You have permission to edit this article.
Opinion/Commentary: In guarding privacy, census could badly distort data

Opinion/Commentary: In guarding privacy, census could badly distort data

  • 0
Only $5 for 5 months

As an immigrant, a proud naturalized American citizen and a professional demographer, I feared the citizenship question proposed for the upcoming census would discourage participation. Fortunately, the question has been scrapped and will not be on the census.

Unfortunately, another new approach by the U.S. Census Bureau presents an even bigger concern. If implemented, this new approach will make the 2020 census the first in which the actual counts for counties, cities and towns will be distorted before they are released.

Like a family portrait, the decennial census is a snapshot of the nation and of every community, big or small. The census has always reported the number of people counted in each state, county, city and town, as well as the average household size, the residential vacancy rate and the number of people by age, gender, race and ethnicity.

Not anymore. Not if the Census Bureau implements a new “differential privacy” procedure that will distort 2020 census results for counties, cities and towns.

Why? The bureau believes data distortion prevents reconstruction of individual records including age, gender, race and homeownership, even though that basic information already is easily accessible through the internet.

The bureau’s proposal is well-intended, but the consequences are disastrous.

If the Census Bureau proceeds with this plan, we no longer will have accurate information about our communities. The data distortion might misrepresent a city’s population size by 25% or more, or in the case of an age group (such as 5- to 9-year-olds) by more than 100%.

As a result, school boards might not hire enough teachers; emergency services might delay purchasing additional but needed vehicles; and businesses won’t know if enough people of working age are in an area to justify a new location in the community. And these are just a few examples.

To demonstrate the impact of the new procedure, the Census Bureau provided sample data that distorted, as an example, the number of girls ages 15 to 19 in the city of Emporia from the actual count of 185 girls in 2010 to only 30. When you put an inaccurate number like 30 into a formula, bad things happen.

There were 20 teen births recorded that year in Emporia. Twenty teen births among 30 girls (30 being the distorted number) generates a distorted increase in the teen pregnancy rate from 10% to 66% — two out of every three girls. This in turn could lead to misguided, unnecessary interventions and irresponsible use of taxpayer dollars — all because of inaccurate data.

The bureau’s plans will diminish, if not eliminate, the value of census data to local charities, schools, employers, road and home builders, planning and zoning groups, and all organizations of governing.

Allocation to communities of hundreds of billions of federal and state funds no longer will be fact-based and likely will be inequitable. Many formulas, such as funding to school divisions, allocate resources on a per-capita basis. If the numbers of school-age children are distorted, two school divisions with identical numbers of children might receive significantly different funding. Students might suffer, teachers and school security staff might not be rehired, bus routes might be cut.

Second, census data no longer will be a reliable source for enforcing voting rights and for drawing equitable voting districts.

Third, census data will no longer be useful in planning and emergency management. When a hurricane strikes, we will not have an accurate number of how many housing units are occupied in the affected area. Anticipating needed nursing home beds, projecting community college enrollment, evaluating business sites and planning road construction all require accurate local data.

Finally, the decennial census data has always been the benchmark for federal, state and local statistics. For example, tracking regional rates of opioid addiction or measuring regional progress in cancer treatment won’t be possible.

The Census Bureau has long had effective privacy protections in place, and acknowledges that this new procedure represents a sea change in how census data is reported. Both internal reviews by Census Bureau staff and external reviews by expert users from federal agencies, local governments, academics, businesses and nonprofits have raised serious concerns. We hope these concerns lead the Census Bureau to reconsider these plans that would remove our ability to understand and to serve the populations of our counties, regions, cities and towns.

If you serve, or are served by, any of the organizations mentioned as examples, please contact the Census Bureau or the offices of your governor, senators, congressional representatives and city/town councilors or board supervisors to express your opinions. No one wants to be forced to live with the disastrous consequences of erroneous data throughout this new decade.

Qian Cai directs the Weldon Cooper Center’s Demographics Research Group at the University of Virginia. Contact her at This commentary originally appeared in the Richmond Times-Dispatch.

Catch the latest in Opinion

* I understand and agree that registration on or use of this site constitutes agreement to its user agreement and privacy policy.

Related to this story

  • Updated

Students aren't necessarily hard to reach in the same ways as other under-represented populations, but changes in the census may shift officials' ability to reach them. This is the first census where large amounts of phone and online responses are expected.

Get up-to-the-minute news sent straight to your device.



Breaking News

Breaking Sports News

News Alert