You may have missed the news that the head of the Census Bureau, John Thompson, resigned a few months ago.
In normal circumstances, the departure of a government statistician would not be worth highlighting. But Thompson’s departure adds to the growing uncertainty surrounding the success of the 2020 decennial census. About that, you should worry.
The U.S. Constitution has fewer than 5,000 words, not much more than five times the length of this column. Only a small fraction of what the government actually does is in this founding document. Yet the Constitution requires the United States government to conduct a census every 10 years to determine how many seats each state will have in the House of Representatives.
And the importance of government data is vast beyond that. Businesses use Census Bureau information every day to make critical decisions that rely on an understanding of the demographic and economic trends that influence their markets. The importance of this data to business is less understood than it should be. (To help remedy that, we wrote a report on the subject.)
If you walk into a Target store in suburban Florida, the items on the shelves are different from what is in a Target store in downtown Washington D.C. Target makes these decisions in large part using government data. An entrepreneur deciding whether to expand his business to a new city will use the data to study his potential customer and employee base. Businesses use official government statistics to benchmark wages, forecast consumer demand, monitor consumer spending patterns, and more.
Even as companies increasingly collect their own statistics and use them for their analysis, they often use government data to complement and validate their internal data. Official statistics are comprehensive, since the Census Bureau's objective is to create a set of facts about the U.S. economy and society as a whole.
Census information matters for public-policy purposes as well. The data are used to determine a wide range of federal appropriations to local communities -- funds for children’s health insurance, for example, unemployment insurance and school meals.
Historically, the U.S. census has been the gold standard of data quality, due in part to the rigor and the great lengths to which the bureau's workers go to ensure a complete and accurate count. If data quality slips, we will have a less reliable picture of the U.S., especially of groups that are traditionally hard to reach, including rural residents, immigrants and the poor.
There are two related threats to the coming decennial census: budget mismanagement and a leadership vacuum.
The cost of the census has been rising over time. The bureau has an innovative plan to reduce the collection cost per household in 2020 -- despite the challenges posed by the increasing distrust of government, for example, and the larger population of households that have traditionally been hard to reach. The plan is estimated to save a whopping $5.3 billion in total costs.
The innovations involve more technology and better leveraging of existing data sources. They are expected to pay off for the 2020 census and for decades to come. But these techniques must be developed, tested and integrated successfully in advance of 2020.
The bureau has estimated its funding needs to carry out the changes, with budget requests that steadily ramp up, especially in the 2017-2019 period when the planning and testing are most intense. A recent General Accounting Office report found that the upfront investment needed may be somewhat higher than what the Census Bureau requested. Nonetheless, Congress appropriated $122 million less than required for 2017, and the president's budget proposes an additional shortfall of $112 million in 2018.
This lack of investment is penny-wise and pound-foolish. If the Census Bureau does not have adequate resources to invest in developing a new modern infrastructure for 2020, it will have to do one of two things. It could resort to the old, expensive approach to ensure an adequate count, driving up costs to taxpayers. Or it could conduct a lower-quality census, diminishing the value of the data.
Both of these scenarios worry us. Diminishing the quality of data could hurt the ability of businesses and government to make critical decisions that affect our lives and our pocketbooks. And a run-up of costs late in the game that could have been prevented by some upfront investment is fiscal imprudence at its worst.
Not only does the Census Bureau need an adequate budget; in the wake of the director's departure, it also urgently needs a strong leader to guide the process.
Counting several hundred million Americans is a massive undertaking, relying on the groundwork done now in developing new systems and preparing to hire and manage temporary staff, including several hundred thousand workers to canvass addresses.
There’s no time to waste. The Trump administration should immediately nominate a new director, and Senate Democrats need to allow for the nominee’s quick confirmation.