By Brian Points
As the United States continues to fight its way out of the Great Recession, more attention has been directed to the question of why is has taken so long for workers to find re-employment. In economist parlance, this is primarily a question of “structural unemployment.” This describes the type of unemployment that results from a mismatch of worker skills and the skills demanded by employers.
As of April 2011, there were 13.2 million unemployed workers and 2.9 million unfilled job openings. In other words, April’s bulky 8.7% unemployment rate could have been lowered one percentage point (to 7.7%) if just half of the advertised job vacancies were filled by unemployed workers. Obviously, it is not realistic for every position to be filled immediately—it takes time for employers to find the right workers, and vice versa. But the odd pairing of high unemployment and high job vacancies illustrates a structural employment issue, which may have worsened in recent years.
Historically, when the economy is growing, the unemployment rate is relatively low and the job vacancy rate is relatively high, indicating more job openings than there are workers to fill those positions. Likewise, when the economy is shrinking, the unemployment rate is relatively high and the vacancy rate is relatively low, because there are more workers looking for work than there are jobs. This pattern held between 2008 and mid-2009 but from the second half of 2009 through mid-2011, the vacancy rate has remained surprisingly high when compared to the unemployment rate. [1. See page 5 of the Bureau of Labor Statistics’ Job Openings and Labor Turnover highlights from May 2001 for an up-to-date illustration of this relationship: http://www.bls.gov/web/jolts/jlt_labstatgraphs.pdf). It is also worth mentioning that some economists point to the special extension of unemployment benefits that occurred during the recession as a contributing factor to unexpectedly high unemployment.]
A question that has perplexed jobseekers and economists alike is how there can be so many people looking for work and yet so many unfilled positions in the economy? In an attempt to answer this question, EMSI has taken a fresh look at the skill gap issue using historic jobs and earnings data to determine which segments of the labor market are growing and which are diminishing. Often when examining shifts in the labor market, analysts will look solely at employment changes and highlight the occupations that have increased or declined in total employment, but we believe this is somewhat shortsighted. This method may not tell the whole story. For example, it is possible for employer demand for a certain occupation to increase or remain the same while actual employment levels drop. [2. This situation occurs when either a) supply drops due to workers’ unionization or the advent of new worker certification requirements that did not previously exist, and/or b) the skills for a certain job category become so specific and technical that only a select group of workers can perform them.] Therefore, the addition of the earnings measurement over time adds a great deal to this analysis.
In order to describe this method, a bit of basic economic theory needs to be explained. One of the chief tenants of economics is that in a market that is not exceedingly manipulated by outside forces, demand and supply will meet at a point that is mutually beneficial for both producers and consumers. To put this in labor market economics terms, producers are individuals offering their time and labor for a wage, and consumers are employers seeking the labor of workers in exchange for a wage. The magical meeting place where both groups settle is called “market equilibrium.” Although both parties may not be completely satisfied with the arrangement, they are at least content enough to accept the terms of employment.
Aggregating the data shows that of all occupations in the potential skills shortages category, 66% are in the fields of healthcare; education; business and finance; and architecture and engineering. Conversely, of all occupations in the potential surpluses category, 63% are in the fields of production; construction and extraction; and installation, maintenance and repair.
Following this theory, we can expect that any given occupational category (SOC code) will have a wage and employment level that best represents the demand for workers, and the required compensation level for employees. [3. With this theory, it is assumed that each group of workers is “homogenous.” In other words, no one worker in any occupational category is more knowledgeable or skilled than any other. Of course, in the real labor market some workers are much more capable than others. In such cases, the higher skilled and lower skilled workers each belong to their own occupational groups, which have their own market equilibrium points with different wage and employment levels.] To complicate matters, the economy is never stationary but is in a continuous state of adjustment and realignment. Although the market for certain workers may be at equilibrium in the fourth quarter of one year, changes influencing supply and demand will likely cause that equilibrium to shift such that the equilibrium will be different in the first quarter of the following year. (Think of the demand for commercial fishermen in the Gulf Coast before and after Hurricane Katrina in 2005).
Based on these theories, EMSI has dug into historic labor market data to look for two characteristic groups:
- Potential skill shortages: where employer demand had pushed both employment and earnings upward over time
- Potential skill surpluses: where worker availability has exceeded demand and pushed both employment and earnings down over time
The key word in both of these categories is “potential.” These shortage/surplus measurements are, in fact, only half of the equation. A “skill shortage” only exists if workers have failed to acquire the requisite skills to perform the required tasks at a rate equal to demand. Likewise, a “skill surplus” exists only if workers have failed to retrain and find employment elsewhere after losing their jobs. Both of these measurements are difficult to pin down. In the next post, we will examine whether the potential shortage/surplus occupations have received the requisite amount of workers over the past couple of years, but for the moment it will suffice to examine these increases and decreases in demand over time.
