Noah Durst is an Assistant Professor in the School of Planning, Design and Construction’s Urban & Regional Planning Program at Michigan State University. On September 11, 2019, he spoke at an IPPSR Public Policy Forum entitled “Census 2020, Possible Outcomes for Michigan.” His presentation was on the implications of an accurate count on research, policymaking, and funding.
On September 16, Tjolsen conducted a follow-up interview with Durst about the Public Policy Forum. A lightly edited transcript of their conversation follows:
Tjolsen: Much of your presentation at the IPPSR Public Policy Forum was about how the census affects federal funding. Why is the census so important for funding? And does it only affect federal funding, or can it have effects on state and local budgets as well?
Durst: Estimating how exactly the census affects funding can be complicated, but it only directly affects federal funding. It does impact state and local budgets more indirectly as federal dollars are often given to states in the form of block grants, which states can then distribute to local governments.
Tjolsen: You mentioned in your presentation that Detroit and the surrounding areas are the most at risk of being undercounted in Michigan, why is this?
Durst: I don’t know why exactly this is in a causal sense, but when you look at ROAM [the Response Outreach Area Mapper, run by the Census Bureau], in Michigan it is primarily urban areas that are predicted to be undercounted by the census bureau. So, Detroit is largely undercounted at least partly because it is one of the largest cities in the state. Similar patterns can be seen in other major cities. In 2010, the census also found that communities of color and renters were largely undercounted. Detroit is a city with many renters and many low-income and black residents, so this is another reason it is largely undercounted.
Tjolsen: You also mentioned that some of your research looks at gerrymandering and that racial gerrymandering is impossible to document without census data. But would it be correct to say that gerrymandering can still occur without that data? And how does having access to the data help to combat it?
Durst: This is a really great question. I may have to think about that one. Yes, my work looks primarily at gerrymandering of city borders. The expansion of city borders happens with municipal annexation. The evidence suggests that race is an important factor in how cities draw borders.
In regard to the gerrymandering of electoral districts, every 10 years states draw new electoral districts through redistricting. Recently, states have used more advanced statistical approaches that rely on census data such as race in order to favor one party over another. Efforts to gerrymander based on race are always done, but precise data makes it easier. I would say that manipulation of boundaries can happen with or without this precise data, but an inaccurate count will not prevent gerrymandering. An accurate count is important for so many other reasons, regardless of how it may affect gerrymandering.
Tjolsen: A number of other panelists also mentioned the importance of all children being counted on the census. What do you think can be done to ensure this and what are the potential impacts of children being undercounted, in terms of funding?
Durst: I don’t know specifically how an undercount may affect children. It probably depends on how certain federal programs may count the number of children into their funding, such as a health and human services program. However, I am not sure if any such program does count children in this way. Each program has a different formula on how it allocates funding. If this is something of interest, I would recommend looking at the George Washington Institute of Public Policy (where I got much of my data). They have the best dollar estimates I have found when it comes to how an undercount may affect funding.
I don’t know what can be done specifically to help count them, but I would say that relying on trusted community leaders is the best way. Getting leaders that are already active and organizing in a community to advocate for a complete count is the best way to achieve it. This is especially true with the Trump administration’s effort to include a citizenship question on the census. There is now understandable doubt within certain communities about whether to feel comfortable completing the census. Academics can help by conducting research and engaging with legislators, but community leaders can really make the biggest difference.
Tjolsen: Lastly, another panelist mentioned on Wednesday that the census is required in the Constitution and has been used since the beginning of the United States. Why do you think the founding fathers found it necessary that each and every living person in this country be counted and are you glad that they made that decision and we have a census today?
Durst: Fundamentally, the census is the means by which the nation’s population is counted for the purpose of apportioning congressional seats and redrawing congressional districts. The census also provides the most accurate count and basic description of everyone in the country. Academics rely on other data that the census bureau creates beyond the decennial census – such as the American Community Survey, or ACS, but these are largely surveys that are not as accurate. Unlike the census, these surveys are random samples that have confidence intervals and are prone to measurement error. The census of course still has measurement error as well, but those errors are much smaller, and the census is still the most accurate measurement that we have. The census is central to much of the research that academics do. For example, my work on city boundaries would not be possible without accurate census data. In large part, this is true because data from the decennial census is available at a smaller geography than ACS data. They are therefore very useful for measuring small changes in cities’ borders.