In my last blog, I introduced social network analysis as a tool for better understanding your local economy and visualizing data in new ways. For the second installment, social network analysis (SNA) will be used to analyze Occupational Employment Statistics published by the Bureau of Labor Statistics. While the following data is decidedly not a social network, it does demonstrate the use of SNA as a tool for visualization.
A brief note on methodology. The Bureau of Labor Statistics defines each two-digit NAICS sector by the occupations that compose it at the national level. The following supposes that the definitions act as a connection between an occupation and the twenty two-digit NAICS sectors, allowing the use of network analysis to analyze the relationships between occupations and business sectors.
The problem: Suppose you are a college freshman who is a little distracted. Things catch your interest but you tire after a given subject after a few years. You’re at your guidance counselor’s office attempting to choose a career path (for the sake of the blog lets define career path as an occupation). Now, you’re fairly certain that whatever occupation you choose you’ll want to work in a variety of sectors throughout your professional career rather than being stuck in, say, healthcare till retirement. By using social network analysis, we can identify which occupations share overlap with the most sectors thereby giving you a list of occupations from which to choose.
The above graph is the preliminary result of the occupational network analysis. At first: chaos. However, with the help of a little highlighting, the semblance of a pattern emerges. The yellow dots (called vertices) represent the twenty NAICS sectors. The gray lines (edges) represent a connection between an occupation and a business sector. Thus, the edges are nothing more than a graphical representation of the BLS’s occupation by industry definitions. Near the center of the graph, proximate to the yellow vertices, are orange vertices which represent the most connected occupations. That is to say, these occupations appear on each of the twenty two-digit sectors. In contrast, the red dots at the graph’s periphery are those least connected, or rather those that share only a single sector in common. For organizations sake, let us group the occupations by their respective occupational categories.
The above graph displays the exact same information as the first graph, only that the vertices are now grouped by their respective major occupational groups. The black dots to the lower right hand center represent the twenty NAICS sectors. Each of the other symbols represents a single occupation.
Now with the housekeeping out of the way, let’s revisit the college student. Now, you’d like to know which occupations are connected to the most number of sectors in hopes that you can hop around to different sectors throughout your career.
The above graph filters out all of the occupations that appear in 19 or less sectors of the economy, leaving only those occupations that appear in all business sectors. Now, which occupation should you choose? You should undoubtedly be a chief executive. They are connected to all twenty sectors and make bank, to the tune of $166,910 a year (annual median wage). Jesting aside, the list of occupations is quite revealing and may signal (as the title of the graph alludes to) those occupations which are more resilient than other occupations assuming, of course, that we define resilience as any occupation in which an individual can transition from one sector of the economy to another. This is easier said than done since many sectors require specialized skills that are pertinent solely to that sector. The chief executive of an IT firm may not have the skills or tacit knowledge required to manage a chemical company.
The list of resilient occupations, however, is revealing. The majority of the “resilient” occupations serve office and administrative support functions followed by sales and related occupations, transportation and material moving, management, and business and financial operations. More specifically are occupations you’d probably guess right off the bat: accountants and auditors; database administrators; bookkeeping, accounting, and auditing clerks; and general maintenance and repair workers. Of course, these occupations may, or may not prove to be resilient as technology evolves, labor market dynamics shift, or larger structural shifts occur throughout all industry sectors. But, for you the college student, any of these occupations would potentially allow you to shift between sectors during your career, provided you learn the specific skills necessary to succeed in a new business sector. Finally, you can filter the edges displaying career paths whose average annual wage is $55,000 or more.
Now, along with a list of “resilient” occupations, we can also develop a list of occupations that are less resilient. These occupations are dependent on a single sector of the economy.
These occupations are potentially susceptible to shocks. If the sector were to suddenly vanish, than the worker would certainly require retraining or be forced to leave the labor market. As displayed in the graph, the least resilient occupations are education, training, and library occupations and, specifically, post-secondary educators. Admittedly, this deserves a bit of a laugh considering the unquestioning importance we place on educational attainment. However, given the diminishing affordability of higher education and state budget cuts, educators are not immune to layoffs. Other less resilient occupations include: farm labor contractors, orthodontists, fish and game wardens, post office clerks, mine cutting and channeling machine operators, shoe machine operators and tenders, and, interestingly, embalmers, among others.
Social network analysis, or in this case, network analysis has allowed us to organize occupational employment data in such a way as to obtain a general sense of the resiliency of occupations in the national economy. Combined with more qualitative data along with perhaps survey data, network analysis could allow practitioners to identify commonalities between different occupational categories and economic sectors. So, if say your community has a large displaced worker population, you may easier assess which sectors the displaced worker could transition their skills. Once again, it requires a qualitative component, but as a tool for visualizing and organizing these connections, network analysis will serve you well.
