Universal Voter Registration on a national scale
by Shai Sachs, Fri Dec 05, 2008 at 04:37:29 AM EST
The Progressive States Network's focus this week is on universal voter registration. This is an idea whose time is long overdue. The kind of attention and pitched battles we fight over voter registration is nothing short of ridiculous, if you stop to think about it. Basically, the purpose of voter registration is to allow the state to build a database of eligible voters, and to locate those voters in cities and precincts. This is fundamentally a technical challenge, and really, not a very difficult one. The fact that voter registration is a difficult and mammoth task is, in the age of sophisticated data matching algorithms and large scale database systems, a national embarrassment.
PSN, whose focus is on state policy reforms, rightly suggests a number of incremental steps towards making voter registration easier and more widespread. These steps include aggressively mining databases to track citizens as they move about the state, making voter registration available at more places besides the DMV, etc.
That is a good start, but it seems to me that Congress could one-up the states, without too much effort. Simply by matching records from the social security database with the list of tax returns each election year, the federal government could produce a "pretty good" list of all 18-year-old citizens, and the addresses for each of those people. There would be some rate of error - as in the case of young adults who live in one place when they file their taxes in April, and have moved by the time elections come around in November - but I think it would be a reasonably good approximation. Why not make that list the default voter roll, and then allow citizens who moved, or who didn't file taxes for some reason, to add themselves to the rolls after the fact? To make things a little smoother, citizens could be given a web-based and phone-based interface to the database, to allow them to check their own registration status, indicate an address change, or perhaps even to find their polling place or request an absentee ballot by mail.
Having worked on identifier matching algorithms in the past, I am reasonably certain that this is technically a fairly simple idea, and won't be a budget-buster. In fact, to the degree that it allows states to slim down their voter registration infrastructure, it might even come out costing the combined federal and state budgets less than the current regime. (A small cost would be added to the US budget, and many small-ish costs subtracted from state budgets.)
The problem, as I understand it, is one of jurisdiction: voter eligibility has always been a state concern, except where the US Constitution bars certain practices (poll taxes and sex discrimination, for example). So while Congress can produce a first-guess voter list, it can't mandate that such a list become the default voter list for any state. However, the simple creation of this database might be enough to get the ball rolling. Giving states the chance to cut costs and simultaneously make voting easier for citizens could be a good enough incentive to sign on some early-adopter progressive states. As the system gains popularity, it could eventually become the de facto standard, exerting pressure on more conservative states to follow suit. Congress could, presumably, add more and more "carrots" to encourage states to adopt the first-guess voter list as their default.
This policy is simple, cheap, and low-profile enough that, I think, it could sail through Congress relatively easy. It's the kind of "Government 2.0" reform which it seems will be all the rage in Obama's Washington. And it could be a progressive positive feedback loop which would help make the country more progressive in the long run.