Wednesday, April 17, 2013

Impact of National Debt: Rogoff-Reinhart vs. Herndon-Ash-Pollin

A lot of economic research sits on shelves collecting dust or, more recently, is read online by others doing economic research.  In recent days, some research has been reported by various business news services and social media.  Ken Rogoff and Carmen Reinhart have received much acclaim in recent years for research about financial crises (for example, see This Time is Different).  Before most people could imagine a financial crisis taking place in the United States, they provided research indicating that it was a distinct possibility (click here).  A follow up piece of research examined the impact of national debt on economic growth, with one of the main conclusions being that nations with a national debt of more than 90% of GDP tend to grow more slowly ("Growth in the Time of Debt").  Here's a relevant quote from the abstract of the paper:

"Above 90 percent, median growth rates fall by one percent, and average growth falls considerably more."

Rogoff and Reinhart provided their data to other economists who claimed to have found some errors in the data set which have some impact on the conclusions (see Herndon, Hash, and Pollin).   Rogoff and Reinhart responded by acknowledging one of the mistakes while defending other parts of their research.  The mistake involved a simple excel error that left out certain countries.  How does the new research affect the conclusions?  As far as I can tell, there is not a significant change in the median growth rate, but the average growth rate for high-debt countries following WW2 goes from -0.1% to 2.2%.  The critics point out that this suggests that countries shouldn't worry as much about high debt while Rogoff and Reinhart point out that this still implies slower growth for high-debt countries (3.1-3.2% for those between 30-90% debt levels and 2.2% for those above 90%) and also is now more in line with other research they have published since then covering a longer time period.  A reduction in growth of 1% per year adds up over time (think about the US experiencing a weak recovery with an average growth rate of 2% instead of a stronger recovery with growth of 3%).  There are a lot of other issues involved, but this is a blog post, not a research paper.  Why is it getting so much public attention?

Most of the attention is coming from critics of austerity and those on the left side of political spectrum.  They view the updated results as suggesting that countries shouldn't have to worry as much about high national debt.  Some also claim that the initial results provided the support for austerity programs being implemented around the world.  Are they correct?

First, Rogoff and Reinhart were careful to indicate that they thought they found a correlation, not a causation.  In other words, they didn't explicitly state that their results indicated that high debt levels led to slower growth, though that was implied.  Critics point out that weakening economic conditions result in higher debt, so it's possible that the weak growth led to high debt rather than the other way around.  Rogoff and Reinhart acknowledge this possibility, but point out that they looked at countries with weak growth and high debt for extended periods of time, not simple recessions.

Second, does it make sense for 90% to be a threshold?  I don't see any economic reasoning for a threshold number.  Some countries can have higher debt loads before suffering consequences while others may face debt crises at lower debt loads (implied by Rogoff and Reinhart as they drew different thresholds for developing economies as opposed to developed economies).  Other factors play a role such as the domestic savings rate, monetary policy, perceived credibility of the government's fiscal policy, etc.

Did this research provide the basis for austerity?  I think that's a very hard claim to justify.  Austerity programs in Europe resulted from various governments being unable to finance their debt.  For example, Greece is being forced to implement austerity as a condition of the European bailout.  In addition, since investors won't purchase Greek bonds at reasonable interest rates, there's pressure on the government to reduce their need for financing (i.e., austerity).  Is this the best approach?  That's debatable; but the motivation was not academic research, but real world pressure.

Did political conservatives try to use the research to support their perspective?  Of course.  That's one of the points of this blog; it's important to separate economics from politics.  Too many people with political agendas choose to use research that supports their bias.  Similarly, it would be improper for liberals to say there's no need to ever worry about debt loads (I'm not saying that's being said).

Does the national debt matter?  The recent experiences in Europe as well as other historical experiences clearly indicate that it does.  Nations with high debt loads that have difficulty financing their debt will experience higher interest rates that hurt economic growth, result in higher interest on the debt which crowds out other programs, etc.  Policies that seek to reduce the budget deficit in the short run (austerity) hurt economic growth and must be implemented with care.  Ideally, governments should put policies in places that avoid debt crises such as running budget surpluses during good economic times so there's more room for budget deficits during bad economic times.  As those in Europe can attest, pressure to implement austerity when the economy is weak should be avoided if at all possible.

Did Rogoff and Reinhart deliberately make mistakes?  That's hard to believe.  They have been serious economists for years and had already received praise for their work on financial crises.  In the past, they had been criticized by political conservatives for showing that recoveries from financial crises tend to be slow (and thus President Obama can't be fully blamed for the slow recovery).  Now they are being criticized by those on the left for research about problems resulting from high debt levels.  They made an embarrassing mistake, but seem to be researching economics without an agenda.

