Written by: Director Robert Groves
One of the issues I think about a lot these days is maximizing the credibility of statistical information that we provide to the United States.
That statistical information is a key feedback loop for the American public. Our estimates describe almost every aspect of our lives – income and educational attainment, retail and wholesale sales, occupational distributions in small communities, foreign trade flows, and changes in job distributions. They allow the citizenry to assess how things are going and to evaluate actions of the government that may affect their well-being. If the Census Bureau statistics are not believed, if they’re not found to be credible, we have failed.
We go through many steps to assure that the statistics we produce are of the highest quality given our resources. Our success at that depends on the professionalism and technical skills of our staff. In short, we believe that to be found credible our statistical information must be sound statistically.
The credibility of our statistics, however, relies on more than just their technical qualities. We have to be transparent in describing our methods. To be credible we must describe both the strengths and weaknesses of our statistics. In a world filled with advertising that only describes the strengths of products and not their weaknesses, this is an unusual activity. However, this is important for us to do. None of the statistics that we deliver are perfect; we know that. So we tell users about their imperfections so that they can take those into account. By being completely honest about our data products, we hope to win the trust of the users and enhance the credibility of our statistics.
A challenge in solely using transparency to enhance the credibility of our data products is that our statistical information is often delivered by a set of intermediaries – print, digital, and video journalists. The time-constrained environment of the 24/7 news world usually doesn’t permit their going beyond reporting only the numbers themselves.
Since many people learn of our estimates from intermediaries, without detailed descriptions of how they were assembled, their reaction to our estimates is affected about whether we are viewed as a trusted source of information. Does our name connote professional, objective, nonpartisan measurement and reporting? Are we viewed as independent of whatever controversies might exist about the phenomena we’re reporting? Do we “tell it like it is?”
Informing such judgments is long term work. As current stewards of the reputation of the Census Bureau, we need to do everything we can to preserve the trust of the American public so that to them our estimates can be believed and used effectively.