John Perilli: The Power of Data-Driven Politics
Wednesday, April 09, 2014
For me, this was the fact that 19.5 percent of Rhode Island children––almost 1 out of every 5––live below the poverty line, a fact highlighted by Rhode Island Kids Count at their annual policy breakfast on Monday. Think about it. Pick any five children in the entire Ocean State at random, and odds are one of them is in poverty. To me, that’s stunning.
Thus is the power of data-driven politics. Sure, anecdotes can play on your heartstrings, but a cold, hard number hits both your senses of emotion and reason. Both in Rhode Island and nationally, organizations ranging from advocacy groups to media outlets are embracing the politics of data. And no one is happier about this than I am. Data makes for better advocacy, better policy and better government, something we need desperately here in Rhode Island.
But even with this collective rush to the spreadsheets, we shouldn’t get ahead of ourselves. Raw numbers by themselves are dull and dead. They can’t inspire action without a picture frame to put them in. Worse yet, numbers can be used for deceit, and since they are trusted, they are all the more sinister when they’re false. So let’s embrace the numbers game of politics, but let’s do so, as any good statistician would, with a requisite amount of restraint.
An Ocean of Numbers
The advocacy group Rhode Island Kids Count’s annually compiled “factbook” is quite a read. It is a comprehensive list of over sixty education statistics that track everything from student enrollment figures to child health and welfare. It takes a meticulous effort to put together, and finds a large audience: Over 500 people including scores of elected officials and candidates attended Kids Count’s event on Monday.
(Full Disclosure: I am currently taking a class taught by Elizabeth Burke Bryant, the Executive Director of Rhode Island Kids Count.)
This, to me, shows an undeniable respect for the rigorous politics of statistics and data here in the Ocean State. When advocates can compile data, then go to the State House and use it to influence lawmakers, it goes a long way towards a more effective government.
Aiding this new data-driven advocacy is the sheer fact that there is so much more available data today, a veritable fountain of statistics on every measure imaginable versus the selective calculator tape rollout fifty years ago. Have a look at Gallup’s Presidential approval ratings through time, and compare Harry Truman, the earliest President they track, to Barack Obama. Obama has many more data points than Truman, painting a much more accurate picture of public opinion. Now multiply this by every measure anyone has ever tracked, and you will begin to understand the wealth of data we have accessible to us.
There are also hopeful signs that the statistical approach is being adopted by groups of all political persuasions. It is here that I must grudgingly give credit where it is due to the Rhode Island Center for Freedom and Prosperity. I may not agree with them at all, but in their recent report on state government, they took the time and diligence to make use of data, crafting a numerical lens for their vision of paring down state government.
Do we want better governance in the Ocean State? This is how it is done. Let our lawmakers swim in a sea of statistics, and give them access to the sharpest analysts around, so that we can craft better policy and solve our myriad problems.
A Nationwide Trend
Of course, two data points in America’s smallest state do not necessarily indicate a trend. To make that claim would not be rigorous of me at all! But I do believe that data politics in Rhode Island are tied to a nationwide movement which is being led by journalists like Nate Silver and Ezra Klein.
In 2012, when every horse-race pundit in America was off and galloping, blogger and statistician Nate Silver stared them all in the eye and used polling data to predict the Presidential election. No one dared believe him, until he called all 50 states correctly.
This success has given rise to data-driven journalism projects like Silver’s own FiveThirtyEight (named after the number of votes in the electoral college) to former Washington Post writer Ezra Klein’s new explanatory journalism venture, Vox. With such large and devoted audiences, these two writers are trendsetters, and will lead more analysts to back up their talk with numbers.
I only hope that they use their data for good.
Numbers (Can) Lie
The reason Nate Silver’s statistical model was so successful was because it was so limited. It had a simple goal: to predict the results of the Presidential election, as well as a reliable source of data, in polling firms. However, when dealing with complicated policy issues such as education or healthcare, no set of statistics can draw a complete picture.
Numbers themselves are certainly honest, but their presentation need not be. At least in theory, you would have to look at every piece of data in existence to have a complete picture of the world, and this is simply impossible. This means that by cherrypicking numbers, it is very easy, whether by accident or on purpose, to tell untruths, and quite stubborn ones at that.
Consider that every month, the Bureau of Labor Statistics releases figures on the unemployment rate and the labor force participation rate––the number of adults either working or looking actively for work. National Democrats rightly trumpet that our unemployment is going down, but conservative critics counter by pointing out that the amount of adults in our labor force is at a historic low. Both facts are true, but each by themselves is not the whole truth.
Another point to note is that numbers cannot explain every issue. What set of data could explain women’s rights, civil rights, or our basic constitutional liberties?
Even with topics numbers can explain, such as economics, there are often many interpretations of the data. And thats if we laypeople can understand the data in the first place!
Politics driven by numbers instead of notions, by statistics instead of speculation, is a promising field, but we must treat it carefully. It will never fully replace the more traditional narrative approach to politics, nor should it. Stories are still powerful instruments for activism and political engagement, and it would be foolish to do away with them. Even with the rise of mass data, they ought still to hold an important place in our political discourse.
That is, of course, if they are fact-checked first.