news article background image
2018-06-07Macro `n Cheese

It’s Now or Never!

This week Roger ponders just what is going on with the US labor market – are wages on the rise, or not?!

Now, over the past couple of weeks, as #UscITA engulfed all other data and events, I have (out of respect for the matchless Caruso) been humming along to some Americanized version of the great Italian theme – O’ Sole Mio – while hammering away at the keyboard. But as the lyrics, ripe with yearning and eagerness, began to sink in, I started to contemplate my own desires (as an economist, of course).


I seek clarity (Did I make a poor career choice or what?). But not in a spiritual or particularly high-minded way. No, I am just tired of the endless, ex-post, discussions and tweaks of the most basic tenets of our economic explanatory models.

 

Do supply and demand decide prices after all?

After yet another look at the US employment and wage inflation nexus, however, I cannot help but think that it’s now or never. That we will soon have answers, real answers.


Why all the unbecoming optimism you might wonder? Well, for starters, the labor markets are currently in a situation of low unemployment rates and at least “some” real earnings growth. The last time this happened was the months and quarters preceding the “Great Recession” in 2007/08. In the graph below I have plotted out how labor markets have gone full circle from low unemployment rate and high real earnings growth in 2007 to a situation with increasing unemployment rates and, initially higher real earnings growth, but later low real earnings growth via decreasing unemployment rates, and back again to the current, 2007-like situation. Here, real wages and/or inflation (nominal wages) should rise in turns, accompanied also by higher productivity growth (i.a., as investments increase to replace expensive labor). Or, at least, this is what we have expected.

 

Admittedly, not all economists agree on what comes first, inflation or wage inflation? Some, mainly practicing economists, believe that a tight labor market makes wage pressures increase, and inflation follows suit as companies try to safeguard their profits. Others, often more academically inclined economists, think that it runs the other way around; in times of high demand, companies hike prices – inflation – and household’s (labor) compensate for higher costs by pushing up wages.


Truth be told, being where we are in the economic and labor market cycle, this shouldn’t make much of a difference. Most measures point to the economy being balanced or even slightly strained and, hence, both the FED and financial markets demonstrate (at varying degrees) a conviction for both higher real wages and inflation.


However, a swift look at the last few data points in the graph above shows that this is not the case, real wage growth has actually stagnated towards ½% y/y. Such developments become downright unfathomable considering that productivity has grown at a rather healthy clip, 1-1½% y/y, over the past few quarters.


Put another way, the stronger productivity growth experienced should give room for stronger real earnings growth, especially as labor markets are becoming gradually more strained with an unemployment rate now below what most analysts deem sustainable in the long-term. Yet, the productivity gains seem to mainly accrue to capital owners in the form of increased profits. How is this possible?

 

I think the short answer is: It’s coming. Most of the discussion above is not completely new (the somewhat higher productivity growth is a welcome change and improvement). The main difference is that there are now more tangible signs of accelerating wages popping up everywhere (yes, slight exaggeration). For instance, Atlanta FED produces various measures of wage growth, i.a., wage growth for job-switchers, which is depicted in the graph above. Clearly, the normal cyclical pattern of pursuing higher paid jobs in economic upswings is alive. The very last outcome demonstrates that while overall median earnings grow at a meagre 3.3% y/y, median earnings growth among job-switchers is a more healthy 4.0% y/y, in line with developments during the last cyclical peak (in 2007).


Such, job-switching behavior is also becoming more widespread, as can be seen in the so called “quit-rate”; the share of employees who resign before getting a new job offer. The quit-rate is now back at the levels preceding the Great Recession. This implies that labor is growing increasingly more optimistic about their labor market prospects and voluntarily choose to resign from (low-paid) jobs in pursuit of “better offers”.


Similarly, from the employers’ perspective, we see that the “openings rate” has surpassed the “hire rate” over the past few months1 . As this denotes a situation where demand for labor is outpacing supply it has drawn a lot of attention from both market pundits as well as media and often put forward as a clear sign of matching problems.

