Census data for downtown Wichita workers

Is the presentation of the number of workers in downtown Wichita an innocent mistake, mere incompetence, or a willful lie?

There’s a question regarding how many people work in downtown Wichita, the Wichita Eagle reports.¹ Other sources have noticed a discrepancy.²

At issue is the meaning of “working” in a certain location. Data that the Center for Economic Development and Business Research at Wichita State University supplied to the Wichita Downtown Development Corporation indicates about 26,000 people work in downtown Wichita, for these purposes defined as zip code 67202. This number is used in a wide variety of ways, including in Wichita city budgets and federal grants made by the city.

It’s appropriate, then, to understand what the 26,000 number means. The Eagle article mentions “a likely mistake in how the number of jobs downtown is calculated.”³ The same article quotes Jeremy Hill, director of CEDBR, as saying, “It looks very obvious and plausible that it is an error.”

There is no “mistake” or “error” in this Census data, which is known as LEHD Origin-Destination Employment Statistics, or LODES. But we need to be curious or cautious enough to investigate what this data means. Documentation from the Census Bureau for LODES data gives the definition of the place of work and a cautionary note: “A place of work is defined by the physical or mailing address reported by employers in the QCEW (formerly ES-202) or Multiple Worksite Reports. An address from administrative data may or may not be the actual location that a worker reports to most often.”

The Census Bureau continues with another warning regarding this data: “Nonreporting of multiple worksites is especially common with state and local governments and school districts. In such a case, LEHD infrastructure files assign all workers for that employer (within the state) to the main address provided.”⁴

In the case of downtown Wichita, the mistake was made in the application of this data, which is the claim that there are 26,000 workers in downtown Wichita. There may be that many people who draw a paycheck from an administrative office located in downtown. But large numbers of these don’t come to downtown to perform their jobs.

The LODES data reports a one square block in downtown that holds 7,740 workers. This is the block that holds the administrative office building for the Wichita public school district. Regarding this, the Eagle article reports: “One of the most likely reasons for the difference, according to multiple local academics, including Hill, is that the Census is reporting that every employee for USD 259 works downtown. Most USD 259 employees work in buildings across the city, but the central office is located downtown.” This is something the Census Bureau warns users to consider.

There’s another area of erroneous application, too, and it isn’t mentioned in the Eagle article. This concerns the second largest concentration of workers in downtown Wichita (according to the LODES data) in a Census block which has 3,437 employees. This is the block that holds Wichita city hall. In 2014 the city had 3,270 employees. But they don’t all work at Main and Central. They’re dispersed throughout the city in police stations, fire stations, and other sites.

How was this missed?

Nearby is an example of using the Census OnTheMap application.⁵ This is the source of LODES data that the WDDC cites in its footnotes to its annual report. When using the application for zip code 67202, there are two — and only two — large dark blue dots. These represent the census blocks with the greatest number of workers, 7,740 and 3,437. I’d like to think that if someone at CEDBR, WDDC, or city hall looked at this map and saw those two big blue dots, they might ask a few questions. Wasn’t someone curious as to how a single block of downtown Wichita manages to hold so many employees? Which companies do they work for? What can we learn from the success of these companies that employ so many people? Can we duplicate this success in other parts of downtown?

But I don’t think anyone asked these questions. No one — not at CEDBR, WDDC, or city hall — was inquisitive enough to really look at this data and see what it means. It’s either that or there was a willful misrepresentation.

The Eagle article also reports this: “This won’t make much of a difference to most businesses downtown, according to Hill. They already know how big the market is because they have experience with it. … The best companies will look at census data when coming up with their business plans, Hill said, but every business relies on several numbers, so even if there are thousands of fewer jobs downtown than previously thought, it’s unlikely that it would have much of an impact.”

On these remarks, I would say that first, we’re trying to recruit new businesses to downtown Wichita. It’s those business firms that this data speaks to. While the “best” companies may use other sources of data, I don’t think we want to discriminate. All companies are welcome to Wichita, I hope.

Second, Hill says companies “will look at census data.” Well, this is census data.

Third, Hill says this mistake won’t have “much of an impact.” In the future, I think we’ll need to ask CEDBR, WDDC, and city hall if the data they supply is intended to have an impact, or is it for something else.

Fourth, there is other census data. The United States Census Bureau tracks business data by zip code.⁶ The data that is available includes the number of business establishments, the number of employees, and the annual payroll, expressed in thousands of dollars not adjusted for inflation. It includes private-sector workers only, so it does not count all workers.

Nearby are results for zip code 67202. For 2015 the number of jobs is 13,581, not much more than half of what city leaders have told us. Again, these are private-sector workers only.⁷

Not only are these numbers much smaller, the results since 2007 show fewer business establishments, fewer people working downtown, and lower earnings generated in downtown Wichita. In all cases, the trend is lower. The LODES data is on a downwards trend, too.

—
Notes