[EL] von Spakovsky's latest Backgrounder from Heritage

[Doug Hess]

Hans von Spakovsky has a new report out on voter registration. There's
a lot that could be said about it, but I'll focus on just a few things
(some of which may only make sense if you read the report):

1. He frequently confuses what the CPS data on non-voting means or how
to assess it, usually by conflating terms and measures or by leaving
out other information. The CPS data is on this page:
http://www.census.gov/hhes/www/socdemo/voting/

For instance, as evidence that registration barriers do not exist, he
points to the fact that the percentage of people stating they didn't
register because they were not interested in politics is larger than
the percentage stating they didn't know where or how to register.
However, he does not discuss other responses that could point to
registration policy or election administration as a source of
problems. Instead, he mentions them in passing without considering the
possible implications of these responses. E.g., "difficulty with
English" as a reason for not registering to vote strikes me as pretty
sad statement about the quality of voter registration access!
Likewise, not meeting residency requirements or the registration
deadline can also be interpreted as statements about registration
policy (or public misunderstanding). In fact, I believe, all three of
these issues have been treated as serious barriers by federal law. He
may not support those laws, but he shouldn't cherry pick the data like
this.

He points out that these answers should be taken with a grain of salt.
Certainly I often doubt that people know why they did something...and
why they didn't do something seems an even more complex item to
report. However, he only applies this skepticism against the responses
he doesn't like.

2. He uses the small percentage of those registered who said they did
not vote mainly because of "registration problems," as evidence that
registration problems do not exist. Well....that's not how I would see
this. If somebody thinks they are registered (correctly or not) and
cannot vote (but apparently tried...although we don't know exactly
what happened), I would say that this is a huge problem with
registration! In 2008, the problem was stated by ~900,000 adult
citizens (of course that number is an estimate from the survey
sample). Again, that's not a minor problem. People often confuse
percentages with substance, but a small percentage of a large number
is a big result. I'll take one-hundredth of one percent of Bill Gates'
wealth any day.

More importantly, this response to this question does NOT mean that
registration problems are rare. This question and this response are
about problems with registration for those who (think they) are
registered. Big difference. It was nothing about registration access
overall.

3. He ignores the fact that people could only choose one answer on why
they didn't vote if they were registered. The distribution of the
single choice by education or income (etc.) is not easy to interpret
in this situation. For instance, the percentage reporting that serious
illness or disability was the reason they didn't vote increases with
age and decreases with education (as one would expect). Thus, the
existence of a problem with their registration is less likely to show
up for those groups (i.e., if they had more than one barrier, we don't
know). Percentage distributions of "check only one" questions are prey
to this problem.

Again, this is NOT the question to look at for assessing voter
registration access (it does, however, indicate that many people, as
other surveys have found, think they are registered but are not...for
whatever reason, their error or the election officials' error...maybe
some of them got knocked off the list due to list cleaning).

In short, his statement "The Census survey, in other words, actually
demonstrated that less-educated voters had fewer registration
problems," is based on a basic misunderstanding of the survey. And,
one last time, the question says nothing about registration problems
among the non-registered, which is where you should look.

4. He ignores peer-reviewed, published research showing that
registration policy (and post-registration policy) matters.

A lot more could be said (e.g., on costs, other data and research, his
odd portrayal of the action of various nonprofits, etc.), but I'll
stop with that.

The link to his report is here:

http://www.heritage.org/research/reports/2013/03/mandatory-voter-registration-how-universal-registration-threatens-electoral-integrity


Douglas R. Hess, PhD
Washington, DC
ph. 202-277-6400
douglasrhess at gmail.com

Starting Aug. 2013:

Assistant Professor
Department of Political Science
Grinnell College
1210 Park Street
Grinnell, IA 50112-1670