Many DC neighborhoods have undergone massive growth in businesses over the past few decades, but certain areas of the city remain under-resourced and underserved, as we discussed in our last blog post. To build a vibrant economy, DC policies must encourage development that includes all of the District’s neighborhoods and residents. In addition, neighborhood changes should be monitored to ensure that long-term residents are benefiting from growth rather than displaced by it.

But good policy is built on good data—and that’s what we’re missing.

The data that researchers and policymakers have about local businesses is poorly categorized. Most business data is based on NAICS codes, the system used by federal agencies to classify businesses. But the way businesses are grouped and defined makes it hard to examine them in detail. In addition, there is no clear incentive for businesses to categorize themselves correctly (or at all), so the data are unreliable.

For example, policymakers might want to know which neighborhoods have a surplus of payday lenders and check-cashing services, but the category for these businesses also includes travelers’ check services and more mainstream providers, making it hard to get a clear picture of where alternative financial services are located.

Without better data, the DC government can’t answer questions related to economic development and equity, such as whether tax incentives are successful at encouraging businesses to locate in certain areas, which types of businesses are struggling where, or—as we tried to determine—how businesses change when neighborhoods gentrify.

Business changes in Columbia Heights, Shaw, and Anacostia

The Columbia Heights and Shaw neighborhoods in DC underwent dramatic changes during the late 1990s and the 2000s, including increased property values and demographic shifts often associated with gentrification. By comparison, Historic Anacostia, east of the river, remained fairly unchanged during this period.

Gentrification in the Columbia Heights neighborhood cluster (which includes Mount Pleasant, Pleasant Plains, and Park View) and in the Shaw-Logan Circle area brought with it an increase in the total number of businesses. Although Historic Anacostia did experience some growth in business, this growth lagged behind the city as a whole and particularly far behind gentrified neighborhoods like Columbia Heights and Shaw.

Understanding what types of businesses entered and exited these neighborhoods during this period is more difficult. For instance, we might have expected full-service restaurants  (or sit-down restaurants) to increase in the Columbia Heights and Shaw neighborhood clusters during the late 90s and 2000s and for limited-service restaurants (or counter-service restaurants) to decrease. However, this was not the case.

              
These trends don’t match predictions because the most disaggregated category for limited-services restaurants in the NAICS system does not differentiate between high-end cafes and fast food joints,  making it impossible to tell whether changes are fueled by businesses like Starbucks or by businesses like McDonalds. Anecdotally, it’s likely that the spike in limited-service restaurants in Columbia Heights and Shaw in the early 2000s was due to the increasing popularity of coffee shops around that time, but the data aren’t clear. Similar difficulties arise when attempting to parse out changes in things like fitness centers and liquor stores.

NAICS was not designed to inform local economic policies

The NAICS coding system presents other challenges as well. For one thing, businesses are required to self-designate into a NAICS category and they can usually only pick one category.

Also, it is often difficult for businesses to figure out exactly which NAICS code they fit into—if they fit into just one—and there’s no real incentive to choose correctly unless there is a tax incentive to do so, like DC’s supermarket tax credit.

The NAICS system was developed for the collection and publication of data on the United States as a whole and was not intended to meet the needs of procurement and/or regulatory applications. This makes it difficult for cities to use NAICS data to monitor and track businesses over time for the purposes of local policy.

What needs to be done to improve our data on businesses?

To fix this problem, businesses need help with, and incentives for, categorizing themselves into the correct business codes. And they should have the option to choose more than one category, at least in city databases. Smaller, more precise categories are also needed, so that cities can monitor changes in specific types of businesses over time.

Finally, city agencies should work together to share and improve the data that they already collect on businesses, such as licensing, unemployment insurance, and tax revenue data. In addition, the data collection procedures for these datasets should be augmented over time to improve the ability of cross agency analysis and planning.

Good data on businesses is necessary for strategic economic development and planning. Increased equity throughout the District will not be realized until policymakers and practitioners have the tools they need to monitor and evaluate changes in businesses over time.