Lying with numbers

Lies, damned lies, and statistics. Recently, I’ve been getting a graduate-level course in how to lie with numbers. The first lesson comes from the Heritage Foundation, which has a chart showing that despite spending tons of money on anti-poverty programs, the poverty rate hasn’t budged in years.

rectorchart

The chart is by Robert Rector, in a column titled “The War on Poverty Has Been a Colossal Flop.” What’s wrong with this picture? The vast majority of means-tested welfare programs, like Medicaid, are not counted by the official poverty standard. So basically, the chart proves that anti-poverty programs don’t reduce poverty when you ignore the effects of anti-poverty programs.

Strangely, Rector acknowledges this point, which makes it even weirder that he would go through the trouble to make the chart in the first place. His argument seems to be that poor people aren’t really poor because they have microwaves (seriously). So…if there are no poor people, then didn’t we win the War on Poverty? Rector still says no, because the poor don’t all have good paying jobs. I agree that’s a problem, I’m just not sure taking away their health insurance will solve it.

But again, why bother making a chart you know proves nothing, unless your goal is to mislead? And he is very committed to making this data into a chart; he included a similar graph in an earlier post titled “This Chart Proves the War on Poverty Has Been a Catastrophic Failure.

The second lesson comes from American Enterprise Institute, which is usually more honest than Heritage, but seems to have dropped their standards in this case. Mark J. Perry argues that rising economic inequality is an “imaginary hobgoblin” (points for mixing Spiderman villians into an economics column), and he has a chart to prove it:

incomeshare1-600x416

He says since the lines are (mostly) flat, income inequality has not really increased (much). But what’s missing? Say it with me Occupy Wall Street: the 1%. Given that the top 1% has been the focus of income inequality debates, their absence is odd.

And indeed, the top 1% are pulling away from everyone else. Not only that, when you separate the top 1%, you find that the income share going to the rest of the top 5% has not actually changed much. Also, American history did not start in 1993, as Perry’s graph does, and looking back further reveals more evidence of rising inequality.

Screen shot 2014-09-16 at 7.42.40 PM

Perry defended his graph by pointing out that the Census data he was using do not break out the top 1%. If only someone had data on the top 1%…

Screen shot 2014-09-16 at 7.43.49 PM

Those two examples are both from conservative think tanks, but progressives don’t have a monopoly on the truth either. Here’s MSNBC showing skyrocketing corporate profits:

steve-benen62B0C485-A5B6-D5A1-41BD-C60EEBAEEE5F

That looks like a shocking trend, until you realize that it’s not adjusted for a growing economy. The U.S. economy is much bigger than it was 50 years ago, so of course corporate profits are also larger. Using the same presentation that MSNBC chose, I made a graph showing that wages are also skyrocketing:

fredgraph

To be clear, wages are not really skyrocketing. To show a meaningful trend in corporate profits, it’s important to put them into economic context. I did that recently at Center for American Progress, showing corporate profits as a share of national income:

CorporateProfitsTaxes_fig1This still shows corporate profits at a record high, but reveals that corporations only started capturing a growing share of income in a sustained manner around the year 2000. It still makes an important point, just a bit less dramatically.

In all three of these examples of lying with numbers, notice how the numbers were always correct. The data was good, but the application of that data was not. In the case of the poverty numbers, two trends that have nothing to do with each other were shown on the same graph, with the implication that their lack of correlation proves something. In the case of the income inequality numbers, the graph left out the most important trend regarding rising income inequality. In the case of corporate profits, failing to adjust for the growing economy can make any economic indicator appear to skyrocket (although even a more honest presentation does give cause for concern in the case of corporate profits).

Numbers don’t lie. People lie with numbers.

Advertisements

One thought on “Lying with numbers

  1. Linwood

    Hello admin do you need unlimited content for
    your website ? What if you could copy article from other
    sources, make it pass copyscape test and publish on your site – i know the right tool for you, just search in google:
    loimqua’s article tool

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s