The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.
- BLS Handbook of Methods
- Bureau of Economic Analysis
- Bureau of Labor Statistics
- Congressional Budget Office
- Economic Data - FRED® II, St. Louis Fed
- Office of Management and Budget
- Statistics: Releases and Historical Data, Board of Governors
- U.S. Census Bureau Economic Programs
- White House Economic Statistics Briefing Room
July 06, 2016
When It Rains, It Pours
Seasonally adjusted nonfarm payroll employment increased by only 38,000 jobs in May, according to the initial reading by the U.S. Bureau of Labor Statistics (BLS), and the total increase for the prior two months was revised down by a cumulative 59,000. Although the May increase was depressed by 35,100 striking workers at Verizon Communications, observers widely anticipated this distortion (the strike started April 13). Nonetheless, the median forecast of the May payroll gain from a Bloomberg survey of economists was 160,000, still well above the official estimate. The disappointing employment gain in May, I believe, is statistically related to the downward revisions to the seasonally adjusted gains made over the prior two months.
In contrast to the revision to the seasonally adjusted data, the nonseasonally adjusted level of payroll employment in April was only revised down by 3,000 in the May report. So most of the downward revision to the seasonally adjusted March and April employment gains was the result of revised seasonal factors (the difference between 59,000 and 3,000). In the chart below, the green diamond (toward the left) is the downward revision of 56,000 that resulted from the revised seasonal factors plotted against the Bloomberg survey forecast error for May (the difference between the actual estimate of 38,000 and the forecast of 160,000). The other diamonds represent corresponding points for reports from January 2006 through April 2016. The data points indicate a clear positive relationship and—based on the May Bloomberg forecast error—a simple linear regression would have almost exactly predicted the total downward revision to the March and April employment gains coming from revised seasonal factors.
To gain some insight into the positive relationship in the above chart, I used a model to seasonally adjust the last 10 years of nonfarm payroll employment data (excluding decennial census workers). Note that although I followed the BLS's procedure of accounting for whether there are four or five weeks between consecutive payroll surveys, I did not seasonally adjust the detailed industry employment data and sum them up, as the BLS does.
According to my seasonal adjustment model, the seasonally adjusted April employment level using data from the May employment report is 60,000 below the seasonally adjusted April employment level estimated with data from the April report. My seasonal adjustment model only using data through April from the May report predicts a nonseasonally adjusted increase of 789,000 jobs in May instead of the BLS's estimated increase of 651,000 jobs. The difference between these two estimates is similar to the Bloomberg survey forecast error noted above.
Further, when I replace the BLS's nonseasonally adjusted estimate for May with the model's forecast, the estimate of seasonally adjusted April employment is only 2,000 less than the model estimated with data from the April employment report. Hence, almost all of the model's downward revision to seasonally adjusted April employment appears to be the result of adding fewer jobs in May than the model expected.
The above analysis illustrates that, when it comes to looking at seasonally adjusted employment data, the number of jobs next month will affect the estimate of the number of jobs this month. This is not a very appealing notion, but when using seasonally adjusted data, it comes with the territory. Fortunately, analyzing the nonseasonally adjusted data allows us to gauge the impact of a surprise in the current estimate of seasonally adjusted employment growth on revisions to the prior two months. So when the June report is released on Friday, we will be paying close attention to both the seasonally adjusted headline numbers as well as the revisions to the nonadjusted data.
June 29, 2016
Pay As You Go: Yes or No?
The Atlanta Fed's 2015 Annual Report focused on the graying of the U.S. economy. Part of the report and a follow-up webcast discussed how aging is driving the depletion of the U.S. Social Security and Medicare trust funds.
Based on current projections from the Congressional Budget Office, the Social Security trust fund is forecast to run dry around 2030 (see the chart); the Medicare trust fund in 2025. Barring a change in legislation, once the trust funds are depleted, benefits will be cut so that outlays match revenues. In the case of Social Security, this reduction will mean a 31 percent decline in benefits. To balance the Medicare budget, certain Medicare benefits will also face significant reduction.