 
1The hire rate is the number of newly employed in relation to the number or employees.
 
 

This is where it starts to get interesting from a monetary policy perspective. I, for one, am not ready to completely give up on the idea that high demand – low supply will lead to higher prices (i.e., higher wages in this case). And if real wage growth indeed accelerates, without companies being able to cut other costs, corporate profit growth will suffer. Likewise, if low (real) wage growth lingers (and inflation remains stable), it suggests lower consumption space, which would of course severely inhibit the FED’s intentions to remove policy accommodation.

 

Luckily, very little points to a retreat in consumption. Consumption growth is healthy, and savings, if anything, are decreasing. This could be seen as a signal of household confidence in future earnings, which is also reflected in both the NY FED and U Mich surveys of households/consumers (see chart above).


To sum up this initial discussion: Companies should start to experience major difficulties in impeding real wage growth soon. Workers are obviously becoming more inclined to quit jobs in search of (i.a.) higher pay, consistent with a continued tightening of monetary policy. Recent and decent productivity growth emphasizes that this path is finally open for workers to exploit. Now, to be sure, this is also embedded in the FED’s forecasts as well as financial market expectations. But it is worth highlighting that it is only recently that THE stars have aligned for this to happen. And if this does not pan out, we are indeed in uncharted territory – It’s now or never!

 

The skills-mismatch discussion

Normally, I would have (and probably should have) ended on that punchy note. However, I am so enjoying this new role where I can use both hands without the risk of getting one severed that I will continue by discussing some ideas that at least partly run counter to the argumentation above. Let’s start with the discussion on matching.

 

If we only look at the private sector overall, it is hard to disagree with the conclusion that good help is hard to find. However, when considering developments within different sectors a different, and somewhat bewildering, pattern emerges.


When I think of matching problems, I normally think of it is as a problem of finding skilled labor, as in well-educated and experienced labor. A problem mainly in more advanced industries, such as those found in the professional and business services sector. That said, if this was all true, a quick glance at the chart above raises some questions.


I would, for example, expect the retail trade, and accommodation and food services sectors to show not only higher differences between hires and openings than the business professional services (reflecting a relative ease to find labor), but perhaps also that the differences would be less volatile (reflecting the more crucial nature of labor in more advanced industries and the ongoing debate of rapid technological advancements).


By and large, business and professional services have indeed demonstrated a smaller difference (in levels) between hires and openings than Retail trade and Accommodation and Food services confirming some of our predispositions. Lately, however, this difference has increased in business and professional services, indicating that it has become easier (!) to fill open positions over the past few months. Furthermore, it is now even higher than in both retail trade, and accommodation and food services (!!), making it hard to square recent developments with the skills mis-match discussion currently filling most economic conversations and media2.


Not only that, but from a longer-term perspective, comparing current developments with those of the last cyclical peak (the 12 months prior to the Great Recession) suggests that the improvement in labor market conditions seen in this expansion has mainly been an effect of increased demand for “unqualified” labor (see also box in above chart) rather than “qualified” labor. Again, when scratching the surface ever so gently, these occurrences underline how hard it is to simply dismiss actual economic developments as a result of skills-mismatch. Together with the decrease in lay-off and discharge rates these elaborations, nonetheless, fit nicely with the low income growth in general and the low productivity growth in particular.

2Of course, there are various sub-categories also in ‘Business and Professional Services’ where developments are “better” (as well as worse).

 

Monopsony or pent-up wage deflation?

Apparently, not all are as convinced that higher earnings growth is just a job-switch away3. Let’s start with the simple observation that for production and non-supervisory employees (graph below) wages have not shown the vigorous rebound in growth that is customary in this phase of the economic cycle. Admittedly, some acceleration is visible towards the end of the time series and hopefully this will continue, but as the next few sections will show, darker forces might also be at play here.

