By Julius Probst
In this blog post I quickly want to summarize what is happening in the US labor market and at the same time ask (and tentatively answer) the question: will there will be a return to normal any time soon?
Last Friday, the Bureau of Labor Statistics (BLS) released the job numbers for the month of April. Anybody who has been following the data closely already knew that this would be the most horrible jobs report in history. We at Macrobond have also been posting the weekly job claims number regularly on our Twitter account, and it is also a Watchlist item in our platform.
The crazy chart below shows that over the last two months, some 26 million people have filed for unemployment claims, thus erasing all the job gains from the previous 10-year expansion within just a few weeks. While a normal weekly job claims number was just North of 200.000 last year before Corona, it rose to more than 6 million in two consecutive weeks this April.
The US labor force fell by more than 20 million, reaching a low that was last seen at the height of the Great recession in 2010/2011. This is already driving down the labor force participation rate, which fell by several percentage points within just one month and therefore reached a new historic low for the 21st century.
The unemployment rate number that came out last Friday was truly astonishing, in a bad way, since it rose to a level that the US economy hasn´t seen since the Great Depression in the 1930s. The unemployment rate increased by 10 percentage points in April from a little more than 4% to close to 15%. For context, the largest ever monthly change during the postwar period was a 1.3 percentage point increase in 1949 and the largest ever decrease was a 1.5% decline during the same year.
I have also compared the monthly path of the unemployment rate using all postwar recessions for the US economy. As you can see, the unemployment rate increases by more than 50% in the typical recession and even 4 years out it is still some 30% higher than at the beginning of the downturn.
The Great recession of 2008 was already a big outlier since the unemployment rate more than doubled over the course of 2009. Moreover, even 4 years out it was still some 70% higher than at the start of the recession. As we all know, it took the Fed almost an entire decade to return to a state of full employment and to push the unemployment rate below 4% again.
Macrobond Moment: I have compared all postwar US recessions here using our Slice tool, which allows you to overlay series at different points in time. I have then used our Cross Section tool to calculate the highest, lowest and average unemployment performance for US recessions after WWII.
With the current COVID shock, the unemployment rate has already increased by more than 420% within just two months. And surely there is more bad news to come. According to some economists, like Justin Wolfers, the true unemployment rate of 15% is understated for technical reasons, including the unprecedented size and speed of this labor market shock. There is good reason to believe that the actual unemployment rate is much closer to and probably already exceeding 20% right now.
As always, the most vulnerable groups in society are suffering the most from this extreme economic shock. The unemployment rate for African Americans is almost always roughly twice as high as the unemployment rate for white Americans. Minorities suffer the most during periods when the labor market is not at full employment. It is easier to discriminate during periods of labor market slack than when the labor market is running hot and workers are scarce.
Macrobond Moment: Our charting tool easily allows you to put one series like a ratio on a second pane like I have done in this example. Take the opportunity to try it for free!
Similarly, the unemployment rate by education shows that workers that are high school drop-outs have an unemployment rate that is double compared to workers with a college degree. With the current economic shock, low-education workers are now recording an unemployment rate of more than 20%, while college educated workers are recording an unemployment rate of “only” 8%.
The current shock to the labor force also has an effect on average nominal earnings across sectors we have never seen before. Low-wage workers are affected the most and have been laid off to a much greater extent than high-wage workers across every sector in the economy. As a result, average hourly earnings have been rising at an unprecedented level. Just to reiterate, this is simply due to a composition effect. Nobody is getting any substantial pay increases right now. The average is only increasing because so many low-wage individuals have been let go.
Macrobond Moment: You can also easily create tables in Macrobond using our Scalar function and then adding a bar chart.
While womens’ unemployment rate has been below the men’s unemployment rate in recent decades, especially during the Great recession, the current COVID shock is mainly affecting services. As women are more likely to work in the service sector, the female unemployment rate increased more strongly in April and is now exceeding men’s unemployment significantly for the first time since the late 1970s.
In fact, the current economic shutdown will be the first service sector recession that the US is experiencing. As the chart below shows, personal expenditures for services is significantly less volatile than personal expenditures on goods and private domestic investment for example (note that I had to put private investment on an alternate scale on the left had since investment is highly procyclical and much more volatile than the other components of GDP).
Even during the Great recession, service expenditures barely went negative while during all other postwar recessions it never went below zero. However, this time is different as the economic shutdown has led to a huge contraction in the service industry.
Macrobond Moment: Using “Fill Range”, you can quickly add recession bands to a time chart.
As the table below shows, it is the service industry that has suffered the most this time with some 17 million job losses during the month of April. Unsurprisingly, leisure and hospitality accounted for some 7.5 million job losses alone.
