The Employment Report Simplified

 

The total number of employed people is rising, and the unemployment rate is dropping.

The key to understanding the employment report is defining employment.

The employer survey isn't reflecting the strength of the labor market.

The divergence between the household survey and employer survey is a cyclical phenomenon.

The potential of a monetary or policy mistake is very high.

 

Both the household survey and the employer survey showed a rise in the number of employed people. However, there is a substantial difference in the degree to which they reflected labor market conditions.

Both the household survey and the employer survey showed a rise in the number of employed people. However, there is a substantial difference in the degree to which they show the employment gain. One survey showed that about 1.1 million more people were employed in November than in October. (The household survey showed 1,136,000 more workers,) The other surveys show only about 200,000 more people were employed. (According to the employer survey total nonfarm payroll employment rose by 210,000 in November.)

The difference hinges on how the two different surveys define employment. One is dependent upon a definition of employment that involves an employee and employer. (The easy way to collect that data is to survey employers. There are less of them to be surveyed than employees.) The other depends only upon whether an individual considers himself or herself gainfully employed with or without an employer. (The household survey which is actually a survey of individuals is the only way to collect data based on the second definition.)

Each survey has its strengths and its weaknesses. Clearly, the employer survey can only measure one dimension of the functioning of the labor market. That dimension is employer hiring. Employer hiring is usually the most important dimension to measure. However, it is not true that it is the only measure or that it's equally important at all points in the business cycle. Further, it is not really measuring employment. It's measuring the number of people employers have hired.

In November the number of unemployed persons fell by 542,000 to 6.9 million. The unemployment rate fell by 0.4 percentage point to 4.2 percent in November. Saying that over 500,000 fewer people are unemployed sounds quite different from saying that there were only 200,000 more employed.

That is particularly true when one notes that the labor force has actually grown. The labor force participation rate edged up to 61.8 percent in November. Almost 600,000 people joined the workforce. In other words, there were 600,000 more people who were either unemployed or employed. The employment-population ratio increased by 0.4 percentage point to 59.2 percent in November.

It's easy to dismiss one source of data or the other as inaccurate. There are undoubtedly issues related to the accuracy of the reports that should be addressed. The most important of those is the degree to which the employment survey can be adjusted for changes in the universe of employers. If there are more small employers and the sample is not adjusted in real time to reflect that, the survey can become less accurate. (Inaccuracies of this type are often adjusted away in subsequent months.) Similarly, if the seasonal adjustments become inaccurate because of changes in the mix of employers by industry or size, they can lead to misleading data.

However, each survey is measuring what it was designed to measure. The question isn't which is accurate. The question is which is more important at this point in the business cycle. At some points in a business cycle unemployment and labor force participation are more important.

So, one has the option to view two employment reports from two different sources. One indicates how many employees have been added during the month (i.e., net hiring). The other indicates employment growth is coming from two different sources. About half is coming from a reduction in unemployment and the other half is coming from an increase in the labor force participation rate.

Regardless of the difference in the numbers reported as the result from each survey, one is providing more information about the health of the labor market than the other. The household survey is reporting on more dimensions of the health of the labor market. As was said above, which report is more important can change over the business cycle, but right now clearly the more important information is that about half the gain in employment is from reductions in unemployment and half is from growth in the labor force.

The reason the household survey is currently more important is that it provides a more accurate basis for assessing whether the growth in employment can continue. Note, the statement was “can continue”, not will continue.

The number of people who are unemployed can continue to fall. The number of people unable to work because their employer closed or lost business due to the pandemic is still reported as being quite high (3.6 million) and did not change despite the drop in unemployment. This represents a pool of unemployed individuals that can be reduced, and the existence of the pool is extremely important because it can offset the drop in number of people on temporary layoff.

There are also significant numbers of people who are still available to join the labor force. (The number in the labor force who currently want a job was 5.9 million and those marginally attached to the labor force was 1.6 million). Despite the rise in the labor force participation rate, the size of this pool that could be attracted into the labor market did not change significantly last month. So, the labor force participation rate could continue to rise and continue to add to employment.

Generally, the household survey becomes more important when confidence is rising enough for individuals to begin to feel that they can be gainfully employed without having to rely upon an employer. That is a cyclical phenomenon that I've written about before. (August 4, 2010 entitled "An Article about a Fiction and the Employment Report" and April 9, 2017 entitled “The Curious Case Of Friday's Employment Report”).

It would be quite difficult to argue that confidence has been rising currently. Neither surveys of consumer confidence nor for surveys of employer confidence would support the argument. At the same time, there is considerable evidence that people are going out and starting a business or becoming self-employed instead of seeking employment with an employer. For example, the WALL STREET JOURNAL on November 30, a couple of days before the employer report, had an article entitled “Workers Quit Jobs In Droves In Order to Become Their Own Bosses.” The article recites a number of facts that support the argument that workers are choosing to be self-employed or start a business rather than seeking employment. The types of data it cites include:

“The number of unincorporated self-employed workers has risen by 500,000 since the start of the pandemic, Labor Department data show, to 9.44 million. That is the highest total since the financial-crisis year 2008, except for this summer…. Entrepreneurs applied for federal tax-identification numbers to register 4.54 million new businesses from January through October”

It's worth noting that the article talks about people quitting jobs in order to form their own business, rather than unemployed individuals becoming self-employed. If that's an accurate reflection of what's happening, it would suppress the data from the employer survey regarding net new hires because these individuals are leaving their jobs to start a business. At the same time, it would increase the employment count in the household survey. So, the thesis of the article is totally consistent with the employment report. It's also consistent with the anomalous fact that both job openings and the rate at which workers quit jobs have been at record highs in recent months.