As my coauthors and I explain in a recent Oxford University Press blog post, our research has found that pay-as-you-go programs for retirees such as Social Security and Medicare, on average, make people worse off, whereas means-tested social insurance programs for retirees, such as Medicaid and Supplemental Security Income (SSI), improve welfare.
These findings are based on comparing the welfare of individuals born into economies with different types of social insurance programs available. We find that, given the hypothetical choice between having or not having Social Security, the average individual would choose to be born into an economy without Social Security. However, when we ask if an average individual would prefer to be born into an economy with or without means-tested retiree programs, we find that he or she would strongly prefer the economy with these programs.
The preference for an economy without universal pay-as-you-go programs like Social Security is consistent with findings in the literature more generally. These programs are large (Social Security was 4.9 percent of U.S. gross domestic product [GDP] in 2013) and have distortionary effects. In standard economic models, the distortions lead to such large reductions in savings and labor supply that they tend to outweigh the programs' insurance benefits.
In contrast, means-tested social insurance programs for retirees, such as Medicaid and SSI, are much smaller. Together, outlays from these programs for the elderly were only 1 percent of GDP in 2013. These programs provide transfers only to individuals with limited income and assets or with impoverishing medical expenses. However, it is in these states of world, when one is poor and/or sick, that such transfers are most valuable, which is why we find that these programs improve welfare.
Researchers have found that means-tested transfer programs for working-age individuals are highly distortionary because they implicitly tax income and assets. However, we find that such distortions are less severe for means-tested transfer programs for retirees, since individuals cannot use these programs to finance working-age consumption and medical care.
Our findings suggest that one potential solution to the sustainability problems plaguing Social Security and Medicare may be to make these programs means-tested as well. Under such a system, the government would still provide protection against the risks of ending up old, sick, alone or poor, but with programs that are significantly less costly.
Of course, saying that individuals would prefer to be born into a hypothetical economy A instead of economy B is not the same thing as saying that current U.S. citizens want to make such a transition. Moving from the current system to one in which Social Security and Medicare benefits are means-tested would not be attractive to wealthier individuals who are already retired or on the verge of it. A compensation scheme would likely have to be devised and financed through taxes or government debt.
Once the cost of compensation is taken into account, we may find that such a transition is too costly to undertake. And as the population ages and the ratio of retirees to working-age individuals increases, the fraction of individuals in the economy who need to be compensated will increase further. This reality adds impetus to dealing with the Social Security sustainability issue sooner rather than later.
June 22, 2016
Was May's Drop in Labor Force Participation All Bad News?
The unemployment rate declined 0.3 percentage points from April to May, and this was accompanied by a similar drop in the labor force participation rate. It is tempting to interpret this as a “bad” outcome reflecting a weakening labor market. In particular, discouraged about their job-finding prospects, more unemployed workers left the labor force. However, a closer look at the ins and outs of the labor force suggests a possibly less troubling interpretation of the outflow from unemployment.
To get a handle on what is going on, it is useful to look at the number of people that transition among employment, unemployment, and out of the labor force. It is not that unusual for an individual to search for a job in one month and then enroll in school or assume family responsibilities the next. In fact, each month millions of individuals go from searching for work to landing a job or leaving the labor force, and vice versa.
The U.S. Bureau of Labor Statistics (BLS) publishes estimates of these gross flows. Analyzing these data shows that there was indeed an unusually large number of unemployed persons leaving the labor force in May. Curiously, the outflow was concentrated among people who had only been unemployed only a few weeks. It wasn't among the long-term unemployed. Therefore, it seems unlikely that discouragement over job-finding prospects was the main factor. Although it is plausible that people who say they are now doing something else outside the labor market feel disheartened, the number of unemployed who said they gave up looking because they were discouraged was largely unchanged in May.
So why was there an increase in the number of short-term unemployed who left the labor force in May? One clue is provided by the fact that the short-term unemployed tend to be relatively younger than other unemployed. Moreover, the single most common reason that unemployed young people leave the labor force is to go to school. Hence, there is a very distinct seasonal pattern in the outflow. It tends to be relatively low around May when school is ending and high around August when school is starting. Seasonal adjustment techniques correct for these patterns by lowering the unadjusted data in the fall and raising it in late spring.
