions, or BVARs) and spliced in the shadow rate starting in 1Q 2009. The shadow rate is currently hovering around minus 2 percent, suggesting a more accommodative environment than what the effective fed funds rate (stuck around 15 basis points) can deliver. Given the extra policy accommodation, we'd expect to see a bit more growth and a lower unemployment rate when using the shadow rate.
Before showing the average forecasts that come out of our models, we want to point out a few things. First, these are merely statistical forecasts and not the forecast that our boss brings with him to FOMC meetings. Second, there are alternative shadow rates out there. In fact, St. Louis Fed President James Bullard mentioned another one about a year ago based on work by Leo Krippner. At the time, that shadow rate was around minus 5 percent, much further below Wu and Xia's shadow rate (which was around minus 1.2 percent at the end of last year). Considering the disagreement between the two rates, we might want to take these forecasts with a grain of salt.
Caveats aside, we get a somewhat stronger path for real GDP growth and a lower unemployment rate path, consistent with what we'd expect additional stimulus to do. However, our core personal consumption expenditures inflation forecast seems to still be suffering from the dreaded price-puzzle. (We Googled it for you.)
Perhaps more important, the fed funds projections that emerge from this model appear to be much more believable. Rather than calling for an immediate liftoff, as the standard approach does, the average forecast of the shadow rate doesn't turn positive until the second half of 2015. This is similar to the most recent Wall Street Journal poll of economic forecasters, and the September New York Fed survey of primary dealers. The median respondent to that survey expects the first fed funds increase to occur in the third quarter of 2015. The shadow rate forecast has the added benefit of not being at odds with the current threshold-based guidance discussed in today's release of the minutes from the FOMC's October meeting.
Moreover, today's FOMC minutes stated, "modifications to the forward guidance for the federal funds rate could be implemented in the future, either to improve clarity or to add to policy accommodation, perhaps in conjunction with a reduction in the pace of asset purchases as part of a rebalancing of the Committee's tools." In this event, the shadow rate might be a useful scorecard for measuring the total effect of these policy actions.
It seems that if you want to summarize the stance of policy right now, just maybe...the shadow knows.
By Pat Higgins, senior economist, and
Brent Meyer, research economist, both of the Atlanta Fed's research department
The spigot of credit to small businesses appears to be turning faster. As of June 2013, outstanding amounts of small loans on the balance sheets of banks were 4 percent higher than their September 2012 levels, according to the Federal Deposit Insurance Corporation. While they are still 12 percent off 2007 levels, the recent increase is encouraging.
The turnaround in small loan portfolios is not the only sign of improved credit flows to small businesses. The Fed’s October 2013 senior loan officer survey indicates that credit terms to small firms have gradually eased since the second quarter of 2010. Approval ratings of banks and alternative lenders, as measured by Biz2Credit’s lending index, have also risen steadily over the past two years.
In addition to these positive signs, the Atlanta Fed’s third-quarter 2013 Small Business Survey has revealed signs of improvement among small business borrowers in the Southeast. The survey asked recent borrowers about their requests for credit and how successful they were at each place they applied. We also asked, “Over ALL your applications for credit, to what extent were you total financing needs met?” This measure of overall financing satisfaction showed some signs of improvement in the third quarter.
Chart 1 compares the overall financing satisfaction of small business borrowers in the first and third quarter of 2013. The portion of firms that received the full amount requested rose from 28 percent in the first quarter to 42 percent in the third quarter. Meanwhile, the portion that received none of the credit requested declined from 31 percent of the sample in the first quarter to 22 percent in the third quarter.
Further, financing satisfaction rose across a variety of dimensions. Chart 2 shows how average financing satisfaction changed for young firms and mature firms, across industries and by recent sales performance. In all cases, there were increases in the average amount of financing received from the first to the third quarter of 2013.
This broad-based increase in overall financing satisfaction is encouraging. Greater financial health of the applicant pool helped fuel the improvement in borrowing conditions. In the October survey, 52 percent of businesses reported that sales increased while 34 percent reported decreases. Sales have improved significantly from a year ago, when about as many firms reported sales increases as reported decreases. Measures of hiring and capital improvements over the year have also improved for the average firm in the survey (see chart 3).
