Last week I was at the annual meeting of the Association of Computer-based Systems for Career Information. This year, for the first time, the meeting was held jointly with America’s Career Resource Network Association, and the two organizations now are talking about a possible merger. The ACSCI meeting is always fascinating to me because it is all about the latest developments in career information.
One of the highlights of the meeting was a presentation by Brian N. Rae of the Alaska Department of Labor and Workforce Development. (A PowerPoint slide show similar to what he showed at the meeting is available here.) The Research and Analysis office there has done a lot of number-crunching in anticipation of a massive pipeline construction project that has the goal of connecting Alaska’s abundant reserves of natural gas to the pipeline system in the lower 48. Although Governor Sarah Palin believes that the pipeline is “God’s will,” most of us expect that some human participation will be required to accomplish this project, and to that end the Research and Analysis (R&A) office is now assessing the readiness of the state’s labor market.
The R&A analysts realized that the availability of workers to fill certain needed occupations depends on more than simply how many workers are currently in that occupation, plus how many can be expected to graduate soon with the appropriate educational credentials. A third source of recruits is occupational mobility–the workers who are poised to enter the needed occupation from another occupation. Of course, we’re rarely going to find your typical bricklayer stepping right into a position as a civil engineer (or vice versa). Therefore, to estimate the pool of workers available via occupational mobility, the R&A office needed to learn what career moves actually do take place. In other words, they had to construct career ladders.
The R&A analysts had access to a rich set of data. The state’s Unemployment Insurance system had records for every working person for the years 2001-06. For each quarter during that time span, they had each worker’s Social Security number, occupation (in the Standard Occupational Classification taxonomy), and earnings. If records for two consecutive quarters shared a Social Security number but had a different SOC code, that meant a shift of occupations. By examining a large number of records, the analysts could identify shifts that were so typical that they represented job ladders.
In most cases, a ladder was thought to exist if the shift from occupation A to B was frequent; if workers in A rarely shifted to occupations other than B; if workers in B rarely arrived from occupations other than A; and if workers in A rarely arrived from B. The analysts got a measure of how far apart A and B were hierarchically by comparing the estimated wages the two occupations, as well as the typical requirements for education and work experience. The analysts did not rely totally on statistical patterns; they also exercised judgment, considering the relationship of the job titles in the SOC taxonomy and the shared occupational characteristics reported in the O*NET database, among other factors. Based on all of this evidence, the R&A office was not only able to identify career ladders, but also made judgments about the strength of the links between occupations.
You can see the fruits of their labors at the Alaska Career Ladder Home Page. There, you can choose an occupation from an alphabetical listing and see where it falls in a career ladder. Then you can click on an occupation above or below it in the ladder to see what’s above or below that occupation. Each page for an occupation has links to the SOC definition of the occupation and to information about it on the O*NET Online site. A note warns you that you do not necessarily have to progress through each rung of the ladder to move upward; sometimes it is possible to skip a rung to reach a target occupation.
The Alaska Career Ladder is a valuable resource, but it does have certain limitations, most notably the small size of the workforce that provided the data. The working population of Alaska is only about 300,000 people, roughly equivalent to that of the Greenville, South Carolina, metropolitan area. In addition, the mix of industries in Alaska is not very representative of the mix in the other 49 states. Other states should follow Alaska’s lead and conduct their own career-ladder studies. Not all states have access to the same data topics that were available to Alaska, but I hope that those states that do will replicate Alaska’s pioneering work.