The unemployment rate has declined to a low 4.1%, a level lower than 90% of the past 50 years and a level at which the U.S. and most countries worldwide have experienced much higher levels of inflation than we are experiencing today. Where is the Phillips Curve and is it dead? Many economists have been struggling to explain the slow uptake in inflation. Some have suggested that the cause is one of demographic changes while others believe it is related to the slower population growth, among other root causes. But it may be that the most plausible explanation is that it is temporarily being held off by the internet.
In November of 1998, I began to look for a job in advance of my graduation in May 1999. I was looking for a career in finance, and had a good resume which included three summer internships within the industry, a strong GPA, and a solid university degree in Economics. The job market was the hottest in decades since the US was in the midst of the dot com heyday. To find a job, I used my network, the college career center, trade magazines, and newspapers. The internet’s ability to help people find a job at that time was in its infancy. Indeed in 1999, hotjobs.com bought a $1.6 million commercial during Super Bowl XXXIII, but had revenue barely over $2 million. Monster.com wasn’t launched until January 1999 and also had a Super Bowl ad in the same year. It was all just starting. The world was not yet ready to use the internet en masse to help find a job. The “old” ways of finding a job still dominated. Yet pursuing job leads from trade journals, magazines, and newspapers was painstakingly slow and inefficient. Job descriptions were brief, often limited to a handful of lines. Also, a limited number of readers were able to view the advertisement, since they were in a local paper. And because descriptions were brief, potential applicants were unsure of the quality of fit. Listings were only very broadly organized by job trade, and large groups of careers were lumped together. This structure created a limited number of applicants where qualifications were often mismatched with the underlying need. Personally, I probably applied to 20 different positions. Not only was it was laborious, but many were not a good fit for my goals, and I only came to learn this at a later time when I was selected for interviews. From an employer’s vantage point, a relatively limited number of people applied, and the chances of a high-quality applicant match was low.
Economically speaking, this created a lot of friction in the market, which made it much harder and more time consuming for employers to find employees and vice versa. The natural rate of unemployment (NAIRU) at that time was often cited as being between 5 and 5.5%. At this level the economy was thought to be fully employed and lower unemployment would cause inflation to increase more materially due to increased wage pressures. Unemployed labor still existed, it was just harder to find.
Contrast this experience with what might happen today. In 2018, the internet offers tremendous tools to more efficiently match an employer with a potential employee. At Advisors Capital we have experienced this first hand. In the past year we’ve hired for several new positions. One position was a trading role, a position for which we received over 1,400 applicants within a single month. For each of our postings, we received no fewer than 400 applicants within a month, while most postings (if not all) triggered over 200 applicants within one week of its initial posting. Because each listing had a lot of information about the job’s responsibilities, our office, culture, and other attributes, those applying for the position were able to determine how ideal a match it would be for them, further filtering the results. We received this volume of applicants despite an unemployment rate of 4.1%.
The traditional relationship between inflation and unemployment is known as the Phillips Curve.
As the unemployment rate declines, wage inflation increases as employers offer higher levels of compensation to attract labor. The high levels of friction of 20 years ago, however, spread out this inflationary pressure. The labor supply at an unemployment rate of 5.5% still had excess labor; it was just harder to find the right candidates. Thus the tradeoff between unemployment and wages (and so inflation) was more gradual. In the internet age, this relationship has probably become less gradual. Employers can find candidates relatively easily until the labor supply is very low, by which time almost nobody is left unemployed. Once the economy reaches this point, wage pressures could build quickly as employers attempt to fill the vacant seats and offer higher levels of compensation in order to attract them from other employers.
While this thesis is purely hypothetical, its implications for coming wage and inflation growth warrant careful attention, especially if the unemployment rate declines further. The U.S. economy is still absorbing about 100,000 net job searchers per month, which amounts to over one million per year. While this pace would not historically be a cause for rapid inflation, it may suggest a growing risk for higher inflation than most are anticipating, and it could happen quickly. The Fed would be unable to respond fast enough.
While the data has not yet traced out this sharper tradeoff—and we are not yet seeing indications of a rapid burst in inflation—it is a risk to watch in the coming 12 months. If it appears, blame the internet.