It is not, however, consistent with the idea that the phenomena being reported reflects increasing confidence. The article goes on to describe a number of cases of individuals who've taken this step and generally characterizes the step as resulting from “looking for flexibility, anxious about Covid exposure, upset about vaccine mandates or simply disenchanted with pre-pandemic office life.”

Whenever migration occurs, whether it's geographic migration or migration from an employer to self-employment, it is motivated by both push and pull, both positive attractions of the new status and repulsions of the previous status. If one looks at the data and reports of people refusing to discontinue remote work, refusing to get vaccinated, and afraid of front line Covid exposure, the balance seems to indicate that the push is more important.

In terms of push or pull it's easy to interpret wage data either way. One can argue that the increase in money wages, eliminates an incentive to try self-employment. However, a more sophisticated look that realizes that real wages are not increasing, but actually decreasing due to inflation, creates an incentive to try self-employment.

Regardless of how one interprets the move to self-employment, the differences between the motivations in this business cycle, with the presence of Covid and the dislocations it has caused, and previous business cycles, increases the risk of a policy mistake. That risk is higher currently than would normally be the case at this point in the business cycle because the recovery is more fragile at this stage than would normally be the case. An economy that is believed to be strong as evidenced by high levels of consumer and business confidence is much less fragile than one where both consumers and businesses are skeptical.

There's also a risk inherent in having to make decisions with data that isn't reliable; and where employment is concerned, clearly, there is some doubt as to what's going on. Some of it is a cyclical phenomenon, but some of it is unique to the current situation. For example, the pandemic has reduced response rates by employers, according to Labor Department figures, which could result in volatile data with wide revisions.

Seasonal adjustments are another potential source of error that is being affected by the pandemic. The industry makeup of employment is influenced by the fact that certain industries are now less appealing as a source of employment because of the associated Covid risk. Those are the industries that we referred to as employing frontline workers, hospitality and retail trade are just two examples. Without the seasonal adjustment, payrolls grew by 778,000 in November. So, the seasonal adjustments were large enough that if incorrect, they affect the picture that one gets from the data.

The risk of policy mistakes also rises when it is necessary to make policy decisions because one can no longer paper over problems with meaningless rhetoric. One of my favorites right now is Biden's (and the Fed’s) referring to the current inflation as a supply or a shortage issue. It seems to me that anybody who has taken even one economics course knows that prices are the product of the interaction of supply and demand.

It is absolutely silly to refer to it as being due to one or the other. Inflation is always a demand issue and a supply issue because it is the interaction of the two that changes price. It is much harder to develop a policy that responses to reality than to avoid facing reality by using language that pretends it doesn't exist. A policy that balances the need for increased supply while maintaining demand is far more complex than that implied by some mumbo-jumbo language to avoid the issue.

However, that's only a small example of the use of terms that are designed to paper over an issue. Another of my favorites was the use of the word transitory when describing inflation. It was completely meaningless, and thus, now that the Fed has finally admitted that inflation wasn't transitory people are treating it as if the change in wording means something. As long as they could paper over the issue with meaningless terms like transitory, no one had to address the more difficult issue of what the appropriate policy should be to reduce the rate of inflation. The complexity of that issue and the potential for a policy mistake is illustrated by Chairman Powell’s reference to the fact that there is uncertainty regarding the relative impact of Covid on supply and demand.

I also get a kick out of the use of the term stagflation. People ignore the fact that it was a product of a different period of time under a different regulatory environment. They pay no attention to the fact that regulation Q, limiting the interest rates banks could pay, and usury laws, limiting the interest rates banks could charge, were common during the period when inflation built up. There is a danger that a policy response to the current environment that resembles the policy response that was appropriate during the 70s would be totally inappropriate in the current regulatory environment.

Heaven help us if the demagogues like Elizabeth Warren get their way and start heavy-handed bank regulation. I have no confidence that they know what they're doing. I think no one should be allowed to talk about regulation unless they prove they can define the difference between insolvent and illiquid and explain why it's important. It's easy to conceive of Congress or the president enacting regulatory changes that would totally confuse the impact of either fiscal policy or monetary policy.

Regardless of whether one thinks heavy-handed bank regulation or a potential response to another Covid wave are desirable, is clear they will complicate policy decisions surrounding how to lengthen the recovery. That, in a nutshell, is why the potential for policy mistake is higher. There is a strong potential that there will be conflicts between what's desirable depending upon the significance one attaches to different social objectives.