The following chart shows the seasonally adjusted and unadjusted flow from unemployment to departure from the labor force. Although the trend has been declining during the last few years, a relatively large increase in the seasonally adjusted outflow took place in May of this year.
When I looked at the unadjusted microdata from the Current Population Survey (CPS), I found that the number of people who were unemployed in April 2016 but in May said that they were not in the labor force because they were in school did not exhibit the usual large seasonal decline. Therefore, when the seasonal adjustment is applied, the result is an increase in the estimated flow from unemployment to out of the labor force.
Taking the seasonally adjusted data at face value, it's not obvious that this is bad news. We know that people who leave unemployment to undertake further education tend to rejoin the labor force later. Moreover, they tend to rejoin with better job-finding prospects than when they left. Alternatively, it could be just a statistical quirk of the May survey. After all, the CPS has a relatively small sample, so the estimated flows have a large amount of sampling error. Either way, I don't think it is wise to conclude that the decline in the labor force participation in May reflected a marked deterioration in job-finding prospects. In fact, the job-finding rate among unemployed workers improved in May from 22 to 24 percent, contributing to the decline in the unemployment rate.
June 21, 2016
Wage Growth for Job Stayers and Switchers Added to the Atlanta Fed's Wage Growth Tracker
The Atlanta Fed's Wage Growth Tracker (WGT) moved higher again in May—the third increase in a row and consistent with a labor market that is continuing to tighten. At 3.5 percent, the WGT is at a level last seen in early 2009.
As was noted in an early macroblog post, when the labor market is tightening, people changing jobs experience higher median wage growth than those who remain in the same job. Median wage growth for job switchers has significantly outpaced that of job stayers in recent months. For job stayers, the May WGT was 3.0 percent, the same as in April, whereas for people switching jobs the median WGT increased from 4.1 percent to 4.3 percent in May (the highest reading since December 2007; see the chart).
Because these patterns over time can help shed light on the relative strength of the labor market, we have added downloadable job stayer and job switcher WGT series to the Atlanta Fed's Wage Growth Tracker web page.
I should note that it is not possible to completely identify people who are in the same job as a year ago according to data from the Current Population Survey. Instead, we define a "job stayer" as someone whom we observe in the same occupation and industry as a year earlier, and with the same employer in each of the last three months. A "job switcher" includes everyone else (a different occupation or industry or employer). We'll be monitoring these data in coming months to see if discernable trends begin to emerge, and we'll discuss any findings here.
June 16, 2016
Experts Debate Policy Options for China's Transition
After nearly three decades of rapid economic growth, China today faces the challenge of economic rebalancing against the backdrop of slow and uncertain global growth. Although investment and exports have been a motor for growth, China is increasingly experiencing structural issues: widening inequality, overcapacity as a consequence of policy distortions, unsustainable environmental costs, volatile financial markets, and rising systemic risk.
On April 28–29, I attended the First Research Workshop on China's Economy, organized jointly by the International Monetary Fund (IMF) and the Atlanta Fed. The workshop, held at the IMF's headquarters in Washington DC, explored a series of questions that have emerged as China shifts toward a new growth model. Is this the end of the growth miracle? Will the Chinese renminbi one day be as important as the U.S. dollar? Should the rapidly increasing shadow banking activity in China be a source of concern? How worrisome is the rapid rise in China's housing prices?
Panelists shared their views on these and other issues facing the world's second-largest economy (or largest, if measured on a purchasing-power-parity basis). Plans are under way for a second workshop to be held in 2017.
The following is a nice summary of the research discussed at the workshop. It was originally published in the IMF Survey Magazine, and was written by Hui He, IMF Institute for Capacity Development, and Nan Li, IMF Research Department. Thanks to the IMF for allowing me to repost it here.
Is China's economic growth sustainable?
Understanding the source of China's tremendous growth was a recurring theme at the workshop. "China's economy combines enormous dynamism with huge distortions," observed Loren Brandt (University of Toronto). Brandt described his research based on China's firm-level data and emphasized that firm dynamics (entry and exit), especially firm entry, have been the main source of the productivity growth in the manufacturing sector.
