By Chris Wilson
There's an old joke in math circles about Richard Nixon that goes like this:
"In the fall of 1972, President Nixon announced that the rate of increase of inflation was decreasing. This was the first time a sitting president used the third derivative to advance his case for reelection."
(A derivative, one recalls, is a rate of change of a value.Nixon was saying that the rising price of goods and services—inflation, our first derivative—was itself rising—second derivative—but that the rate at which it was increasing was on the decline—third derivative.)
Historians (Wikipedia) trace this observation to the October 1996 edition of the "Notices of the American Mathematical Society," in which editor Hugo Rossi wondered if voters fell for it:
"Was President Nixon telling us that the economy was getting better? Did his listeners understand that in fact the inflation rate was still increasing and thus the economy still worsening?"
When it comes to economic data, there is almost always a time frame and a measure of calculus that can make the news sound good. The 2012 election arguably hinged on whether voters saw the economy in gross terms—how many total jobs were being created—or as a rate of change—how quickly the economy was improving. In the first presidential debate, President Barack Obama emphasized the former: "Over the last 30 months, we've seen 5 million jobs in the private sector created. "By the third debate, Republican nominee Mitt Romney was mournfully trumpeting the latter: "You can't have an economy that over the last three years keeps slowing down its growth rate." (That's a second derivative, for those of you keeping score at home.)
This is not your basic "lies, damn lies, statistics" situation. It's a foundational question about how far back our memories go when judging the health of the economy, and how far forward our imaginations go when projecting current trends.
Last week, the Census released an incredibly detailed set of figures on the number of people employed in over 1,500 industries in 2011. This dataset, called "County Business Patterns," is an annual survey of just about every sector of the U.S. economy, with the notable exception of government employees and some agricultural workers.
The absolute figures in this sort of data can be hard to make sense of if, like me, you are not an economist. Is the fact that 504 people were employed in the formal wear and costume rental business in Illinois in 2011 good news or bad news? Without some specific knowledge of the industry, the easiest thing to do is look at how much employment figures have changed over time. As it turns out, the formal wear rental business shrunk by 25 percent from 2010 to 2011, according to the Census data.
In the following interactive map, you can compare any two years of data for any industry and see how they measure up. Use the slider at the bottom to change the time frame and use the menu to select an industry. You can type in a few letters of the word you're looking for to narrow down this long list.
You can also toggle between viewing the map in terms of total jobs gained or lost, or as percentages—something like the "Obama View"and the "Romney View," to put it crudely.
To see the epitome of the Romney View, select "Total for all sectors" on the menu and set the time frame at 2008 to 2011. Whether you look at totals or percentages, you will find that only two states are in a better position today than they were before the worst of the recession hit: North Dakota and Alaska. This can largely be attributed to fossil fuels.
For a happier picture, just drag the left handle of the slider to 2010. For the most-recent two years of data, all but nine states have added jobs after shedding them like the plague the previous two years.(You can drag the blue region between the handles on the slider to move both at once.)
Think creatively if you can't find a specific industry; the government often words things very precisely. If you're worried about your friends at Yahoo News, for example, search for "Internet publishing." (Though this one is a bit misleading since the data doesn't show up until 2008. You have to drag the leftmost handle on the slider to 2009 or 2010 to see a real effect.)
Extra credit: How many anti-recession industries can you find, which show growth when the rest of the country shows decline? For example: Set the slider to 2008-2011 and select "Repossession services." Found a good one? Let me know: @chriswilsondc.
You'll see some employment figures as a range, like 20-49.This happens when the Census is blurring data to protect the privacy of a company whose payroll could otherwise be imputed from the data. For the purposes of comparison, this map assumes the lower value. Thanks to the several different employees of the Census County Business Patterns office who took my calls to explain this. Questions? cewilson@yahoo-inc.com.Want to try it at home? My source code is on Github.