Lending standards have been improving and small businesses have been slowly gaining momentum, but many obstacles remain. Open-ended questions in our survey revealed that small businesses are still concerned about a number of factors, including the general political and economic uncertainty, the impact of the Affordable Care Act, the higher collateral and personal guarantees required to obtain financing, and regulatory requirements that restrict lending. So while conditions on the ground seem to be improving for small businesses, there still appear to be headwinds that may be holding back a greater pace of improvement.
By Ellie Terry, an economic policy analysis specialist in the Atlanta Fed’s research department
Last Friday, Atlanta Fed President Dennis Lockhart delivered a speech at the University of Mississippi, the bottom line of which was reported by the Wall Street Journal's Michael Derby:
Federal Reserve Bank of Atlanta President Dennis Lockhart said Friday that central bank policy must remain very easy for some time to come, although he cautioned the exact mix of tools employed by the central bank will change over time...
"Monetary policy overall should remain very accommodative for quite some time," Mr. Lockhart said... "The mix of tools we use to provide ongoing monetary stimulus may change, but any changes will not represent a fundamental shift of policy"...
That's a pretty accurate summary, but Derby follows up with commentary that feels somewhat less accurate:
The big question about Fed policy is what the central bank does with its $85 billion-per-month bond-buying program. It had widely been expected to start slowing the pace of purchases starting in September, but when it didn't do that, expectations went into flux. Ahead of the jobs data Friday, many forecasters had gravitated to the view bond buying would be trimmed some time next spring. Now, a number of forecasters said the risk of the Fed slowing its asset buying sooner has risen.
Now, the views that I express here are not necessarily those of the Federal Reserve Bank of Atlanta. But in this case, I think I can fairly claim that what President Lockhart was saying was that the big question is not "what the central bank does with its $85 billion-per-month bond-buying program." The following part of President Lockhart's speech—reiterated today in a speech in Montgomery, Alabama—is worth emphasizing:
The FOMC [Federal Open Market Committee] is currently using two tools to maintain the desired degree of monetary accommodation—the policy interest rate and bond purchases. Importantly, the FOMC has stated that it intends to keep the short-term policy rate low at least until the unemployment rate falls below 6 1/2 percent. This "forward guidance" is meant to convey a sense of how long short-term interest rates will stay near current levels.
There is some confusion about how the Fed's forward guidance and asset purchase program relate to each other. I will give you my view.
In the toolkit the FOMC has at its disposal, there is a sense in which asset purchases and low policy rates are complementary. Asset purchases and forward guidance on interest rates are complements in the sense that they are both designed to put downward pressure on longer-term interest rates....
But there is also a sense in which these tools are substitutes. By substitutes I mean that guidance pointing to a sustained low policy rate and asset purchases are discrete tools that can be deployed independently or in varying combinations. They can be thought of as a particular policy tool mix chosen to fit the circumstances at this particular phase of the recovery.
In other words, there is an important difference between changing the amount of monetary stimulus and changing the tools deployed to provide that stimulus. When the only tool in play is the federal funds rate, equating adjustments in the Fed's policy rate with changes in the stance of monetary policy is, while not completely straightforward, relatively simple. With multiple tools in use, however, gauging the stance of monetary policy requires that the settings of all policy instruments be considered.
Suppose that the FOMC does scale back or end its asset purchases? Can that possibly be consistent with maintaining a constant degree of monetary stimulus? Sure, and one obvious option is to use adjustments to the forward guidance portion of the FOMC's current policy to provide additional stimulus as asset purchases are scaled back. There are pros and cons to that approach, many of which surfaced in the discussion of this paper, by the Federal Reserve Board's Bill English, David Lopez-Salido, and Bob Tetlow, which circulated last week. (See, for example, here, here, and here.)