Echoing Brandt's message, Kjetil Storesletten (University of Oslo) discussed regional growth disparities and showed that barriers preventing firms from entering an industry account for most of the disparities. Such barriers are more severe for privately owned firms in regions in which state-owned enterprises (SOE) dominate, he said.
In his keynote speech, Nicholas Lardy (Peterson Institute for International Economics) offered an upbeat view on China's transition to a new growth model, one in which the service sector plays a larger role than manufacturing. The bright side of the service sector, he noted, is its continued strong productivity growth. The development of financial deepening and the stronger social safety net are contributing to increased consumption, which helps to rebalance the economy.
However, he emphasized, SOE reforms remain critical as the service sector cannot provide a silver bullet for a successful transition.
Central bank's policy decisions
Several participants tried to discern how the People's Bank of China (PBC) conducts monetary policy. Tao Zha (of the Atlanta Fed's Center for Quantitative Economic Research and Emory University) found that the PBC reacts sharply when the gross domestic product's growth rate falls below its target, increasing the money supply by 11.5 percentage points for every 1 percentage point shortfall.
Mark Spiegel (Center for Pacific Basin Studies) discussed the trade-offs involved in Chinese monetary policy—for example, controlling the exchange rate versus maintaining inflation stability. He also argued that the heavy use of reserve requirements on banks as a monetary policy tool might have an unintentional consequence to reallocate capital from SOEs to more efficient privately owned firms and could therefore offset the resource misallocation caused by the easy credit to SOEs that banks granted in the high growth years.
Renminbi versus the dollar
Eswar Prasad (Cornell University and Brookings Institution) argued that China's capital account will become more open and the renminbi will be used more widely to denominate and settle cross-border transactions. But he also noted that legal and institutional constraints in China were likely to prevent the renminbi from serving as a safe-haven currency as the U.S. dollar does today.
Moreover, he said, the current sequencing of liberalization initiatives—that is, removal of capital account restrictions before appropriate financial market supervision and regulation and exchange rate reform—poses financial stability risks.
Shadow banking and the housing market
Recently, volatile Chinese financial markets and continued housing price appreciation have raised serious financial stability concerns.
Michael Song (Chinese University of Hong Kong) argued that rapidly rising shadow banking activity is an unintended consequence of financial regulation. Restrictions on deposit rates and loan-to-deposit ratios have led to the issuance by banks of "wealth management products" to attract savers with higher returns. Because these restrictions had a greater impact on small banks, the big state banks had more room to undercut the smaller banks by offering wealth management products with higher returns and then restricting liquidity to them in interbank markets, ultimately making the banking system more prone to liquidity distress and runs.
Hanming Fang (University of Pennsylvania) found that, except in big cities such as Beijing and Shanghai, housing prices in China's urban areas between 2003 and 2013 more or less tracked rising household incomes. In his view, the Chinese housing boom is thus unlikely to trigger an imminent financial crisis. He warned, however, that housing prices may fall rapidly if economic growth slows dramatically, and that such a development could, in turn, amplify the economic downturn.
Rising wage inequality
China's rapid growth over the past two decades has been accompanied by rising wage inequality, an issue highlighted by two conference participants. Dennis Yang (University of Virginia) explored the distributional effects of trade openness in China and found a significant impact on wage inequality of China's accession to the World Trade Organization in 2001.
Chong-En Bai (Tsinghua University) argued that the decline after 2008 of the skill premium—that is, the ratio of the skilled labor wage to the unskilled labor wage—can be explained by the Chinese government's targeted credit extension to unskilled labor-intensive infrastructure sector (as part of the fiscal stimulus following the global financial crisis). Such distortionary policies might have short-run growth benefits but could lead to long-run welfare losses, he said, especially when rural-to-urban migration has run its course.
June 09, 2016
It’s Not Just Millennials Who Aren't Buying Homes
In recent years, much attention has been focused on the growing tendency of millennials to rent. Theories for the decrease in homeownership among young adults abound. They include rising student debt levels that crowd out additional borrowing, a tendency to live in more urban areas where the cost to buy is relatively high, a generally tougher credit environment, and even shifts in the perception of homeownership in the wake of the housing bust. The ideas have been widely debated, and yet no single factor seems to neatly explain the declining share of the millennial population opting to buy a house. (See this webcast by the Atlanta Fed's Center for Real Estate Analytics for a discussion of these issues.)