To perform this analysis, EMSI analyzed 661 SOC codes in terms of jobs and earnings between 1999 and 2010. In order to get the data to line up properly, self-employed workers and every SOC code that ends with “all other” have been excluded. An occupation appears in the potential shortage category if the wage and employment growth between 1999 and 2010 have exceeded the average by a significant degree; and an occupation classifies in the surpluses category if both wages and employment have decreased by a significant degree. [4. All wages are in real terms, adjusted for inflation to 2010 dollars. The cut-off point for significance is 0.5 standard deviations from the median. Please be aware: we are not treating this as a standard econometric model in which we are attempting to show a consistent relationship between earnings change and employment change. For this reason, we are not utilizing the same measures for statistical significance that are common to econometric models.]
Tables 1 and 2 show the results of this analysis. These tables are ranked by employment in 2010 to provide some gauge of the significance of the potential shortage or surplus. Percent change in employment and percent change in earnings are also shown in these tables.
Table 1: Top 25 Occupations Facing Potential Skill Shortages
[table id=108 /]
Table 2: Top 25 Occupations Facing Potential Skill Surpluses
[table id=109 /]
So what can be gleaned from this analysis? To start at the highest level, this certainly indicates employers’ preferences are shifting away from manual labor occupations and toward knowledge-based occupations. Aggregating the data shows that of all occupations in the potential skills shortages category, 66% are in the fields of healthcare; education; business and finance; and architecture and engineering. Conversely, of all occupations in the potential surpluses category, 63% are in the fields of production; construction and extraction; and installation, maintenance and repair.
To examine some more specific cases, it is interesting that two of the occupations regularly at the center of skill-shortage discussions, registered nurses and accountants, are at the top of this list. (We must emphasize again that this does not indicate that there is a skills shortage for these occupations but rather that the demand for such workers has increased at a rapid rate over the past 11 years; whether or not the output of students has remained apace with this demand will be explored in the next piece.) It is also not surprising that 10 other healthcare positions land on this list, including occupations such as medical managers, pharmacists, and speech-language pathologists.
There are also some surprises on this list, such as the contingent of occupations in the business and financial operations category (e.g., loan officers, claims adjusters, and financial analysts). The prevailing theme with these occupations is that each requires individuals with strong interpersonal skills, as well as strong computational and analytical skills. Over the past decade, both the increase in the rate of information sharing and increased complexity of this information can likely explain why employers have been investing higher wages in these workers.
Management analysts, for example, experienced a wage increase of 11% and employment increase of 78% over the past decade. Their presence on this list highlights the importance of technology in creating job change, as well as changes in business trends. In the past decade, businesses in the professional and technical services sectors have been increasingly hiring businesses and consultants from outside of their own companies to handle departmental work such as advertising, payroll, and human resources. We can account this change, in large part, to the power of technology to move information quickly and efficiently.
Many occupations in the manufacturing category have declined sharply in both wages and employment due to offshoring. On this point, we must specifically state that manufacturing skills are not declining on the global scale. Looking worldwide, there are likely more individuals working as industrial truck and tractor operators and tool and die makers in 2010 than there were in 1999, but today many of these positions are now in developing countries. These reflect situations where without the effect of protectionist policies (such as quotas or tariffs) foreign competition has a competitive advantage over American workers because foreign workers are willing to work for lower wages.
Offshoring is not the only reason that occupations on this list have declined. Just as with the potential surpluses list, technology is the catalyst for many notable changes. Occupations such as stock clerks and order fillers have become less valued in the labor market due to labor-saving technology that efficiently catalogs inventory and computer programs that allow people to make orders for equipment and merchandise without the aid of a middle-man. Likewise, positions such as telephone operators and desktop publishers are quickly becoming obsolete due to advancements that have made telecommunications more accessible for a wider audience.
The large cohort of construction jobs on this list are a consequence of the precipitous drop in construction employment between 2007 and 2010, and these may or may not represent an actual skill surplus. For example, employment for carpenters had increased every year between 1999 and 2007, but between 2008 and 2010 employment decreased by an average of 14% per year; indicating that this may represent a temporary, or cyclical change. On the other hand, wages consistently decreased for all construction jobs by about 0.5% per year over the last decade. This could indicate that a sustained oversupply issue among construction occupations has allowed employers to pay workers slightly less for their labor. Time will tell whether there are too many or just the right number of people in the workforce with construction skills, but it is difficult to say right now.
The dynamic nature of the economy causes routine changes in labor market demand. These data illustrate an important and often overlooked fact: the labor market is driven by all other markets (e.g., markets for cellular phones, houses, and doctor’s office visits, etc.). Over time, we can see labor market changes occurring, for instance, when the number of product orders conducted over the internet increase because there are jobs required to support that increase. At the same time, there are jobs that will be lost because they are no longer the most efficient way to address consumer demand. It is easy to see how skills shortages naturally arise in a market-based economy. When such changes occur, it is imperative that public education, the workforce system, and economic development agencies are able to cope with the changes, and assist workers in the process of moving from areas of skill surplus into areas of skill shortage.
In the next blog post we will analyze these potential skill gaps from the supply perspective to see whether or the supply of talent has grown at the same rate as the demand for the workers identified here. We will also analyze the knowledge, skills, and abilities (KSAs) that are incumbent to the potential skill shortage occupations in order to see which KSAs could be undersupplied in the labor market.
Illustration by Mark Beauchamp