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

COVID-19 RESOURCES AND INFORMATION: See the Atlanta Fed's list of publications, information, and resources; listen to our Pandemic Response webinar series.

About


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

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

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

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

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

July 28, 2011

Lots of ground to cover

In my last post I noted that the pace of the recovery, now two years old, is in broad terms similar to that of the first two years of the previous two recoveries. The set-up included this observation:

"Though we have grown used to thinking of the rebound from the most recent recession as being spectacularly substandard, that impression (which I share) is driven more by the depth of the downturn than the actual speed of the recovery."

The context of the depth of the downturn is not, of course, irrelevant. One way of quantifying that context is to look at measures of the "output gap," that is, the difference between the level of real gross domestic product (GDP) and the economy's "potential." An informal way to think about whether or not a recovery is complete is to mark the time when the output gap returns to zero, or when the level of GDP returns to its potential.

There are several ways to estimate potential GDP, but for my money the one constructed by the Congressional Budget Office (CBO) is as good as any. And it does not tell a pretty story:

Real GDP-Real Potential GDP

It is worth noting that the CBO's measure is not a just a simple extrapolation of a constant trend, but a calculation based on historical relationships among labor hours, productivity growth, unemployment, and inflation. Their trend in potential GDP growth rates implied by this methodology, described here, is anything but linear:

Real Potential GDP

Note that the output gaps in the first chart are at historical lows (by a lot) despite the fact that potential GDP growth is at historical lows as well.

These estimates provide one way to assess the pace of the recovery. For example, the midpoints of the Federal Open Market Committee's (FOMC) most recent consensus forecasts for GDP growth are 2.8 percent (2011), 3.5 percent (2012), and 3.85 percent (2013). If those forecasts come to pass, approximately 60 percent of the CBO-implied gap will be closed. This would still leave, in real terms, more resource slack than existed at the lowest point in the past two recessions.

Put another way, if the economy grows at 4 percent from 2012 forward, the output gap won't be closed until sometime in 2015. At a growth rate of 3.5 percent—the lower end of FOMC participants' projections for the next two years—the "full recovery" date gets pushed back to 2016. If, however, the FOMC projections are too optimistic and the economy can only manage to grow at an annual pace of 3 percent (which is currently the consensus view of private forecasters for 2012) output gaps persist until 2020.

The conventional view of the macroeconomy that motivates the CBO estimates of potential GDP (and hence output gaps) at least implicitly embeds the assumption that time heals all wound. But the healing won't necessarily be fast.

David Altig By Dave Altig, senior vice president and research director at the Atlanta Fed

July 20, 2011

Is consumer spending the problem?

In answer to the question posed in the title to this post, The New York Times's David Leonhardt says absolutely:

"There is no shortage of explanations for the economy's maddening inability to leave behind the Great Recession and start adding large numbers of jobs…

"But the real culprit—or at least the main one—has been hiding in plain sight. We are living through a tremendous bust. It isn't simply a housing bust. It's a fizzling of the great consumer bubble that was decades in the making…

"If you're looking for one overarching explanation for the still-terrible job market, it is this great consumer bust."

Tempting story, but is the explanation for "the still-terrible job market" that simple?

First, some perspective on the pace of the current recovery. Though we have grown used to thinking of the rebound from the most recent recession as being spectacularly substandard, that impression (which I share) is driven more by the depth of the downturn than the actual speed of the recovery. The following chart traces the path of real gross domestic product (GDP) from the trough of the last three recessions:


In the first two years following the 1990–91 and 2001 recessions, output grew by about 6 percent. Assuming that GDP grew at annual rate of 1.5 percent in the second quarter just ended—a not-unreasonable guess at this point—the economy will have expanded by about 5.3 percent since the end of the last recession in July 2009. That's not a difference that jumps off the page at me.

Directly to the point of consumption spending, it is certainly true that consumer spending has expanded at a slower pace in the expansion to this point than was the case at the same point in the recoveries following the previous two recessions. From the end of the recession in the second quarter of 2009 through the first quarter of this year (we won't have the first official look at this year's second quarter until next week), personal consumption expenditures grew in real terms by just under 4 percent. That growth compares to 4.8 percent in the first seven quarters following the end of the 2001 recession and 5.9 percent in the first seven quarters following the end of the 1990–91 recession.