To the extent that these factors are true, they may be affecting the decisions of other generations as well. Chart 1 below shows the overall average homeownership rate and homeownership rates by age group from 1982 to 2015. It's clear that homeownership rates have declined for everyone during the past 10 years, not just for millennials.
In fact, homeownership among young Generation Xers has fallen by a bit more than the millennial generation since the housing peak—declining 11 percentage points since 2005 compared with a decline of 9 percentage points for those under 35 years old.
Another interesting point of comparison is the mid-1980s to mid-1990s, a period in which the United States had a relatively stable share of owner-occupied housing of around 64.0 percent. During the subsequent housing boom, the homeownership rate climbed to a peak of 69 percent in 2004, only to fall back down to 63.7 percent in 2015, a level similar to that prevailing before 1995. However, each age group under age 65 has a somewhat lower homeownership rate than their same-aged peers had during the 1986–94 period.
The fact that the average U.S. homeownership rate is close to rates seen in the mid-1980s and mid-1990s while homeownership rates within age groups (under 65) are currently lower than their respective averages in the mid-1980s to mid-1990s suggests that factors other than age may be affecting the average person's decision to buy or rent.
To investigate what else may be going on, charts 2 and 3 show homeownership rates by family type and race. Between 2005 and 2015, the trend mirrors what's happening by age group. The tendency to own a home has been falling for all family types and races over the past decade. In general, economic incentives (or cultural attitudes) appear to have shifted the population toward renting and away from buying.
However, the picture is quite different when you compare homeownership rates by family type and race to the pre-1995 period. While homeownership rates within age groups are generally lower today, married couples, one-person households, and nonmarried, multiperson households were all more likely to own their home in 2015. Homeownership rates across race (except for blacks) were also higher in 2015 than in 1994.
So how do we interpret the fact that the overall homeownership rate is close to its average in the 1986 to 1994 period? Are millennials to blame? Yes. But so is everyone else under the age of 65. The data suggest that whatever is affecting millennials' homeownership decisions is applicable to older individuals as well. Further, it seems there are other, possibly larger, factors affecting homeownership, such as the changing face of America. Although homeownership rates by family types and racial groups are a bit above the level seen in 1994, the average person in 2015 was about as likely to live in a home that is owned or being bought. Thus, the shift in the distribution of the population toward racial groups and family types (and likely other factors) that tend to have lower homeownership rates is likely exerting an important influence on the overall homeownership rate.
June 06, 2016
After the Conference, Another Look at Liquidity
When it comes to assessing the impact of central bank asset purchase programs (often called quantitative easing or QE), economists tend to focus their attention on the potential effects on the real economy and inflation. After all, the Federal Reserve's dual mandate for monetary policy is price stability and full employment. But there is another aspect of QE that may also be quite important in assessing its usefulness as a policy tool: the potential effect of asset purchases on financial markets through the collateral channel.
Asset purchase programs involve central bank purchases of large quantities of high-quality, highly liquid assets. Postcrisis, the Fed has purchased more than $3 trillion of U.S. Treasury securities and agency mortgage-backed securities, the European Central Bank (ECB) has purchased roughly 727 billion euros' worth of public-sector bonds (issued by central governments and agencies), and the Bank of Japan is maintaining an annual purchase target of 80 trillion yen. These bonds are not merely assets held by investors to realize a return; they are also securities highly valued for their use as collateral in financial transactions. The Atlanta Fed's 21st annual Financial Markets Conference explored the potential consequences of these asset purchase programs in the context of financial market liquidity.
The collateral channel effect focuses on the role that these low-risk securities play in the plumbing of U.S. financial markets. Financial firms fund a large fraction of their securities holdings in the repurchase (or repo) markets. Repurchase agreements are legally structured as the sale of a security with a promise to repurchase the security at a fixed price at a given point in the future. The economics of this transaction are essentially similar to those of a collateralized loan.
