Monday, May 01, 2017

Why Bad Politicians Hate Good Academics - Mouhcine Guettabi Explaining His Research on Economic Impacts On Two Alaskan Boroughs

There's a war going on against academics as well as against the media.  People who try to uncover the truth are never the favorites of people in power and often they are persecuted.  (See this and this, for example. Journalists are being killed in a number of countries including Trump's role model, Russia.  But first he wants to get rid of the First Amendment.  While that may be just distraction or hyperbole, the trouble with liars is you don't know when they are serious. You get the idea.  Tracking down fake news takes up time that could be better spent.)

Dr. Mouhcine Guettabi's Presentation at ISER

Friday at noon, I went to a display of what good scientists do to get near the truth.  And why bad politicians hate good scientists.

Economics professor and researcher at UAA's Institute for Social and Economic Research (ISER,) Mouhcine Guettabi,  gave a talk.

I don't claim to understand all the details of his formuli, but what was so refreshing was listening to how he treated the 'truth.'

He never claimed to have it.  He qualified everything he said.  And, in fact, the whole presentation was a preliminary look at work in process (he even said he didn't think he could say 'in progress') in front of about 25 others, many of whom know the subject and the methodology well.  He wanted them to critique his methodology, to challenge his tentative claims, to help him get closer to the elusive truth.

After hearing so many politicians attack or embrace statistics to 'prove' that their policies are working, regardless of whether the claim has any validity, I found Guettabi's talk to be like a walk in the wilderness.  Instead of the noise of politicians trying to spin everything in their favor, we just heard the pure sound of good minds struggling with the difficulties of really knowing things.

Impacts of Red Dog Mine on NWA Borough and Oil Price Increase on North Slope Borough

These were the issues Guettabi was wrestling with:
  1. How did the opening of the Red Dog Mine in the Northwest Arctic Borough impact people in the borough?
  2. How did increasing oil prices affect employment in the North Slope Borough?
Of course, in situations like this, you don't really have a control group to measure against.  The big problem was how could he know what each of those boroughs would have looked like if a) there had been no mine or b) there had been no sharp increase in oil prices?   

He attacked that problem by creating hypothetical 'twin' boroughs that he could compare them with on a variety of measures.  He took all the Alaska boroughs and found the handful whose numbers in the years prior to the Red Dog Mine's beginning (and before the oil price spike) and used those 'twins' to compare the 'after' years.  How did the hypothetical 'twin' boroughs do compared to the actual NWA Borough and the NS Borough? So he had different groups of boroughs for different measures (since the ones with the closest matches were not the same for the different statistics he was looking at.)

Here's my transcript of what he says on the video below:
"However, I have all these boroughs that had no mine, so let me use information from those other boroughs in order to construct a unit that looks as much as possible like the Northwest Arctic Borough before the mine was created, and therefore the difference, essentially, between the outcome I actually observe, right?  So this is, for example, employment in the Northwest Arctic Borough in 1992 minus the weighed average of employment in that ‘twin’ would give me, arguably, the effect that I’m interested in, which is what is the actual effect of the  mine itself."

I don't want to go into all the details - well, even if I wanted to, I couldn't get too far.  But I just want to highlight
  • how difficult it is to really know how living communities or political subdivisions are impacted by things
  • how painstakingly good scientists work to find ways to answer these questions
  • how humble good scientists are and how welcoming they are to other experts reviewing and challenging their thoughts and methods
  • how different this process is from the way bad politicians 'prove' cause and effect

And while President Trump offers an extreme example of a bad politician* I think you can all find lots of examples of other federal, state, and local politicians who range from careless with numbers to  creating their own totally unjustifiable conclusions from the data they have.

And let me also note, that these are simply attempts to check on the impacts in these boroughs.  Guettabi recognizes that there are many factors that could go wrong and he worked hard to consider them in his model.  Most, if not all, of the data Guettabi is using (this is a work in process as he said) comes from governmental data sources.

And my guess is that this is precisely why some politicians are trying to defund data collection in key agencies.  Without good data, scientists, journalists and others, can't demonstrably refute politicians' false claims.  The NRA's successful campaign to limit the Center for Disease Control's (CDC) gun violence data bases) is a good example of this.

Data sets aren't as photogenic as physical violence, but destroying them is a form of structural violence and in many ways just as important.  It's an assault on how we know things, how we get closer to 'truth.' And whether we have good data to make good policy decisions.  

*"bad politician" here means one whose approach is to highlight whatever makes him look good (whether it's true or not) and attack his opponents or detractors (whether it's true or not) rather than focus on getting everyone together and solve our collective problems.   

No comments:

Post a Comment

Comments will be reviewed, not for content (except ads), but for style. Comments with personal insults, rambling tirades, and significant repetition will be deleted. Ads disguised as comments, unless closely related to the post and of value to readers (my call) will be deleted. Click here to learn to put links in your comment.