In any event, a decision to replace asset purchases with some other form of stimulus—be it extending forward guidance or another alternative—would necessarily raise the question: Why bother? One answer might arise from the cost and efficacy considerations that the FOMC has identified as part of the calculus for whether to continue with asset purchases.
Here again, the fact of multiple tools is germane. With the option of different policy mixes, altering the asset purchase program on grounds of cost or efficacy need not mean that the costs of the program are large or the purchases themselves lack effect. It need only mean that the costs might be larger, or the purchases less effective, than providing the same set of stimulus with some alternative set of tools. I give the last word to President Lockhart:
Going forward, it may be appropriate to adjust the policy tool mix. That will depend on circumstances and the economic diagnosis of the moment.
By Dave Altig, executive vice president and research director at the Atlanta Fed
One possibility is that Bernanke and the other FOMC leaders… never intended to start tapering…
A second possible explanation is that Bernanke and other Fed leaders were indeed anticipating that they would begin tapering QE in September but were startled at how rapidly long-term rates had risen in response to their earlier statements…
The third scenario is that economic activity was clearly slowing, with the future pace of activity therefore vulnerable to even higher interest rates.
Speaking only for myself, I choose Feldstein's third option. He goes a good way to making the case himself:
The annualized GDP growth rate in the first half of 2013 was just 1.8%, and final sales were up by only 1.2%. Although there are no official GDP estimates for the third quarter, private-sector assessments anticipate no acceleration in growth, putting the economy on a path that will keep this year's output gain at well under 2%.
That unfortunate story was pretty clear on the eve of the FOMC meeting—in particular, the lack of evidence that growth in the second half of the year would be an improvement on the already disappointing pace of the first half. Our own internal "nowcast" tracking model was suggesting third-quarter GDP growth in the neighborhood of the sub-2 percent growth that Feldstein cites. And as this table shows, things have not improved since:
These facts, of course, were reflected in the downgrade of the 2013 growth forecasts published in FOMC participants' Summary of Economic Projections. But that is not all, as Professor Feldstein reports:
In addition, the Fed's preferred measure of inflation was much lower than its 2% target. The annual price index for personal consumer expenditure, excluding food and energy, has been rising for several months at a rate of just 1.2%, increasing the possibility of a slide into deflation.
And even if you don't go in for inflation measures that exclude food and energy, it doesn't much matter, because all-in inflation was, and still is, also running well below that 2 percent target:
Though the August personal consumption expenditures price report finally provided a slight uptick in year-over-year core inflation, there was not even that scant hint of a return to the 2 percent inflation target by FOMC meeting time.
And that was looking a lot like strike number two to me. As Fed Chairman Ben Bernanke explained at the post-meeting press conference, repeating the criteria for adjustments to the FOMC asset purchase program that he laid out in June:
We have a three-part baseline projection which involves increasing growth…, continuing gains in the labor market, and inflation moving back towards objective… we'll be looking to see if the data confirm that basic outlook.
Of the remaining element of the three-part baseline, it is true that 12-month average monthly job gains looked pretty much like they did in June, when the talk of taper got serious:
But the momentum—which I measure here as the ratio of three-month average monthly job gains to the 12-month average—was clearly in the downward direction:
What's more, the revisions in prior months' employment statistics were running in the wrong direction:
As a rule, forecasters don't sweat being wrong. That comes with the territory. But when you are persistently wrong in the same direction, it is time to worry at least a bit.
So, what do we have, then?
- Inflation is low relative to the FOMC's objective—and has not moved in the direction of that objective with any conviction.
- GDP growth has disappointed, with the anticipated pickup in second-half growth nowhere in sight.
- "Continuing gains in the labor market" at the pace seen earlier in the year are looking a little shaky.
I find it pretty easy to see how this fails to add up to satisfaction of the three-part economic conditionality laid out in June by the Chairman (on behalf of the FOMC).
One could argue, I suppose, that the FOMC's explicit tying of asset purchases to improvement in the labor market makes it first among equals in the three-part test (as long as inflation is relatively stable), that similar downward momentum on the job front arose and disappeared in the summer of 2012, and that with a little patience things will appear on track.
