Sunday, April 23, 2017

Interview: Does the media's obsession with polls thwart accuracy? Charlie Cook explains.

This is the third of five articles spanning my discussion with Charlie Cook. The first and second parts are available. 
Story by Joseph Ford Cotto
"Like I've said before .... polls are only good for strippers and cross-country skiers," Former Alaska Governor Sarah Palin told Fox News host Bill O'Reilly two days before last year’s presidential election.
While I have criticized her more than a fair bit during years gone by, it is undeniable that she turned out to be a sage among pundits.
When Palin made her remark, virtually all national opinion surveys – save two highly important yet conspicuously underreported ones – indicated an impending win for Hillary Clinton. Reuters predicted she was set to win 247 electoral votes outright and favored to seize so many more that her chance of victory hovered at 90 percent.
At the Princeton Election Consortium, Dr. Sam Wang – a neuroscientist and prolific author – declared that Clinton enjoyed a 99 percent probability of winning. The platinum-grade forecaster Moody's Analytics also claimed she would triumph in the Electoral College.



United Press International, in conjunction with the polling group CVoter, reported that her Electoral College lead over Trump was substantial: 259 to 209.
After the race was called, Clinton barely eked out 232 votes. Her vaunted 'firewall' of Michigan, Pennsylvania, and Wisconsin crumbled into ash. Only Minnesota and New Hampshire stood -- and not by any means tall. Clinton struggled to win either, despite the former having gone to every Democratic nominee since Richard Nixon's 1972 earth-slide over George McGovern.
Just a few hours earlier, Clinton fans were jubilant over her seemingly assured victory. By the morning after, perhaps more tears were shed than at any other time in twenty-first century America.
Like nearly all other professional election watchers, Charlie Cook believed that Clinton would defeat Trump. He went so far as to declare their contest finished during mid-October. Despite the – to quote an infamous Chicago Tribune headline – "Dewey Defeats Truman" quality of said statement, it is an aberration from the norm. Cook has enjoyed a long, prosperous, and stable career of election forecasting. 
As his must-read publication, the Cook Political Report, explains, he "is considered one of the nation’s leading authorities on American politics and U.S. elections. In 2010, Charlie was a co-recipient of the American Political Science Association's prestigious Carey McWilliams award to honor "a major journalistic contribution to our understanding of politics." In the spring semester of 2013, Charlie served as a Resident Fellow at the Institute of Politics at the Kennedy School of Government at Harvard University.
"Charlie founded the Cook Political Report in 1984 and became a columnist for Roll Call, the newspaper of Capitol Hill, in 1986. In 1998 he moved his column to National Journal. Charlie has served as a political analyst or election night analyst for CBS, CNN and NBC News and has been a frequent political analyst for all three major broadcast news networks and has appeared on Meet the Press and This Week."
Cook recently spoke with me about many issues relative to electioneering in modern America. Some of our conversation is included below.
****
Joseph Ford Cotto: Crunching numbers, especially in our social media-driven, sound bite-prone age, is touted as the begin all, end all of event forecasting. In 2016, perhaps more attention was given to polls than in any previous election. Might the press have generated such a demand for polling data that quantity increased at the expense of quality?

Charlie Cook: As issues become less important in our campaigns, personalities and images more important, with the focus increasingly on the horse-race aspect of campaigns, everything seems binary, somebody is going to win, somebody lose, nothing else is important, it is inevitable that polls get more attention.

Ironically this is all coming at a time when polling is less dependable, it is the only thing that I am familiar with where technology has rendered it less effective than ever.  A lot of people dwell on the switch from landlines to cell phones as the reason for polling becoming less dependable, it is really a fairly minor factor.  Good pollsters have gradually increased the proportion of their cellphone interviews in recent years, I noticed that in the Gallup tracking this year, the cell proportion was up to 70 percent.  

The real culprit was the abuse of telemarketing over the last couple of decades and increasing use of caller ID and voice mail to screen unwanted calls, making it difficult to get a representative sample on the phone, whether to landlines or cellphones.  Fifty years ago, someone felt honored to be asked their opinions about politics and issues.  Today it is ‘who is interrupting my dinner’ or ‘bothering me when I am busy.’  Good pollsters will call one telephone number representing a sampling point several times before moving to an alternative number and person.  The sample becomes one heavily weighted toward couch potatoes or those who have nothing better to do that moment than talk to a pollster.  That’s obviously not good for collecting good data.

Just as polling transitioned from personal, face-to-face interviews with people on their doorsteps or in their living rooms to telephone interviews in the late Sixties and early Seventies, we are undergoing a shift from telephone to on-line interviewing.  I am a bit old-fashioned, I think this transition is inevitable but I don’t think we can yet get more representative samples on-line than on the phone, though it will certainly and eventually happen.  Though plenty of people who are 70 or older or low income people who can and occasionally do go on-line, I do not believe you are getting a cross section of those groups yet who are willing to go through an on-line survey.  Some very good researchers are working hard to accomplish this, along with some pollsters just using a cheaper alternative way to interview, but I would still rather see a live interview than either a IVR (robo) poll or one conducted on-line.  But the day will come.

For elections, I suspect five to ten years from now, what will be used for polling will be a form multi-modal modeling, using some combination of telephone and on-line polling integrated with Census data and other data available from voter files and other data bases.

Cotto: Historical data, analyzed and complied into two studies by SUNY Stonybrook's Helmut Norpoth and American University's Allan Lichtman, respectively, indicated that Donald Trump would win. Many, both in the media and otherwise, chose to ignore these academics even though they have solid track records. After 2016, some might say that history is a better election forecaster than opinion surveys. What is your perspective on this?

Cook: There are dozens of political scientists, economists and other academics who make presidential forecasts, in any election, some are going to be right, others wrong.  Some elections are fairly easy to forecast, others more difficult.  In some presidential elections, most of the forecasts are right, few are wrong.  In others, some might end up being right for the wrong reasons. In some cases forecasts can be right for the wrong reasons.  In some cases there is less there than meets the eye.  I don’t see any percentage in picking a fight with either, suffice it to say among those in the business, they are not considered to be any better than many others.

No comments:

Post a Comment