I chose to look at the dataset of Principal Employers (Non-Government) in the County of Los Angeles, which compares the top top 10 employers in the county in 2005 and 2014. The data types consist of year, name of employing entity, rank, number of employees, and percentage of total count (of all employees in the county). A record consists of the relevant data associated with the individual employers, along with, for comparative purposes, the aggregate information for all other county employers.
While Wallack and Srinavasan do not define ontology as such in their paper, it appears that what they mean by the term is the philosophical underpinning of the way in which real-world events are labeled, categorized, and interpreted (2009:1). In the case of this dataset, then, the ontology appears to derive from a big-business standpoint, in which large entities–whether corporations, universities, or health care providers–employ masses of individuals to provide services or sell things to other individuals. It’s an ontology based in capitalism, and implicitly, as the title indicates, in a bigger-is-better mentality. It also seems to grow out of the the idea of a “company town,” where there are a few big employers and then a host of smaller companies/industries that support them. But given that this is a dataset produced by a local government, the ontology may also reflect the need for the provision of public services: how do all these employees get to work, where are they going to need to live to have reasonable commutes, how much water, electricity, and other utilities are these large centers of commerce going to require to operate? Also, it may be presumed that companies that employ large numbers of people are equally out-sized in terms of their tax contribution to the county coffers.
If number of employees is a sign of a successful business, then this dataset indicates not only what are the most successful businesses in Los Angeles County, but what kind of businesses are successful: this is a set that skews toward a service economy, not a production economy. Although there are only two time periods represented, you can see that, while most of the names remain the same, it’s probably significant that AT&T and Vons, which were on the 2005 list, do not appear on the 2014 list, and are replaced by Home Depot and Providence Health, which probably reflects larger trends in their respective industries as much as it represents the rise and fall of individual companies. However, it’s also significant that these 10 employers constitute only about 4% of all employers in the county, as the “All Other Employers” category represents just under 96% of employers in both years.
This does lead you to wonder why “number of employees” has been chosen to represent–what? Sheer size? Whoever “owns” the most working bodies wins? Is bigger better? The inclusion of USC and Cedars-Sinai on both lists may indicate a need for public transportation to help get employees to and from a limited number of work sites, but there are Home Depot, Ralphs, and Target stores all over the county. Also, there is no breakdown of what kind of employees these are. Boeing and Northrup appear on both lists, but are these manufacturing jobs or management? If this dataset stretched over many more decades, what kinds of trends would we see emerge? Would the big movie studios of the 1920s-1950s appear in the top 10? Would the top 10 employers constitute a larger percentage of the employer pool at different points in time?
If I were starting from scratch, I would be inclined to categorize employees by what kind of work they do rather than who they work for–not only larger categories of industries–health, education, sales, manufacturing–but also job types–for instance, another dataset, Gender Breakdown of City Workers by Category, includes categories like paraprofessionals, technicians, protective services, skilled craft, etc. Given that the top 10 only constitute 4% of employers, this would shift the ontology from a bigger-is-better mindset to a reflection of what citizens are actually doing during their workday. It would also better incorporate data from smaller companies, individual entrepreneurs, and freelancers in an increasingly fragmented economy.