Maybe. But I would point out that the reversal of negative momentum in the labor market the summer before was accompanied by the initiation of "QE3" (or at least the MBS part of QE3). You can draw your own conclusions about causality, but there is a fairly convincing case to be made for the proposition that, with the data in hand at the time, a wait-and-see decision was what patience dictated.
That, of course, begs the main question posed in Feldstein's article: When will it be time to taper? On that, and in the spirit of baseball playoff season, get your scorecard here.
By Dave Altig, executive vice president and research director at the Atlanta Fed
Foresight about the disaster to come was not the primary reason this year’s Nobel Prize in economics went to Robert Shiller (jointly with Eugene Fama and Lars Hansen). But Professor Shiller’s early claim that a housing-price bubble was full on, and his prediction that trouble was a-comin’, is arguably the primary source of his claim to fame in the public sphere.
Several years down the road, the causes and effects of the housing-price run-up, collapse, and ensuing financial crisis are still under the microscope. Consider, for example, this opinion by Dean Baker, co-director of the Center for Economic and Policy Research:
...the downturn is not primarily a “financial crisis.” The story of the downturn is a simple story of a collapsed housing bubble. The $8 trillion housing bubble was driving demand in the U.S. economy in the last decade until it collapsed in 2007. When the bubble burst we lost more than 4 percentage points of GDP worth of demand due to a plunge in residential construction. We lost roughly the same amount of demand due to a falloff in consumption associated with the disappearance of $8 trillion in housing wealth.
The collapse of the bubble created a hole in annual demand equal to 8 percent of GDP, which would be $1.3 trillion in today’s economy. The central problem facing the U.S., the euro zone, and the U.K. was finding ways to fill this hole.
In part, Baker’s post relates to an ongoing pundit catfight, which Baker himself concedes is fairly uninteresting. As he says, “What matters is the underlying issues of economic policy.” Agreed, and in that light I am skeptical about dismissing the centrality of the financial crisis to the story of the downturn and, perhaps more important, to the tepid recovery that has followed.
Interpreting what Baker has in mind is important, so let me start there. I have not scoured Baker’s writings for pithy hyperlinks, but I assume that his statement cited above does not deny that the immediate post-Lehman period is best characterized as a period of panic leading to severe stress in financial markets. What I read is his assertion that the basic problem—perhaps outside the crisis period in late 2008—is a rather plain-vanilla drop in wealth that has dramatically suppressed consumer demand, and with it economic growth. An assertion that the decline in wealth is what led us into the recession, is what accounts for the depth and duration of the recession, and is what’s responsible for the shallow recovery since.
With respect to the pace of recovery, evidence supports the proposition that financial crises without housing busts are not so unique—or if they are, the data tend to associate financial-related downturns with stronger-than-average recoveries. Mike Bordo and Joe Haubrich, respectively from Rutgers University and the Federal Reserve Bank of Cleveland, argue that the historical record of U.S. recessions leads us to view housing and the pace of residential investment as the key to whether tepid recoveries will follow sharp recessions:
Our analysis of the data shows that steep expansions tend to follow deep contractions, though this depends heavily on when the recovery is measured. In contrast to much conventional wisdom, the stylized fact that deep contractions breed strong recoveries is particularly true when there is a financial crisis. In fact, on average, it is cycles without a financial crisis that show the weakest relation between contraction depth and recovery strength. For many configurations, the evidence for a robust bounce-back is stronger for cycles with financial crises than those without...
Our results also suggest that a sizeable fraction of the shortfall of the present recovery from the average experience of recoveries after deep recessions is due to the collapse of residential investment.
From here, however, it gets trickier to reach conclusions about why changes in housing values are so important.
