Wages are flat because unemployment IS so low

The New York Times had a feature on February 2nd on why wages continue to lag despite the extremely low unemployment rates at the moment.

There is a basic presumption in this article that is no longer valid, however: namely that the unemployment rate today is the same kind of unemployment rate from 10 or 20 or even 30 years ago.

In Wisconsin, the news for some time has been how the state’s unemployment rate and benefit payments to claimants are at record lows. For instance, a January 2018 press release from the Department includes the following observations:

Other indicators of the state of Wisconsin’s economy include:

• Initial UI claims ended 2017 at their lowest level in the last 30 years.
• Continuing unemployment claims ended 2017 at their lowest level since 1973.

See also this October 2017 press release. What is notable here is that this decline is well known and part of an apparent plan.

As previously noted here, this decline is occurring because of Department efforts at making it harder for the unemployed to qualify for unemployment benefits and then disqualifying them for not jumping through some state requirement fast enough or alleging unemployment fraud for nothing more than simple claim-filing mistakes.

But, the data in Wisconsin does not explain what is happening nationally. The National Employment Law Project has already noted how unemployment has changed significantly across the nation the last few years. But, thanks to the efforts of some smart folks in Pennsylvania, national unemployment data is now available in a highly convenient format and which produces eye-catching charts.

NOTE: I cannot say enough good things about this unemployment data explorer. Pretty much any unemployment data currently being collected is now available for quick analysis in a chart. Moreover, you can easily see and download the data being used to create the charts. Excellent work.

This data provides some charts that can compare what is happening from state to state. For instance, separation data (how claims are denied because of a quit or a discharge for which misconduct/substantial fault is found) presents the following set of charts:

The red line in these charts is the national rate. As obvious, this chart shows that there is a great deal of variation from state to state. And, because there have been big changes in the number of claims being filed, this data is somewhat incomplete. See “Employer UI taxes declining because more UI claims being denied” (24 August 2016) for an examination of how changes in the number of claims being filed affect Wisconsin’s claim-filing numbers.

But, the variation among the states also reveals some obvious increases in denial rates in Kansas, Maryland, New Mexico, South Carolina, and Wisconsin. On the other hand, Alabama, Colorado, Florida (a surprise), Mississippi, New Hampshire, New Jersey, and New York show significant drops in denial rates.

This 50-state data gets even more interesting when “other” reasons for denying unemployment benefits are examined:

NOTE: “Other” reasons include, for instance, a claimant not being able and available, not completing the required job search actions for that state, not attending call-in or meeting requirements a state has mandated, or not registering for various state-mandated services. Wisconsin specific data on these issues “other” denial reasons is available here from this prior post on the financial impact of Wisconsin’s substantial fault disqualification. Outside of able and available status, these “other” reasons generally encompass requirements an individual state creates as part of its claim-management bureaucracy for supervising the unemployed.

Again, the red line in this chart is a national average of cases being decided for “other” reasons. As evident here, there has been an increase (and even an explosion in some states) in “other” denial reasons the last few years in Florida, Louisiana, Massachusetts, Michigan, Mississippi, Montana, New Mexico, New York, Ohio, Oklahoma, Pennsylvania, South Carolina, and Tennessee. Indeed, the national trend of these “other” cases increasing over time would probably be much more significant if the extremely large populations of Texas and California were removed from the analysis (Texas has been flat, and California has actually declined significantly). As such, national data is masking significant changes in the availability of unemployment benefits in numerous states.

Of course, this table is simply showing the number of cases being decided for “other” reasons. If all of these cases did NOT lead to a denial of an unemployment claim, then there is essentially no harm, no foul in these cases. But, the actual denial rates for “other” reasons reveal a not-so innocent story.

Where California declined and Texas increased slightly and then plateaued, Delaware, Florida, Iowa, Illinois, Indiana, Louisiana, Massachusetts, Maryland, Michigan, Mississippi, Montana, North Carolina, North Dakota, New Mexico, New York, Ohio, Oklahoma, Pennsylvania, South Carolina (showing shocking jumps from year to year), Tennessee, Utah, Wisconsin, and Wyoming have seen obvious increases in denial rates since 2005.

These other reasons for the most part did not exist until very recently and almost none go back a decade in significant numbers in any one state. The Wisconsin data on this issue, for instance, is telling: cases involving profiling registration requirements were in single or low double-digits until 2015 when they sky-rocketed to hundreds and then thousands. What this last chart reveals is that numerous states have essentially created numerous mechanisms for disqualifying claimants even when those claimants are initially eligible for unemployment benefits.

Given that we are all human and can only take so much abuse before moving on, it is extremely likely that most folks have simply stopped filing unemployment claims because of the obstacles states have placed on their eligibility and not because they have found the jobs they have wanted all along.

So, if states are making it much, much harder to receive unemployment benefits when filing a claim, then the low unemployment rates of today are NOT comparable to the low unemployment rates of yesteryear or even to unemployment rates of a decade or so ago. Instead of creating a question about how low wages and low unemployment rates can co-exist, the low unemployment rates of today may actually be placing a brake on wage growth: the state unemployment policies at issue here increase the supply of individuals looking for any work in place of their missing unemployment benefits. That increase in the labor supply, as a result, creates downward pressure on wages. At least, that is what I learned in labor economics 101.