That difference in the growth of consumption across the early quarters of recovery after the 1990–91 and 2001 recessions with little discernible difference in GDP growth across those episodes illustrates the pitfalls of mechanically focusing on specific categories of spending. In fact, the relatively slower pace of consumer spending in this expansion has in part been compensated by a relatively high pace of business spending on equipment and software:


If you throw consumer durables into the general notion of "investment" (investment in this case for home production) the story of this recovery is the relative boom in capital spending compared to recent recoveries:


And what about that "still-terrible job market"? You won't get much argument from me about that description, but here again the reality is complicated. Focusing once more on the period since the end of the recession, the pace of job creation is not out of sync in comparison to recent expansions (though job creation after the last two recessions was meager as well, and we are, of course, starting from a much bigger hole in terms of jobs lost):


So, relative to recent experience, at this point in the recovery GDP growth and employment growth are about average (if we ignore the size of the recession in both measures). The undeniable (and relevant) human toll aside, the current recovery seems so disappointing because we expect the pace of the recovery to bear some relationship to the depth of the downturn. That expectation, in turn, comes from a view of the world in which potential output proceeds in a more or less linear fashion, up and to the right. But what if that view is wrong and our potential is a sequence of more or less permanent "jumps" up and down, some of which are small and some of which are big?

In addition, investment growth to date has been strong relative to recent recoveries and, as Leonhardt suggests, consumption growth has been somewhat weak. So here's a question: Would we have had more job creation and stronger GDP growth had businesses been more inclined to add workers instead of capital? And if that had occurred, might the consumption numbers have been considerably stronger?

David Altig By Dave Altig, senior vice president and research director at the Atlanta Fed

 

May 28, 2010

How "discouraged" are small businesses? Insights from an Atlanta Fed small business lending survey

Roughly half of U.S. workers are employed at firms with fewer than 500 employees, and about 90 percent of U.S. firms have fewer than 20 employees. While estimates vary, small businesses are also credited with creating the lion's share of net new jobs. Small businesses are, in total, a big deal. Thus, it is no surprise that there is congressional debate going on about how to best aid small businesses and promote job growth. Many people have noted the decline in small business lending during the recession, and some have suggested proposals to give incentives to banks to increase their small business portfolios. But is a lack of willingness to lend to small businesses really what's behind the decline in small business lending? Or is it the lack of creditworthy demand resulting from the effects of the recession and housing market distress?

Economists often face such identification dilemmas, situations in which we would like to know whether supply or demand is the driving factor behind changes within a market. Additional data can often help solve the problem. In this case we might want to know about all of the loans applied for by small businesses, whether the loans were granted and at what rates, and specific information on loan quality and collateral. Alas, such data are not available. In fact, the Congressional Oversight Panel in a recent report recommends that the U.S. Treasury and other regulators "establish a rigorous data collection system or survey that examines small business finance" and notes that "the lack of timely and consistent data has significantly hampered efforts to approach and address the crisis."

We at the Federal Reserve Bank of Atlanta have also noted the paucity of data in this area and have begun a series of small business credit surveys. Leveraging the contacts in our Regional Economic Information Network (REIN), we polled 311 small businesses in the states of the Sixth District (Alabama, Florida, Georgia, Louisiana, Mississippi and Tennessee) on their credit experiences and future plans. While the survey is not a stratified random sample and so should not be viewed as a statistical representation of small business firms in the Sixth District, we believe the results are informative.

Indeed, the results of our April 2010 survey suggest that demand-side factors may be the driving force behind lower levels of small business credit. To be sure, when asked about the recent obstacles to accessing credit, some firms (34 firms, or 11 percent of our sample) cited banks' unwillingness to lend, but many more firms cited factors that may reflect low credit quality on the part of prospective borrowers. For example, 32 percent of firms cited a decline in sales over the past two years as an obstacle, 19 percent cited a high level of outstanding business or personal debt, 10 percent cited a less than stellar credit score, and 112 firms (32 percent) report no recent obstacles to credit. Perhaps not surprisingly, outside of the troubled construction and real estate industries, close to half the firms polled (46 percent) do not believe there are any obstacles while only 9 percent report unwillingness on the part of banks.