Friday, April 5, 2013

March Employment Report

Yes, this blog is still active.  This morning's job report raises quite a few questions about the state of the economy.  First the headlines: the unemployment rate dropped to 7.6%, the lowest since December 2008 (good news?  not so fast).  The economy added 88,000 jobs (95,000 in the private sector); not a good number.  However, there were upward revisions to previous months.  Let's look at some of the details.

The labor force participation rate fell to 63.3%, the lowest since May 1979.  In addition, the employment-population ratio declined to 58.5% (compared to 62.9% before the recession in November 2007, 59.4% at the end of the recession in June 2009, and a post-recession high of 58.7%).  The U6 measure of unemployment declined from 14.3% to 13.8%.  Why did U6 fall by so much?  The number of people working part-time for economic reasons dropped significantly.  Also, there was a larger than normal drop off in the number of people not in the labor force who want a job (that number normally drops for seasonal reasons in March, but the decline was much larger than usual).  If that trend continues, it would suggest that the decline in the participation rate may be more permanent (i.e., people may not re-enter the labor force when conditions improve).

On to employment: after growing by 268,000 in February, nonfarm payrolls added only 88,000 in March.  Which sectors slowed down?  After adding 14,600 jobs in February, retail trade lost 24,100 jobs in March, led by declines in clothing stores (down 15,300) and building material stores (down 10,100).  Which sectors added jobs?  Temp agencies added more than 20,000 workers, food and drinking places added 13,000 (both low-paying sectors).  The brightest spot was specialty trade contractors (part of construction), which added over 23,000 jobs.  Was the slowdown a direct result of sequestration?  Though the federal government shed 14,000 jobs, most of that was due to the trimming of postal workers (down 11,700).  Though sequestration is likely to have an impact down the road, it's effects were minimal in March.

What's the initial takeaway?  Given the strong February report, one should not read too much in the March report, but it does raise concerns.  However, this was not a good report.  Both the establishment and household surveys were weak.  There's evidence that more people are "permanently" leaving the labor force.  Also, though construction is growing, other parts of the economy are sluggish.  Sequestration played little, if any, role, but the increase in the payroll tax probably had a more significant impact.  I hope to post more thoughts later today after digging into the report a little more.

Afternoon Update:  Though seasonal adjustments seemed to have distorted the employment data in recent years, that does not appear to be an issue in this case.  For example, using the average seasonal adjustments from before the recession would have resulted in an even a smaller increase in employment in March.  Also, several people have noted the Spring slowdowns in the last two years.  In 2011, the slowdown began in May (a gain of 115,000 compared to 304,000 in April) with a brief rebound in June before a summer slump.  In 2012, the slowdown began in April and lasted through the summer.  So this slowdown is earlier than in recent years.  Of course, one must be cautious in reading too much in one month's report.

Friday, March 8, 2013

February Employment Report

The headlines from the February Employment Report contained some better than expected news about the job market: the unemployment rate fell to 7.7% (from 7.9%) and the economy added 236,000 jobs (246,000 in the private sector.  Let's look at some of the details.  The February numbers look quite good, but should be interpreted along with somewhat weaker numbers for January to assess the trend.  For example, though the private sector added 246,000 jobs in February, it added 140,000 in January, which averages 193,000 per month so far in 2013.  Looking at particular sectors, construction has added 140,000 jobs over the past year, a majority of those in the last 2 months (+73,000 in 2013).  To empahsize the significance of the pickup in construction, it added about 8,000 jobs per month, on average, in 2012, but is averaging more than 36,000 per month so far in 2013.  Other sectors and industries showing significant gains include retail trade, food accomodations, and health care.  One industry that you may not have expected to make a significant contribution, motion picture and sound recording industries, added nearly 21,000 jobs in February.

Why  did the unemployment rate fall?  A combination of more people employed and fewer people in the labor force.  The participation rate declined back to a 30-year low of 63.5% (the last time it was lower was in 1979).  Meanwhile, the employment-population ratio remained at 58.6%, unchanged so far in 2013 as well as from February 2012, but up from the post-recession low of 58.2% in November 2010 and still considerably below it's pre-recession high of 62.9% in November 2007.

What are the key takeaways from the report?  The job market has gotten off to a pretty good start in 2013, led by the rebound in construction.  That said, the employment gains in February were across the board, not concentrated in one sector.  Given the weakness in January, the job market is not signaling a robust economy, but an economy that continues to move forward.  In the coming months, there will be a battle between the rebounding housing market as seen in rising employment in construction and the fiscal drag due to the increase in the payroll tax and sequestration.