The sold and repurchased securities are often termed "pledged collateral." In these transactions, which are typically overnight, the lender will ordinarily lend cash equal to only a fraction of the securities value, with the remaining unfunded part called the "haircut." The size of the haircut is inversely related to the safety and liquidity of the security, with Treasury securities requiring the smallest haircuts. When the securities are repurchased the following day, the borrower will pay back the initial cash plus an additional amount known as the repo rate. The repo rate is essentially an overnight interest rate paid on a collateralized loan.
Central bank purchases of Treasury securities may have a multiplicative effect on the potential efficiency of the repo market because these securities are often used in a chain of transactions before reaching a final holder for the evening. Here's a great diagram presented by Phil Prince of Pine River Capital Management illustrating the role that bonds and U.S. Treasuries play in facilitating a variety of transactions. In this example, the UST (U.S. Treasury) securities are first used as collateral in an exchange between the UST securities lender and the globally systemically important financial institution (GSIFI bank/broker dealer), then between the GSIFI bank and the cash provider, a money market mutual fund (MMMF), corporation, or sovereign wealth fund (SWF). The reuse of the UST collateral reduces the funding cost of the GSIFI bank and, hence, the cost to the levered investor/hedge fund who is trying to exploit discrepancies in the pricing of a corporate bond and stock.
Just how important or large is this pool of reusable collateral? Manmohan Singh of the International Monetary Fund presented the following charts, depicting the pledged collateral at major U.S. and European financial institutions that can be reused in other transactions.
So how do central bank purchases of high-quality, liquid assets affect the repo market—and why should macroeconomists care? In his presentation, Marvin Goodfriend of Carnegie Mellon University concluded that central bank asset purchases, which he terms "pure monetary policy," lower short-term interest rates (especially bank-to-bank lending) but increase the cost of funding illiquid assets through the repo market. And Singh noted that repo rates are an important part of the constellation of short-term interest rates and directly link overnight markets with the longer-term collateral being pledged. Thus, the interaction between a central bank's interest-rate policy and its balance sheet policy is an important aspect of the transmission of monetary policy to longer-term interest rates and real economic activity.
Ulrich Bindseil, director of general market operations at the ECB, discussed a variety of ways in which central bank actions may affect, or be affected by, bond market liquidity. One way that central banks may mitigate any adverse impact on market liquidity is through their securities lending programs, according to Bindseil. Central banks use such programs to lend particular bonds back out to the market to "provide a secondary and temporary source of securities to the financing market...to promote smooth clearing of Treasury and Agency securities."
On June 2, for example, the New York Fed lent $17.8 billion of UST securities from the Fed's portfolio. These operations are structured as collateral swaps—dealers pledge other U.S. Treasury bonds as collateral with the Fed. During the financial crisis, the Federal Reserve used an expanded version of its securities lending program called the Term Securities Lending Facility to allow firms to replace lower-quality collateral that was difficult to use in repo transactions with Treasury securities.
Finally, the Fed currently releases some bonds to the market each day in return for cash, through its overnight reverse repo operations, a supplementary facility used to support control of the federal funds rate as the Federal Open Market Committee proceeds with normalization. However, this release has an important limitation: these operations are conducted in the triparty repo market, and the bonds released through these operations can be reused only within that market. In contrast, if the Fed were to sell its U.S. Treasuries, the securities could not only be used in the triparty repo market but also as collateral in other transactions including ones in the bilateral repo market (you can read more on these markets here). As long as central bank portfolios remain large and continue to grow as in Europe and Japan, policymakers are integrally linked to the financial plumbing at its most basic level.
To see a video of the full discussion of these issues as well as other conference presentations on bond market liquidity, market infrastructure, and the management of liquidity within financial institutions, please visit Getting a Grip on Liquidity: Markets, Institutions, and Central Banks. My colleague Larry Wall's conference takeaways on the elusive definition of liquidity, along with the impact of innovation and regulation on liquidity, are here.