Simply put, why should there be a “wealth effect” at all? If the price of my house falls and I suffer a capital loss, I do in fact feel less wealthy. But all potential buyers of my house just gained the opportunity to obtain my house at a lower price. For them, the implied wealth gain is the same as my loss. If buyers and sellers essentially behave the same way, why should there be a large impact on consumption? *
I think this notion quickly leads you to the thought there is something fundamentally special about housing assets and that this special role relates to credit markets and finance. This angle is clearly articulated in these passages from a Bloomberg piece earlier in the year, one of a spate of articles in the spring about why rapidly recovering house prices were apparently not driving the recovery into a higher gear:
The wealth effect from rising house prices may not be as effective as it once was in spurring the U.S. economy...
The wealth effect “is much smaller,” said Amir Sufi, professor of finance at the University of Chicago Booth School of Business. Sufi, who participated in last year’s central-bank conference at Jackson Hole, Wyoming, reckons that each dollar increase in housing wealth may yield as little as an extra cent in spending. That compares with a 3-to-5-cent estimate by economists prior to the recession.
Many homeowners are finding they can’t refinance their mortgages because banks have tightened credit conditions so much they’re not eligible for new loans. Most who can refinance are opting not to withdraw equity after the first nationwide decline in house prices since the Great Depression reminded them home values can fall as well as rise...
Others are finding it difficult to refinance because credit has become a lot harder to come by. And that situation could worsen as banks respond to stepped-up government oversight.
“Credit is going to get tighter before it gets easier,” said David Stevens, president and chief executive officer of the Washington-based Mortgage Bankers Association...
“Households that have been through foreclosure or have underwater mortgages or are otherwise credit-constrained are less able than other households to take advantage” of low interest rates, Fed Governor Sarah Bloom Raskin said in an April 18 speech in New York.
(I should note that Sufi et al. previously delved into the relationship between household balance sheets and the economic downturn here.)
A more systematic take comes from the Federal Reserve Board’s Matteo Iacoviello:
Empirically, housing wealth and consumption tend to move together: this could happen because some third factor moves both variables, or because there is a more direct effect going from one variable to the other. Studies based on time-series data, on panel data and on more detailed, recent micro data point suggest that a considerable portion of the effect of housing wealth on consumption reflects the influence of changes in housing wealth on borrowing against such wealth.
That sounds like a financial problem to me and, in the spirit of Baker’s plea that it is the policy that matters, this distinction is more than semantic. The policy implications of an economic shock that alters the capacity to engage in borrowing and lending are not necessarily the same as those that result from a straightforward decline in wealth.
Having said that, it is not so clear how the policy implications are different. One possibility is that diminished access to credit markets also weakens policy-transmission mechanisms, calling for even more aggressive demand-oriented “pump-priming” policies of the sort Dean Baker advocates. But it is also possible that we have entered a period of deep structural repair that only time (and not merely government stimulus) can (or should) engineer: deleveraging and balance sheet repair, sectoral resource reallocation, new consumption habits, new business models driven by both market and regulatory imperatives, you name it.
In my view, it’s not yet clear which policy approach is closest to optimal. But I am fairly well convinced that good judgment will require us to think of the past decade as the financial event it was, and in many ways still is.
*Update: A colleague pointed out that my example describing housing price changes and wealth effects may be simplified to the point of being misleading. Implicitly, I am in fact assuming that the flow of housing services derived from housing assets is fixed, a condition that obviously would not hold in general. See section 3 of the Iacoviello paper cited above for a theoretical description of why, to a first approximation, we would not expect there to be a large consumption effect from changes in housing values.
By Dave Altig, executive vice president and research director at the Atlanta Fed
Though never far from the headlines, the Federal Reserve's dual mandate comes front and center again with the announcement today of President Obama's nomination of Fed Vice Chair Janet Yellen as the next chair of the Board of Governors. Inevitably, analysis will turn to discussions of who is a hawk and who is a dove, who cares relatively more about inflation, and who cares relatively more about growth and employment.
That's unfortunate, because such characterizations really do miss the point. The debate among different policymakers is not about whether person A is more concerned about jobs and unemployment than person B, but about legitimate and longstanding conversations about what accounts for the performance of labor markets and what role monetary policy might have in the event that performance is judged to be subpar.