3This section draws heavily on a Kansas City FED note: “Nominal Wage Rigidities and the Future Path of Wage Growth

 

The last labor market data (May 2018) showed that average hourly earnings growth was 2.8% y/y (for production and non-supervisory workers, N.B.). While high for this expansion, it is still below the above 4% y/y that is normally recorded at the current levels of unemployment.


Arguably, relationships between employer and employees are long-lasting. Hence, it might be possible that workers’ wages reflect not only current but also past labor market conditions. For instance, if wages are nominally rigid (to the downside) during a recession, companies could very well try to compensate for this by giving smaller wage increases during the subsequent recovery (implicit contracts). To illustrate, I have plotted the 12-months centred average of the share of individuals with no nominal wage increase relative to the previous year (calculated by SF FED) in the chart below.

 

What is immediately striking is how much the share of workers with zero nominal wages increased during the Great Recession (and that the recovery has been lethargic, at best). In order to more formally gauge how the share of workers with zero nominal wage change impact future wage growth, I have plotted annual wage growth against nominal wage rigidities (lagged one year) in the graph below and estimated the relationship with a simple linear regression. The equation underlying the regression line is shown in the bottom-left box in the chart.

Note: For those of you with access to the Macrobond application, you can of course download the document with statistical tests and assess the relationships yourself. If you don’t have the application, get in touch with us for a demo. I have also performed the calculations using the same controls as in the original KC FED note, but if some of you try different controls and/or get different results, I would love to hear about it.

 

There is an obvious negative relationship between wage growth and nominal wage rigidities. Even when looking at sub-periods the two variables have moved similarly over the two most recent business cycles, with the main difference clearly stemming from sharper dynamics in the period after the great recession.


More recent outcomes, the grey dots, are very much in line with historical experiences and the very last data points are basically on the regression line, indicating that, bar a considerably higher share or workers with zero nominal wage growth, the cyclical patterns of previous recessions remain intact. The coefficient (beta) implies that a one percentage point increase in nominal wage rigidities is associated with a 0.3 percentage point decrease in nominal wage growth a year later. This result is robust to a host of different controls, such as current and past unemployment rates.


Thus, high and persistent levels of nominal wage rigidities explain the lion’s share of the modest wage growth recorded in the recent past. While these results are perhaps not in line with conventional thinking, they go hand in hand with a general decrease in turnover on labor markets, and explain why job-switching behaviour is on the rise. Furthermore, the results suggest that the nominal wage rigidities would have to fall by around 2½ percentage points in a very short period of time to push overall hourly earnings growth up to the cyclical peaks last seen prior to the great recession. Alas, if anything, nominal wage rigidities seem to be on the rise again, suggesting very poor prospects of an acceleration of overall wage growth. Or, for context, after the 2001 recession it took more than two years for nominal wage rigidities to drop 2½%, and I have a hard time finding strong evidence for a swifter improvement this time around.


I’m sure it was a depressing read, but it certainly was a fun topic to explore and write about. However, I still feel somewhat unsatisfied, like something fundamental is missing... What could that be? – I know! Another why! Now, why is it that nominal rigidities shot-up like that, and what can be done to address it? In the heliocentric world of economics most of the debate seems to circle around Krugman’s ideas, but I have long been a fan of this explanation – and it fits nicely with developments of much of the other indicators we have studied here today, not to mention my own, Robert Gordon-esque world-view.

 

Take note

In the last post, on Italy, one of our gifted interns created a Macrobond-template for how to calculate Italy’s debt to GDP-ratio. As a few of our most avid readers noticed there was, nonetheless, a small typo in one of the formulas. We sincerely apologize for any inconvenience this might have caused you. Even though it was swiftly corrected I advise all those who downloaded the MB-document to check if you have the correct version (link).

 

 

 

Get these posts delivered to your inbox

Sign up now to get an email every Friday with the latest research posts from Macro n Cheese

Sign up

 

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 our resident chief economist. While we think he’s very smart, Macrobond Financial does not expressly endorse the views he presents 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 the application you can easily check everything that’s mentioned here, and decide for yourself. If you don’t have the application, now you have a great reason to get it.