These opinions are reinforced by responses detailing the firms' decisions to seek or not seek credit and the outcomes of submitted credit applications.

052810a
(enlarge)

Of the 191 firms that did not seek credit in the past three months, 131 (69 percent) report that they either had sufficient cash on hand, did not have the sales/revenues to warrant additional debt, or did not need credit. (Note the percentages in the chart above reflect multiple responses by firms.) These responses likely reflect both the impact of the recession on the revenues of small firms as well as precautionary/prudent cash management.

The administration has recently sent draft legislation to Congress for a supply-side program—the Small Business Lending Fund (SBLF)—to address the funding needs of small businesses. The congressional oversight report raises a good question about the potential effectiveness of supply-side programs:

"A small business loan is, at its heart, a contract between two parties: a bank that is willing and able to lend, and a business that is creditworthy and in need of a loan. Due to the recession, relatively few small businesses now fit that description. To the extent that contraction in small business lending reflects a shortfall of demand rather than of supply, any supply-side solution will fail to gain traction."

That said, one way that a supply-side program like SBLF would make sense, even if low demand is the force driving lower lending rates, is if there are high-quality borrowers that are not applying for credit merely because they anticipate that they will be denied. We could term these firms "discouraged borrowers," to co-opt a term from labor markets (i.e., discouraged workers).

If a program increased the perceived probability of approval, either by increasing approval rates via a subsidization of small business lending or merely by changing borrower beliefs, more high-quality, productive loans would be made.

Just how many discouraged borrowers are out there? The chart above illustrates that, indeed, 16 percent of all of our responding firms and 21 percent of construction and real estate firms might fall into this category. I add "might" because the anticipation of a denial may well be accurate but based on a lack of creditworthiness and not the irrational or inefficient behavior of banks. Digging into our results, we find that 35 percent of the firms who did not seek credit because of the anticipation of a denial also cited "not enough sales," indicating that a denial would likely have reflected underlying loan quality.

In the labor market, so-called "discouraged workers" flow back into the labor force when they perceive that the probability of finding an acceptable job has increased enough to make searching for work, and working, attractive again. We should expect so-called "discouraged borrowers" to do the same. That's because if they don't, the likely alternatives for them, at some point, would be to sell the business or go out of business. It seems unlikely that, facing such alternatives, a "discouraged" firm would not attempt to access credit. The responses of firms in our sample are consistent with this logic; 55 percent of those who did not seek credit in the past three months because of the anticipation of a denial indicated that they plan to seek credit in the next six months.

Our results also provide some interesting data on an assumption underlying the policy debate: that those small businesses are credit constrained. Of the 117 firms in the survey that that sought credit during the previous three months, the following chart illustrates the extent to which these firms met their financing needs.

052810b
(enlarge)

Based on firm reports of the credit channel applications submitted in the previous three months, we created a financing index value for each firm. Firms that were denied on all of their credit applications have a financing index equal to 1, while firms that received all of the funding requested have an index level of 5. Index levels between 1 and 5 indicate, from lesser to greater, the extent to which their applications were successful. In the chart we plot data on the financing index levels of all firms in our sample and then split according to whether the firm is in construction and real estate. Among construction and real estate firms, 50 percent of firms had an index below 2.5, suggesting most did not get their financing requests meet. In contrast, the median index value of 4.7 for all other firms suggests that most of these firm were able to obtain all or most of the credit they requested. This difference between real estate–related firms and others is really not surprising given that the housing sector was at the heart of the financial crisis and recession. But it does suggest that more work needs to be done to analyze the industry-specific funding constraints among small businesses.