As it happens, the Atlanta Fed's most recent contribution to this discussion came last week in the form of the annual employment conference sponsored by the Bank's Center for Human Capital Studies. Organized, as in past years, by Richard Rogerson (Princeton University), Robert Shimer (University of Chicago), and Melinda Pitts (Federal Reserve Bank of Atlanta), the conference explored the causes of the continued weak labor market recovery in the United States. The existing literature has suggested a number of possibilities: wage rigidities, mismatch between workers' skills and the skills required by new jobs, extended unemployment insurance benefits and other government policy changes, and firms' reorganizing and asking workers to do more. The papers sought to analyze and document the importance of these factors for the slow recovery.
One notable policy change in the recent recession was the unprecedented expansion of unemployment insurance (UI) benefits to as long as 99 weeks for a very large fraction of UI-eligible workers. Did this increase play an important role in high levels of unemployment? Two papers from the conference addressed this question from different perspectives. "Do Extended Unemployment Benefits Lengthen Unemployment Spells? Evidence from Recent Cycles in the U.S. Labor Market," by Henry S. Farber and Robert G. Valetta, assessed the extent to which extended UI benefits result in higher unemployment because workers choose to remain unemployed longer. They find a statistically significant effect of longer UI durations on the duration of unemployment spells, but they conclude that the overall contribution to the unemployment rate was less than half a percentage point. Because the aggregate unemployment rate rose by more than 5 percent, this effect accounts for less than 10 percent of the overall increase.
"Unemployment Benefits and Unemployment in the Great Recession: The Role of Macro Effects," by Marcus Hagedorn, Fatih Karahan, Iourii Manovskii, and Kurt Mitman, offered a different perspective. The authors look at the evolution of unemployment rates in counties that are adjacent but lie in different states. They use the fact that the timing of extended benefits occurs at different times across states to identify the effect of extended UI durations on country-level unemployment. They find that the effects are sufficiently large that the increase in UI duration can account for virtually all of the increase in unemployment.
While seemingly at odds, the results of these two studies are consistent. The first paper shows that the decrease in the job-finding rate for workers with relatively longer benefits did not increase that much compared with the rate for workers with shorter-duration benefits, holding the overall unemployment rate constant. The second paper argues that the job-finding rate decreases for everyone when benefits are extended. The authors find that when some workers have access to longer-duration UI benefits, being unemployed is not as painful for them, which puts upward pressure on wages. To the extent that firms cannot target their job openings toward workers without access to UI, firms may be less likely to create jobs, making it harder for all workers to get job offers. The impact on uninsured workers may be as large as the impact on insured workers, and so the microeconomic estimates in Farber and Valetta will not necessarily uncover UI's total impact on the unemployment rate.
The possible role of wage rigidities has figured prominently in many accounts of the large increase in unemployment during the recent recession. Two papers considered the importance of this explanation. "Wage Adjustment in the Great Recession," by Michael Elsby, Donggyun Shin and Gary Solon, used microdata from the U.S. Census Bureau's Current Population Survey to examine the extent to which wages are sticky. The paper finds that there has been less response in average real wages during the recent recession than in previous recessions, perhaps suggesting that real wage rigidity contributed to the large increase in unemployment. However, they also show that wages at the individual level are really quite flexible. Specifically, relatively few individuals have zero nominal wage growth from one year to the next, and many people experience decreases in nominal wage rates.
A key issue in the theoretical literature is the extent to which wage stickiness affects new hires versus existing workers. In "How Sticky Wages in Existing Jobs Can Affect Hiring," authors Mark Bils, Yongsung Chang and Sun-Bin Kim show that even if wages for new hires are completely flexible, they may nonetheless have large effects on unemployment fluctuations when one allows for an "effort decision" for existing workers. This decision means that in response to negative shocks, firms require existing workers to expend more effort given that their wage is fixed, decreasing the need to hire new workers. The authors show that this effect is quantitatively significant and can come close to resolving the unemployment volatility puzzle, which relates to the large fluctuations in unemployment relative to productivity.