June 02, 2016
Moving On Up
People who move from one job to another tend to experience greater proportionate wage gains than those who stay in their job, except when the labor market is weak and there are relatively few employment options. This point was illustrated using the Atlanta Fed's Wage Growth Tracker in this macroblog post from last year.Given that the Wage Growth Tracker ticked higher in April, it is interesting to see how much of that increase can be attributed to job switching. Here's what I found:
A note about the chart: In the chart, a "job stayer" is defined as someone who is in the same occupation and industry as he or she was 12 months ago and has been with the same employer for at least the last three months. A "job switcher" is everyone else.
The overall Wage Growth Tracker for April was 3.4 percent (up from 3.2 percent in March). For job stayers, the Tracker was 3.0 percent (up from 2.9 percent), and for job switchers it was 3.9 percent (up from 3.7 percent). So the wage gains of job switchers do appear to have helped pull up our overall wage growth measure.
Moreover, unlike the wage growth of job stayers, job switchers are now tending to see wage growth of a similar magnitude to that experienced before the recession. This observation is broadly consistent with the improvement seen during the last year in the quits rate (the number of workers who quit their jobs as a percent of total employment) from the Job Openings and Labor Turnover Survey.
I think it will be interesting to continue to monitor the influence of job switching on wage growth as a further indicator of improving labor market dynamism. An update that includes the May data should be available in a few weeks.
June 01, 2016
Putting the Wage Growth Tracker to Work
The April pop in the Atlanta Fed's Wage Growth Tracker has attracted some attention in recent weeks, resulting in some interesting analysis. What is the tracker telling us about the tightness of the labor market and the risks to the inflation outlook?
We had earlier noted the strong correlation between the Wage Growth Tracker and the unemployment rate. Tim Duy took the correlation a step further and estimated a wage Phillips curve. Here's what he found:
The chart shows that lower unemployment generally coincides with higher wage growth (as measured by the Wage Growth Tracker), but wage growth varies a lot by unemployment rate. In the past, an unemployment rate around 5 percent has often been associated with higher wage growth than we currently have.
If the Wage Growth Tracker increased further, would that necessarily lead to an increase in inflation? Jared Bernstein suggests that there isn't much of an inflation signal coming from the Wage Growth Tracker. His primary evidence is the insignificant response of core personal consumption expenditure (PCE) inflation to an increase in the Wage Growth Tracker in a model that relates inflation to lags of inflation, wage growth, and the exchange rate.
However, I don't think the absence of a wage-push inflation connection using the Wage Growth Tracker is really that surprising. The Wage Growth Tracker better captures the wage dynamics associated with improving labor market conditions than rising labor cost pressures per se. For example, if firms are replacing departing workers with relatively low-wage hires, then the wages of incumbent workers could rise faster than do total wage costs (as this analysis by our colleagues at the San Francisco Fed shows). That said, as Bernstein also pointed out in the Washington Post, it's also pretty hard to find evidence of wage pass-through pushing up inflation in his model using more direct measures of labor costs.
I look forward to seeing more commentary about Atlanta Fed tools like the Wage Growth Tracker and how they can be part of the broader discussion of economic policy.
May 23, 2016
Can Two Wrongs Make a Right?
In a recent macroblog post, I showed that forecasts from the Atlanta Fed's real gross domestic product (GDP) nowcasting model—GDPNow—have been about as accurate a forecast of the U.S. Bureau of Economic Analysis's (BEA) first estimate of real GDP growth as the consensus from the Wall Street Journal Economic Forecasting Survey. Because GDPNow essentially uses a "bean-counting" approach that tallies the forecasts of the various main subcomponents of GDP, the total GDP forecast error can be broken up into the forecast errors coming from each piece of GDP. For most of the subcomponents of GDP, the contribution to total GDP growth is approximately its real growth rate multiplied by its expenditure share of nominal GDP (the exact formulas are in the working paper for GDPNow). The following chart shows the subcomponent contributions to the GDPNow forecast errors since the third quarter of 2011. (I want to note that the forecast errors are based on the final GDPNow forecasts formed before the BEA's first estimates of GDP are released.)
The forecast errors for the subcomponents can sometimes be quite large. For example, for the fourth quarter of 2013, GDPNow underestimated the combined contributions of net exports and inventory investment by nearly 2 percentage points. However, these misses were nearly offset by overestimates of the other contributions to growth (consumption, business and residential fixed investment, and government spending).