By Paula Tkac, assistant vice president and senior economist, of the Atlanta Fed

April 29, 2010

Consumer credit: More than meets the eye

A lot has been made (here, for a recent example) of the idea that banks have shown a surprising amount of reluctance to extend credit and to start making loans again. Indeed, the Fed's consumer credit report, which shows the aggregate amount of credit extended to individuals (excluding loans secured by real estate), has been on a steady downward trend since the fall of 2008.

Importantly, that report also provides a breakdown that shows how much credit the different types of institutions hold on their books. Commercial banks, which are the single largest category, accounted for about a third of the total stock in consumer credit in 2009. The two other largest categories—finance companies and securitized assets—accounted for a combined 45 percent. While commercial banks have been the biggest source of credit, they have not been the biggest direct source of the decline.

042910a
(enlarge)

The chart above highlights a somewhat divergent pattern among the big three credit holders. This pattern mainly indicates that credit from finance companies and securitized assets has been on a relatively steady decline since the fall of 2008 while credit from commercial banks has shown more of a leveling off. These details highlight a potential misconception that commercial banks are the primary driver behind the recent reduction in credit going to consumers (however, lending surveys certainly indicate that standards for credit have tightened).

To put a scale on these declines, the aggregate measure of consumer credit has declined by a total of 5.7 percent since its peak in December 2008 through February 2010. Over this same time period, credit from finance companies and securitizations declined by 16.2 percent and 12.4 percent, respectively, while commercial bank credit declined by 5.5 percent. Admittedly, securitization and off-balance sheet financing are a big part of banks' activity as they facilitate consumers' access to credit. The decline in securitized assets might not be that surprising given that the market started to freeze in 2007 and deteriorated further in 2008 as many investors fled the market. Including banks' securitized assets that are off the balance sheet would show a steeper decline in banks' holdings of consumer credit.

A significant factor in evaluating consumer credit is the pace of charge-offs, which can overstate the decline in underlying loan activity (charge-offs are loans that are not expected to be paid back and are removed from the books). Some (here and here) have made the point that the declines in credit card debt, for example, reflect increasing rates of charge-offs rather than consumers paying down their balances.

How much are charge-offs affecting the consumer credit data? Unfortunately, the Fed's consumer credit statistics don't include charge-offs. However, we can look at a different dataset that includes quarterly data on charge-offs for commercial banks to get an approximation. We can think of the change in consumer loan balances roughly as new loans minus loans repaid minus net loans charged off:

Change in Consumer Loans = [New Loans – Loan Repayments] – Net Charge-Offs

Adding net charge-offs to the change in consumer loans should give a cleaner estimate of underlying loan activity:

  042910c

If the adjusted series is negative, loan repayments should be greater than new loans extended, which would lend support to the idea that loans are declining because consumers are paying down their debt balances. If the adjusted series is positive, new loans extended should be greater than loan repayments and adds support to the hypothesis that part of the decline in the as-reported loans data is from banks removing the debt from their books because of doubtful collection. Both the as-reported and adjusted consumer loan series are plotted here:

042910b
enlarge

Notably, year-over-year growth in consumer loans adjusted for charge-offs has remained positive, which contrasts the negative growth in the as-reported series. That is, the net growth in new loans and loan repayments shows a positive (albeit slowing) growth rate once charge-offs are factored in. Over 2009, this estimate of charge-offs totaled about $27 billion while banks' average consumer loan balances declined by about $25 billion. Thus, a significant portion of the recent decline in consumer loan balances is the result of charge-offs.

Nevertheless, in an expanding economy, little or no credit growth implies a declining share of consumption financed through credit. Adjusting consumer loans for charge-offs suggests that the degree of consumer deleveraging across nonmortgage debt is somewhat less substantial than indicated by the headline numbers.

All in all, the consumer credit picture is a bit more complicated than it appears on the surface. A more detailed look suggests that banks haven't cut their consumer loan portfolios as drastically as sometimes assumed. The large run-up in charge-offs has also masked the underlying dynamics for loan creation and repayment. Factoring in charge-offs provides some evidence that a nontrivial part of consumer deleveraging is coming through charge-offs.

By Michael Hammill, economic policy analysis specialist in the Atlanta Fed's research department