An empirical regularity that has appeared in the last few years is an outward shift in the Beveridge curve, which relates the unemployment rate to the level of vacancies. One interpretation of this upward shift is that the matching of unemployed workers and vacancies has worsened. Yet there is a lot of variety in the job-search effort by workers with different characteristics, such as the length of unemployment, whether they are on temporary layoff, and so on. In "Measuring Matching Efficiency with Heterogeneous Jobseekers," Robert Hall and Sam Schulhofer-Wohl devise a method for incorporating this heterogeneity into the analysis and show that there has indeed been a decrease in the matching rate for workers during the last few years. It will be important for future research to determine how much this decrease reflects a decline in search intensity or whether the lower job-finding rates represent a decrease for a given level of search intensity.
Related to the two issues of nominal rigidities and mismatch, in the paper "Labor Mobility within Currency Unions," Emmanuel Farhi and Ivan Werning study the role of labor mobility in diminishing the effects associated with nominal rigidities. For example, some researchers have suggested that a key difference between the apparent success of the United States relative to the euro zone is U.S. labor is more mobile. Farhi and Werning argue that one should not assume the mobility necessarily reduces the effects of nominal rigidities. In particular, they conclude that mobility eases the effects of nominal rigidities only if goods markets are well integrated.
Two papers focused on the nature of worker mobility across firms in the recent recession. In "Worker Flows over the Business Cycle: The Role of Firm Quality," Lisa Kahn and Erika McEntarfer examine recent changes in flows of workers between firms that offer jobs of differing quality. They find that that lower-quality firms decreased both hiring and separations by large and equal amounts, whereas high-quality firms have much smaller declines in both hiring and separations. The net result is that the fraction of workers in lower-quality jobs tends to increase during recessions.
In closely related work, "Did the Job Ladder Fail after the Great Recession?" by Giuseppi Moscarini and Fabien Postel-Vinay, uses data from the U.S. Bureau of Labor Statistics' Job Openings and Labor Turnover Survey (JOLTS) to study the hiring and separation patterns across firms of different sizes. They determine that the pattern of firm growth across size classes was different during this recession than in previous recessions. In particular, they find that following the Lehman Brothers collapse, smaller firms actually fared worse than larger firms, perhaps because financing constraints had more severe consequences for smaller firms.
As the provisions in the Affordable Care Act (ACA) take effect in the coming months, there may be large effects not only on the market for health care but also on the labor market. In particular, the ACA will implicitly introduce taxes and subsidies that will differ across firms and workers of different types. In "Effects of the Affordable Care Act on the Amount and Composition of Labor Market Activity," Trevor Gallen and Casey Mulligan develop a framework to think about how these provisions will influence labor market outcomes across different sectors and worker types, and they use a calibrated version of the model to quantify the effects. The authors predict that the ACA will substantially reduce the return to market work for low-skilled individuals and that a large number of individuals who currently receive health insurance through their employers will end up purchasing insurance through the exchanges established as part of the ACA.
The conference also featured a presentation by Ed Lazear, "The New Normal? Productivity and Employment during the Recession and Recovery." The talk highlighted three themes from Lazear's recent research. First, productivity did not decline in the recent recession—as it typically had done in previous recessions—perhaps reflecting that workers expend more effort during periods of high unemployment since they fear unemployment more in a weak labor market. Second, the unemployment rate is a less useful indicator of the overall state of the labor market during the current recovery (in recent years the decline in the unemployment rate has not been accompanied by an increase in the employment-to-population ratio, since labor force participation has declined). The third theme is that the deterioration in labor market outcomes during the recent recession should be interpreted as cyclical rather than structural and, hence, a labor market recovery is likely once GDP growth is stronger.
We certainly wouldn't claim that the conference put to rest any of the relevant questions that will confront the Federal Open Market Committee and its new chair going forward. But we do believe that continuing to support the dissemination of the type of research presented at this conference gives us a fighting chance.
By Richard Rogerson of Princeton University and Robert Shimer of the University of Chicago, both advisers to the Atlanta Fed's Center for Human Capital Studies, and Melinda Pitts, director of the Atlanta Fed's Center for Human Capital Studies