Bad Data Science = Bad Business: Census 2020

Tomorrow, the U.S. Supreme Court will hear arguments on whether the government can add a question about citizenship to the 2020 census. By adding this new question, the plaintiffs in the case argue that participation in the census will be suppressed, causing a significant undercount.

At Nielsen, we’ve worked with data for almost 100 years. We know data science. And we know that adding an untested question to this census is bad science and will likely lead to a significant undercount of the U.S. population. In particular, we believe that young, multicultural people—a consumer segment that will be the main driver of economic growth for the next decade—are at great risk of being underrepresented.

To that end, we’ve filed a “friend of the court” brief with the Supreme Court supporting the exclusion of a citizenship question to the 2020 decennial census. We’ve outlined our concerns that such a question could present for businesses today in an Op-Ed in the New York Times.   

The purpose of conducting a census every 10 years is to get a complete, accurate count of the population in our country. If the government is successful in adding the citizenship question to the 2020 census, the census will contain flawed data that will have far-reaching consequences for U.S. commerce and media, retail and consumer packaged goods businesses. Billions of dollars follow these counts.

Nielsen, in particular, relies on the U.S. Census data. We use it at a town-by-town, block-by-block level to determine the makeup of our measurement panels. In turn, this data allows us to provide measurement and analytics services to our clients. It is our panel participants—representative of their particular demographics, including race, age, gender, socioeconomic class, and other characteristics—together with large data sets, that allow us to project the ratings of a particular piece of video content, the market share of a beverage manufacturer, or what products to stock and where to site physical stores.

Nielsen is the gold standard when it comes to measurement and data science. Our rigorous methodology allows us to measure what people watch, buy and listen to while accurately representing U.S. diversity. That’s why we’re committed to advocating for a sound research approach to the Census. If the U.S. Census data is inaccurate, businesses won't know their customers exist—they won’t plan for them and won’t reach them, leaving perhaps hundreds of millions of dollars on the table.

One of the reasons I am proud to lead Nielsen is that we employ trusted and fair data science principles. We spend millions to ensure that every person is represented in our measurement. To maintain the integrity of the data on the population of our country, it’s essential that no one is deterred from completing the census. Counting everyone makes all the difference in the world.