Welcome visitor you can log in or create an account


Consumer Insights. Market Innovation.

Posted by on in Better Graphics
  • Font size: Larger Smaller
  • Hits: 19623
  • Print

Numbers That Don't Add Up

In my last blog I talked about a simple chart on Morning Joe, which was presented by Steven Rattner. I submitted that when we see data presented in the media or especially by politicians, we should judge it in terms of how a researcher would have presented the same data (because of course researchers are free of bias...well let's leave that for another blog). I gave Mr. Rattner a pass last time, but his presentation of a chart on infrastructure was misleading and would only have pleased a client who wanted misleading data to prove a point.

In this case he presented a chart showing infrastructure spending as a percentage of GDP . It showed a massive drop from the high in the 1950's to the low of today. The chart had a y axis that went from 0% to 1.5% which made the drop easier to see. Nothing wrong with that (assuming those viewing the chart understood that it was not based on 0-100%).

The chart led the panel to decide that we need to massively increase spending (to be fair, this was a position they already held so the chart simply backed up prior thinking). Many referenced new airports, highways or railroads in China and how we would no longer be able to compete with them if we didn't match them.

Now of course it is likely we need to spend more on infrastructure, but if I presented this chart to a client, they likely would have asked some pretty tough questions and ultimately found the chart to be of no use even to set a context for decision making. For example, they might have asked:

  • What impact does the rise in GDP have on these percentages? The rise in GDP has been driven by rises in population and by efficiency. The result is we now spend less on the stuff we bought in the 1950's, making it possible to buy stuff we didn't have back then. If I were to chart spending on food I'd see it drop from around 20% in the 1950's to less than 10% today. A view of such a chart might convince me we are all starving to death (given all the epidemic of obesity in this country that clearly isn't so). So some context is necessary to really understand these data.
  • What value does our current infrastructure have? A company that builds a new factory will see their spending on factory building go through the roof in the year they do it. In the years that follow they might not need more productive capacity so their spending goes to nothing. While China might be outspending us now (as a percentage of GDP or even in real terms), my bet is we've outspent them over time. Sure some of our past spending no longer has value, much of it does. For example, the excavations necessary to build the Schuylkill Expressway from Philadelphia to its western suburbs in the 1950's, was massive and expensive. While the highway is no longer adequate for the traffic it supports, the excavation for the existing lanes won't have to be done again.
  • What is China spending? The chart didn't say. All they offered were anecdotes about cool new airports and super fast trains. I could talk about the intersection of several major highways at the suburban end of the Schuylkill Expressway in King of Prussia and how an improvement made a few years back fixed decades old congestion problems, but that doesn't mean we are spending enough anymore than a new airport in China means they are outspending us.
  • What impact does population have on the needs and for that matter what about geography. China is more populated but the US is more spread out. Surely these factor into infrastructure needs. Perhaps looking at spending in terms of GDP per person would be a better meaure.

Ultimately, the chart was intended to answer the question, "What should we be spending on infrastructure?"   I don't believe it achieves that. If I presented a chart like this I likely wouldn't be asked back. Instead, I would need to consider many questions, including those above, and present data such that it became a framework for decision making. I would leave out those details that add nothing and consolidate others into comprehensive charts that quickly make the point. Of course, I'd also build to a chart indicating where the greatest need for increase spending is.

Perhaps this is an advantage we have as researchers. We can decide to leave out superfluous data (as I discussed in my last blog) which are critical to decision making. I suspect that Mr. Rattner, even if he wanted to, would not be given enough air time to explain all the complexities of the data he presented. Fair enough, but if we are faced with a situation in which we don't have adequate data to provide certainty, we have an obligation to point out that fact.

As for Morning Joe, I suggest if they are going to present numbers like this out of context, they'd be better served by skipping the charts and just discuss anecdotes...

President, TRC

Rich brings a passion for quantitative data and the use of choice to understand consumer behavior to his blog entries. His unique perspective has allowed him to muse on subjects as far afield as Dinosaurs and advanced technology with insight into what each can teach us about doing better research.  


  • No comments made yet. Be the first to submit a comment

Leave your comment

Guest Saturday, 24 October 2020

Want to know more?

Give us a few details so we can discuss possible solutions.

Please provide your Name.
Please provide a valid Email.
Please provide your Phone.
Please provide your Comments.
Invalid Input
Our Phone Number is 1-800-275-2827
 Find TRC on facebook  Follow us on twitter  Find TRC on LinkedIn

Our Clients