The pattern of large but largely offsetting GDP subcomponent errors has been attributed to the work of a fictional "Saint Offset," as former Fed Governor Laurence Meyer noted in a 1998 speech. Unfortunately, "Saint Offset" doesn't always come to the forecaster's aid. For example, in the fourth quarter of 2011, GDPNow predicted 5.2 percent growth—well above the BEA's first estimate of 2.8 percent—and the subcomponent errors were predominantly on the high side.
A closer look at the chart also reveals that GDPNow has had a tendency to overestimate the contribution of business fixed investment to growth and underestimate the growth contribution of inventory investment. Although these subcomponent biases have nearly offset one another on average, we really don't want to have to rely on "Saint Offset." We would like the subcomponent forecasts to be reasonably accurate because the subcomponents of GDP are of interest in their own right.Have the subcomponent biases been a unique feature of GDPNow forecasts? It appears not. Both the Survey of Professional of Forecasters (SPF), conducted about 11 weeks prior to the first GDP release, and Blue Chip Economic Indicators, conducted as close as three weeks prior to the first release, provide consensus forecasts for some GDP subcomponents. The following table provides an average forecast error (as a measure of bias) and average absolute forecast error (as a measure of accuracy) of the subcomponent growth contributions for the two surveys and comparably timed GDPNow forecasts.
We see that the biases in GDPNow's subcomponents have been fairly similar to those in the two surveys. For example, all three sources have underestimated the average inventory investment contribution to growth by fairly similar magnitudes.
The relative accuracy of GDPNow's subcomponent and overall GDP forecasts has also been similar to the accuracy of the two surveys. "Saint Offset" has helped all three forecasters; the standard errors of the real GDP forecasts are 20 percent to 40 percent lower than they would be if the forecast errors of the subcomponents did not cancel each other out.
Finally, notice that some GDP subcomponents appear to be much more difficult to forecast than others. For instance, the bias and accuracy metrics for consumer spending are smaller than they are for inventory investment. This differential is not really that surprising, because more monthly source data are available prior to the first GDP release for consumer spending than for inventory investment.
Can we take any comfort in knowing that private forecasters have mirrored the biases in GDPNow's subcomponent forecasts? An optimistic interpretation is that the string of one-sided misses are the result of bad luck—an atypical sequence of shocks that neither GDPNow nor private forecasters could account for. A more troubling interpretation is that there have been structural changes in the economy that neither GDPNow nor the consensus of private forecasters have identified. Irrespective of the reason, though, optimal forecasts should be unbiased. If biases in some of the subcomponents continue, then forecasters will need to look for a robust way to eliminate them.
- Unemployment Risk and Unions
- Cumulative U.S. Trade Deficits Resulting in Net Profits for the U.S. (and Net Losses for China)
- The Slump in Undocumented Immigration to the United States
- A Quick Pay Check: Wage Growth of Full-Time and Part-Time Workers
- Back to the '80s, Courtesy of the Wage Growth Tracker
- Introducing the Atlanta Fed's Taylor Rule Utility
- Payroll Employment Growth: Strong Enough?
- Forecasting Loan Losses for Stress Tests
- Men at Work: Are We Seeing a Turnaround in Male Labor Force Participation?
- What’s Moving the Market’s Views on the Path of Short-Term Rates?
- October 2016
- September 2016
- August 2016
- July 2016
- June 2016
- May 2016
- April 2016
- March 2016
- February 2016
- January 2016
- Business Cycles
- Business Inflation Expectations
- Capital and Investment
- Capital Markets
- Data Releases
- Economic conditions
- Economic Growth and Development
- Exchange Rates and the Dollar
- Fed Funds Futures
- Federal Debt and Deficits
- Federal Reserve and Monetary Policy
- Financial System
- Fiscal Policy
- Health Care
- Inflation Expectations
- Interest Rates
- Labor Markets
- Latin America/South America
- Monetary Policy
- Money Markets
- Real Estate
- Saving, Capital, and Investment
- Small Business
- Social Security
- This, That, and the Other
- Trade Deficit
- Wage Growth