Sunday, June 19, 2016

Lame Counterfactuals for $600, Alex

This has been going around:

Why is this such a dumb counterfactual?  Let me count the ways:
  1. Unlike parliamentary elections, individual voters could not vote for the full range of candidates--they voted in either the Democratic process or the Republican process
  2. Unlike parliamentary elections, the primary/caucus processes in the US were not simultaneous but sequential.  Which meant that folks voting later had few choices and were reacting to the outcomes of the previous primaries/caucuses
  3. Turnout? Primaries have lower turnout than general elections.  In primaries, the most passionate tend to turn out, so you tend to get more extreme wings of the parties show up.  Indeed, this is one reason why Trump is so confused and ill-prepared--he thinks the electorate is the same for the general as for the primaries, so he just has to do more of the same.
  4. Um, this conflates type of political system with electoral laws.  There are parliamentary systems with majoritarian electoral laws (that would be most of the Westminster places--NZ is wacky).  And presidential systems can have a variety of rules for picking the president and then different rules for the legislature.  
Ok, that's just a few (Economist also assumes most proportional of electoral systems), but I just did a bunch of tax forms (F-Bars!), so I could be forgetting some other reasons.  What have I missed?

It would probably be far more accurate to take a look at surveys of the entire electorate on a variety of issues and then figure out how they would shake out if we had a proportional representation electoral system combined with a parliamentary system.

Sorry, to be a killjoy, Economist, but this is the kind of infographic that creates ignorance, not enlightenment.


Unknown said...

One factual that you seem to have missed is that according to the footnote, this is not based on primary results, but rather on a poll carried out between April 22nd and April 26th. ;)

The only thing I find misleading is the label "Parliament". But they're trying to make things concise... "What if the United States Elected Its Representatives Using Proportional Representation, and Each of the Last Five Remaining Presidential Candidates Headed Their Own Party, How Many House Seats Would Each Party win?" wouldn't be quite as snappy, or fit as well.

First-past-the-post, winner-take-all systems (whether parliamentary or Presidential) tend to exaggerate divisions and make different parts of the country seem more different and more polarized than they really are. It's our current system of determining "winners" that's inherently misleading. So I think this graphic is very useful in providing some visual clarity about who Americans actually support, when their preferences are not distorted by the 2-party first-past-the-post system.

Ashley Giles said...

What do you expect from The Economist? I have this perpetual feeling that they never really think through anything they publish.

L'il Steve said...

I'm with Andrew. This is not for an audience of Political Scientists.

Steve Saideman said...

Sure, not for political scientists who can quickly diagnose how lame this is.