Sunday, March 3, 2013

The Impact of a Decline in the Participation Rate on the Uenmployment Rate

Recently, I was interviewed for a story in the Orlando Sentinel regarding the impact of a decline in the pariticpation rate on the unemployment rate.  Since it's hard for reporters to include all relevant information in an article, I decided to fill in some of the details that were missing.  As has been reported in many places, the labor force participation rate has declined significantly since the start of the Great Recession, falling from 66% in December 2007 to 63.6% recently.  If the participation rate was still 66%, the current unemployment rate would be more than 11% (simple math requiring no assumptions).  Should the participation rate be 66%?  A study by the Kansas City Federal Reserve explores it in some detail.  Estimates are that about half of the decline in the participation rate was due to the aging of baby boomers, which was not a result of the recession (though it did cause many to age prematurely!).  Taking that into account, a participation rate of 64.8% would result in the unemployment rate being about 9.6% (down from a peak of 10% in October 2009).  Someone was quoted in the Sentinel making the point that one can't assume that if more people were participating in the labor force, that they would be unemployed.  While this may be true, it's hard to imagine that more people participating in the job market would result in a noticeable increase in the number of jobs (some participants who dropped out of the labor force would have obtained jobs that others ended up getting) ; thus the primary result would still be a significantly higher unemployment rate.  Does that mean there has been little improvement in the job market?  The economy has added over 5 million jobs in recent years, so the job market has definitely improved, but the significant decline in the unemployment rate is due in a large part to the declining participation rate.  That's why a recent consensus forecast from a survey of economists for the Wall Street Journal predicts employment rising by 2 million in 2013 while the unemployment rate only declines to 7.4% by the end of the year.  According to the forecast, the rate of job creation is expected to increase, but the unemployment rate is not expected to decline as quickly as in recent years due to stabilization of the participation rate (and a possible minor rebound).

Saturday, March 2, 2013

A Look at the Past Week's Economic Reports

The government released several economic reports this past week, including a revision of fourth quarter GDP and personal income/spending for January.  The headline numbers show that, instead of shrinking by 0.1% in the fourth quarter, the economy grew by 0.1%.  Personal income declined by 3.6% in January, the worst month in 20 years, after rising by 2.6% in December.  What did we learn from these reports?  Not too much.

The GDP report contained minor revisions and still reflect an economy that is growing at a modest rate (the private sector continues to grow at close to a 2% rate, which is better than 0%, but nothing to brag about).

The personal income/spending report reflects the impact of the anticipation and reality of the tax changes.  Those who could shifted income into December 2012 to avoid the anticipated tax hikes, resulting in a surge in dividends and bonuses in December and a subsquent decline in January.  When was the last time the US saw such a shifting in income?  In the early 1990s after the election of President Clinton, who promised to increase taxes on the rich during the 1992 campaign; the result was a surge in income in December 1992 to avoid the higher taxes and decline in January 1993 (so that's why this January was the largest decline in 20 years).

More importantly, consumer spending, adjusted for inflation rose by 0.1% in both December and January, reflecting a slowdown from the Fall of 2012.  The savings rate, which rose in December due to the temporary surge in income, declined to 2.4% in January, the lowest since November 2007 (right before the start of the recession).  Given the conisderable noise in the data, it's important to watch how things unfold in the coming months.

What are the key takeaways?  The economy was sluggish as 2012 came to an end and consumer spending continues to move forward, but at a slow rate.  It's still a little too early to tell how much of a hit consumer spending will take from the payroll tax hike.  Early evidence indicates that low and moderate income consumers are struggling, which should contribute to a sluggish economy in the first half of 2013.  Though not part of this post, it should be noted that housing is now the strongest part of the US economy (growing at a double-digit rate over the last year), helping to offset weakness in consumer spending and reductions in government spending.

Thursday, February 21, 2013

The Fed Minutes and QE3

Those who follow financial news are aware that the release of the minutes of the Fed January meeting contributed to a decline in the stock market yesterday (Feb 20).  Basically, the minutes revealed that there was significant debate over the costs and benefits of QE3 and raised questions as to how much longer the Fed will engage in quantitative easing (click here for an earlier post describing QE3; earlier posts have also addressed the arguments for and against QE3).  In a speech earlier this month, Jeremy Stein, one of the newest members of the Board of Governors gave a speech, "Overheating in Credit Markets: Origins, Measurement, and Policy Responses," which provided useful insight into the thinking of some members of the Fed as well as ways to assess whether credit markets are functioning properly (i.e., if credit is too easy, whether credit is being misallocated - too much credit going to higher-risk borrowers, etc.).  Stein's speech is insightful, but one needs a solid understanding of finance and economics to follow it.  Here are some highlights.  Junk bond (bonds issued by high-risk corporations) issuance is at a record high as shown below:


The above chart shows that both banks (blue) and non-bank institutions (red) issued record amounts of leveraged loans at the end of 2012 and high-yield corporate bond issuance also hit a record high (top chart).   Meanwhile, credit spreads on high-yield (junk) bonds have declined significantly (see below), but are not out of line with historical averages (note: this is the difference between the interest rate on an average high-yield bond and a comparable US Treasury bond and measures the perceived risk of holding high-yield bonds):


Though the credit spread is not out of line with its historical average, it must be noted that interest rates on high-yield bonds are at record lows.  How does that work?  The credit spread of 400 basis points (i.e., 4%) indicates that interest rates on high-yield bonds are 4 percentage points above US Treasuries.  But interest rates on US Treasuries are still near record lows reached earlier in 2012.  When one adds 4% to a near record-low Treasury interest rate, the result is a record low junk bond rate.

Stein's conclusion is that there are preliminary indications of credit markets may be overheated, but not necessarily posing a significant problem yet (some other indicators are mixed).  He does raise the concern that if the trends continue, there may be a credit bubble that could have a harmful impact on the economy.

OK, what's going on?  QE1 (Fall 2008) was designed to avoid a financial collapse/depression.  QE2 (2010) was smaller and designed out of fear of possible deflation (market-based estimates of future inflation were declining significantly).  QE3 (Fall 2012) is intended to be an economic stimulus, particularly to reduce long-term unemployment which, if left unchecked, may lead to some of the currently unemployed to become unemployable, resulting in a "permanent" increase in unemployment.  Pushing interest rates lower should help interest-sensitive parts of the economy (for example, housing and autos) thus restoring stronger economic growth.  Given that housing and autos are now the strongest sectors of the economy, it seems to be working.  So what's the problem?

As we saw in the mid-2000s, if credit is too cheap, riskier borrowers may obtain credit to which they shouldn't have access, setting the stage for problems down the road as risky borrowers default, etc.  In the short run, it's helpful for more companies and individuals to get credit, but risky lending is harmful over time (recall the housing bubble and its aftermath).  Doesn't the Fed realize this?  Of course.  Those supporting continued quantitative easing perceive that lending is too risk-averse and thus should become a little riskier.  Also, the benefits of short-run stimulus offset the possible problems down the road (and it could be reversed in time to avoid some of the future problems).  Stein's speech points out that credit may already be too easy (or soon will be) and that the potential costs are rising.  In addition, reversing policy to avoid the future problems is easier said than done,

What's the key takeaway?  Conducting monetary policy is very difficult, particularly in the aftermath of a financial crisis.  Though the Fed may want to implement loose policy (policy that makes credit easier to obtain) in order to offset lending policies that are perceived to be too risk averse and to provide an overall boost to the economy, it must consider the underlying details.  Policy that results in too many risky borrowers obtaining credit can result in worse problems down the road.

Friday, February 1, 2013

January Employment Report

The January employment report was released this morning and it contained pretty good news about the job market.  The headline numbers showed that employment rose by 157,000 in January (private employment rose by 166,000) while the unemployment rate rose to 7.9%.  Once one digs into the details, the news is mixed but generally positive.  First, employment for November and December was revised up by 127,000.  Given the revisions, the economy added 181,000 jobs per month, on average, for 2012.  The strength of the job market is reassuring given the weak GDP report.

The household survey was not quite as positive as the establishment survey (the establishment survey is generally regarded as superior when assessing the growth in employment).  Not only did the unemployment rate increase slightly (back to where it stood in October 2012, but down from 8.3% in January 2012) while the employment-population ratio remained at 58.6% (up 0.1% from a year ago).  This continues the pattern of moderate employment growth, enough to accomodate growth in the labor force, but not much more (note: the participation rate didn't change and is down by 0.1% over the last 12 months).  On a related note, the broad measure of unemployment (U6) was unchanged at 14.4% (down from 15.1% one year ago and from the peak of 17.1% in late 2009/early 2010).

Though employment rose in terms of the number of job holders, hours worked was flat to negative in many sectors, with the exception of education and health services.  Average hourly earnings were up slightly in January and now have risen 2.1% over the last 12 months, slightly ahead of inflation (which is about 1.5% according to the latest figures).

What are the takeaways?  Despite the weakness of the GDP report (even after removing the effect of inventories), employment growth remains moderate, countering fears of a possible recession.  The recent trend in employment suggests an economy growing closer to 2% than 0%.  Optimists will point to the upward revisions in employment for late 2012 while pessimists will point to sluggish employment gains that are not strong enough to make a dent in the unemployment rate.  For further analysis, though I haven't looked at it yet, I'm counting on Bill McBride of calculated risk to provide solid analysis of the job report.