Twitter user @glennrice called out a "journalist" for producing the following chart:


You can't say the Columbia Heartbeat site doesn't deserve a beating over this graph. I don't recognize the software but my guess is one of these business intelligence (BI) tools that produce canned reports with a button click.

Until I read the article, I kept thinking that there are several overlapping lines being plotted. But it's really a 3D plus color effect!

Wait there's more. This software treats years as categories rather than a continuous number. So it made equal-sized intervals of 2 years, 1 year, 2 years, and 8 years. I am still not sure how this happened because the data set given at the bottom of the article contains annual data.

The y-axis labels, the gridlines, the acronym in the chart title, the unnecessary invocation of start-at-zero, etc. almost make this feel like a parody.


Aside from visual design issues, I am not liking the analysis either. The claim is that taxes have been increasing every year in Columbia, Missouri, and that the additional revenue ended up sitting in banks as cash. 

We need to see a number of other data series in order to accept this conclusion. What was the growth in tax revenues relative to the increase in cash? What was the growth in population in Columbia during this period? Did the cash holding per capita increase or decrease? What were the changes in expenditure on schools, public works, etc.?

This is a Type DV chart. There is an interesting question being asked but the analysis must be sharpened and the graphing software must be upgraded asap.

PS. On second thought, I think the time axis might be deliberately distorted. Judging from the slope of the line, the cumulative increase in the last 8 years equals the increase in past two-year increments so if the proper scale is used, the line would flatten out significantly, demolishing the thesis of the article. Thus, it is a case of printing cash, graphically.




Comments are closed.