The only piece of good news right now is that a significant chunk of the newly unemployed are classified as temporary layoffs. In fact, there are good reasons to believe that a lot of the current malaise might be temporary, as a substantial part of the unemployed are right now classified as temporary layoffs. This is a main difference to other downturns like the Great recession when most of the increase in the unemployment rate was permanent.
When the current economic shutdown is finally over, we can expect that many of these laid off workers will go back to work. Obviously, a lot will depend on to what extent the current shutdown has forced companies to go out of business and how long the shutdown will last.
As Lars Christensen has argued, the current shock should not affect permanent income to a large extent. According to that theory, consumption growth should resume quickly, which would mean that the service sector can potentially recover very quickly.
However, there are also reasons to doubt this assumption. Private sector balance sheets have deteriorated as a result of this substantial shock. If the Japanese experience has taught us anything, it is that balance sheet shocks can have significant effects on economic performance for many years.
Risk aversion probably has increased significantly. It is not entirely clear to me when and how things will get back to normal. While the current savings shock mostly is a result of ”forced savings”, with the current economic shutdown consumers are simply unable to spend their money, there is an argument to be made that this savings shock will not be reversed so quickly.
Even when the economy opens up again, people might still be reluctant to go out if the virus is still spreading. I do not think it is plausible to assume that the entire leisure and hospitality industry, where most job losses occurred will go back to normal right away. Both international travel but also domestic travel is way down and there will not be a return to normalcy even after the lockdown has ended.
Moreover, there is a reason to suspect that this current economic downturn will have a semi-permanent effect on risk aversion, for both companies and individuals. An increase in risk aversion would translate into a significantly higher savings rate than before the crisis and subtract from current spending. If this would be the case, such a savings shock could negatively affect aggregate demand for some time to come and therefore prevent a quick economic recovery.
In the chart below, I have quickly estimated Okun’s law for the US, which describes the relationship between GDP growth and unemployment, using data for the postwar period. The Okun’s law coefficient tells us that for every percentage point reduction in GDP growth, the unemployment rate rises by 0.6 percentage points (1/1.6) – or alternatively that a 1 percentage point increase in unemployment is associated with a 1.6 percentage point reduction in growth.
Macrobond Moment: I have estimated Okun’s law using our Regression tool and created the chart above using Scatter Plot. Moreover, the R2 and the regression line in the text box are dynamic and updated automatically with incoming data.
Note that this relationship does not really tell us anything about causality. However, when estimating the relationship for different subsamples, one can see that the Okun’s law coefficient is actually quite stable over time (something you can quickly check in Macrobond yourself using my chart as a blueprint by just clicking on it).
For the unemployment rate to decrease to below 10% again by 2021, we would need a 1.5 to 2 percentage point reduction in the unemployment rate in H2 of this year during every single month! Moreover, for a 10% reduction in unemployment to take place, we would need something like a 15% growth rate of real GDP in the second half of the year (on an annualized basis). Given the size of the current shock, this is not impossible per se, but I do find it rather unlikely, and so do many macroeconomic forecasters.
Note, however, that all of this also rests on the assumption that the historical relationship between GDP growth and unemployment holds up. There is some reason to believe that this will not be the case. The current shock is so large and unprecedented in nature that we are mostly flying blind. And this goes for everybody, Central Banks, fiscal authorities, private forecasters, and even financial markets that are giving us conflicting signals on the economy’s future performance (while stock market is relatively bullish, the bond market is quite bearish). For the record, right now I rather believe the bond market and the Fed funds futures market, which tell me that we will be facing a significant demand shortfall for several years.
The CBO is currently estimating an unemployment rate North of 10% by the end of 2021, which would be a quite horrible outcome and would also mean a significant policy failure on the part of the Fed and the fiscal authorities. Moreover, a lot will also depend to what extent social distancing rules will evolve throughout the year and how this will affect consumption and the service sector. Let’s be real, any economic projections right now are nothing more than wild guesses. Precisely for that reason, policy makers need to assume the worst, act accordingly, and implement expansionary policies in order to prevent some of the harm that the economic meltdown is currently inflicting on the labor market.
Disclaimer: We don’t usually have views and opinions about economic and financial states of affairs, (not ones that we express publicly as a company, anyway). We do believe, however, that people can and do appreciate a variety of perspectives. What you’ve just read is the perspective of the author. While we think our writers are very smart, Macrobond Financial does not expressly endorse the views presented here. And, as the old adage goes, you shouldn’t believe everything you read (not without finding the data, performing a few analyses and presenting it in a nice chart). We want to make it clear that we are not offering this information as investment advice. That being said, if you have Macrobond, you can easily check everything that’s mentioned here, and decide for yourself. If you don’t Macrobond, now you have a great reason to get it.