Data. It surrounds us. We rely on it. Our elected officials rely on it for decision making (OK, maybe not enough).
So it’s worth asking if the data tells us what we think it does? In economics we often respond to a question with, “It depends.” And with data, it often does depend whether the data tells us what we think it is telling us.
How many people live in Whatcom County? It depends. The U.S. Census Bureau has two different numbers. Our state’s Office of Financial Management has a third. Why? The data is collected different ways. Different people are under- or over-counted differently each time. While there is no debate that we certainly have people, the number of those people, where they work and where they live is up for some discussion.
Recently two studies were published regarding the economic impact of recreation in our state and regions within our state. They have very different numbers – why? When conducting any research a number of assumptions are required. Large or small these assumptions can drive significantly different results. How study boundaries are constructed, assumptions about consumer behavior and how/where exactly people spend recreation dollars requires the use of different data sets and different error rates.
Sign Up and Save
Get six months of free digital access to The Bellingham Herald
In the case of the two studies, they didn’t rely on the same data. One study included things the other did not. (One considered boat building under the heading of recreation. The other took a more narrow view. Neither is necessarily right or wrong. Just different. And such differences often lurk in the data.)
We read a lot about data driven decision making (think “Big Data” and examples like data-driven health care). But we hear little about the data driving the decision making. For example, we’ll read a story about how Mariah Carey has sold more albums than the Rolling Stones – but we won’t find any mention of whether the sales are adjusted for differences in population when the records were available or whether the price of the albums differed, when adjusted for inflation. (Note: those sorts of things matter.)
Income in Whatcom County? You can look at household income or family income and get a very different image of Whatcom County. Do you know the difference between households and families? Households are simply people living at one address, while families are those living at one address that are related by marriage or blood. In a college community, this simple distinction can make a significant difference in the data that is used for analysis. When you read something about wages, do you know if it is household, family or employment data? Probably not, but they will give you different results.
Should you throw your arms up and declare nothing as trustworthy or accurate? No. Just start thinking more about how we know something and what assumptions drive that knowledge. When you read or overhear something, spend five minutes to look it up yourself to understand the limitations or structure of the data. Understanding the challenges inherent of all data sets is the key to forming better decisions.
Some data issues can be addressed and others cannot. These are the conversations worth having. The Center for Economic and Business Research at WWU conducts analysis to help businesses, non-profits and government agencies make informed decisions. Learn more at cbe.wwu.edu/cebr.
Have a question about our economy? Send an e-mail to firstname.lastname@example.org.
This is one in a monthly series about the Whatcom County economy. James McCafferty is the assistant director at the Center for Economic and Business Research at Western Washington University. Research for this column was done under the direction of Hart Hodges and CEBR staff.
The Center for Economic and Business Research at Western Washington University conducts analysis to help businesses, non-profit and government agencies make informed decisions. For more information online, go to cbe.wwu.edu/cebr.