So here in Massachusetts, my old congressional district (MA-5) elected a replacement for Ed Markey, who became Senator over the summer. I find the winner, Katherine Clark, to be rather uninspiring, but that's another story for another time. The issue I have here is the turnout. Clark won with just a tick over 40,000 votes of the almost 61,000 cast. That sounds like a dominating victory, until you sit back and wonder why we're electing someone to represent roughly 705,000 people with only 40,000 votes.
It is often the inclination to self-righteously proclaim that people are not active enough, and that they don't care enough about their democracy. This is the wrong place for that. Democracy is hard. High turnout levels reflect voter engagement in part, but they also reflect the ease of the process - that's why those polling place changes down south often wouldn't get through the voting rights act preclearance. Making it harder for people to vote makes it less likely for them to vote. That having frequent elections for single seats rather than a main election day suppresses turnout is not a controversial statement.
In Massachusetts there has been a real hullabaloo in the last decade about how to replace a Senator who leaves mid-term. When John Kerry was almost elected President in 2004, many were scared that Governor Romney would appoint his replacement, who would serve the next four years. So they changed the rules to have a special election.
Then in 2009, Senator Edward Kennedy died, and we realized they'd done, excuse my French, a shitty job writing that law. So they changed the law on the fly, having an appointed representative in place until the special election. That election had something of a surprise result but the system worked because people were very motivated to turn out for a variety of reasons.
In 2013, we had yet another special election to replace Kerry. It was won, with fairly poor turnout, by Congressman Markey. (Full disclosure: I was once an intern in Markey's office and volunteered on his Senatorial campaign). That set off the need for ANOTHER special election to fill his seat. Nobody showed up to vote in it. Well, not quite nobody, but check out the town by town comparison with the 2012 election. I use 2012 because with a Presidential race and a hotly contested Senate campaign, that's probably as close to full turnout as we're going to get.
| City/Town | 2012 Total | 2013 Total | % Turnout |
|---|
| Winchester | 12767 | 7205 | 56.43% |
| Melrose | 15427 | 3810 | 24.70% |
| Arlington | 25799 | 5605 | 21.73% |
| Lexington | 18257 | 3750 | 20.54% |
| Lincoln | 3660 | 741 | 20.25% |
| Belmont | 14077 | 2750 | 19.54% |
| Sherborn | 2653 | 458 | 17.26% |
| Wayland | 8260 | 1245 | 15.07% |
| Winthrop | 8895 | 1323 | 14.87% |
| Malden | 20968 | 3100 | 14.78% |
| Framingham | 27704 | 4008 | 14.47% |
| Medford | 27401 | 3898 | 14.23% |
| Holliston | 8267 | 1175 | 14.21% |
| Stoneham | 12368 | 1747 | 14.13% |
| Watertown | 16716 | 2314 | 13.84% |
| Weston | 6621 | 893 | 13.49% |
| Woburn | 19243 | 2553 | 13.27% |
| Natick | 18642 | 2304 | 12.36% |
| Ashland | 8512 | 998 | 11.72% |
| Revere | 16897 | 1973 | 11.68% |
| Waltham | 25066 | 2880 | 11.49% |
| Southborough | 5804 | 622 | 10.72% |
| Sudbury* | 10482 | 1306 | NA |
| Cambridge* | 48787 | 4279 | NA |
Cambridge and Sudbury are only partially in the fifth district, so they are marked with asterisks - the 2012 total is city-wide.
The results are kind of gross. Malden (which Clark represents as a State Senator!) had 3100 votes cast, compared to nearly 21,000 in 2012. Medford, which for whatever reason I've always considered the "district seat" due to its large population and politically-connectedness, had less than 4000 voters. Medford could get 4000 voters if they announced were electing the town dump manager tomorrow - I've never seen anything close to that low in that city. Waltham, something of a bellwether in the state, had only 11.5% of its 2012 turnout! The big outlier was Winchester. Were they more politically active in Winchester, you ask? No, they voted on their new high school at the same time.
There is a very minimalist view of democracy, forwarded by Joseph Schumpeter and others, that essentially reduces democracy to free and fair elections. (Schumpeter specifically has a pretty detailed definition of "free and fair" but we shan't go down that road at this time). This, I think, is wrong - an elected tyrant is still a tyrant, and elections, even fair ones, can be non-democratic. Robert Dahl outlined additional rules for democracy, with one including
effective participation (emphasis added). Having an election in December, four weeks after municipal elections, months after another special election, failed on its face. A United States Representative was chosen by 40,000 people.
What is the alternative? Appointments and empty seats are unpopular, but perhaps sometimes necessary. The 2004 rule change was put into effect because an appointed Senator would have served all of Kerry's remaining term, until January 2009. I agree that an unelected appointee should not serve for four years as a Senator, but I also don't want to be sending people to Congress with less than 40,000 votes, So what is the alternative?
Here is my proposal.
1. Special elections are held on election day. That is, the Tuesday following the first Monday in November. Turnout is naturally higher then - people know that it is time to vote, and they do the proper preparation to do so.
2. Empty seats are filled by a governor's appointee between the open date and election day. I understand the push to institute a rule that the appointee must be filled by the same party as the departing representative, but I don't think it would do a lot of good - it won't take much for someone to change their own party designation and have the D or R label while voting the opposite. I also understand why people want these appointees to not be able to run for the open seat, so as not to gain an incumbency advantage, but I doubt the Constitutionality of such a rule. The enforcement of such a thing would need to come at the electoral level.
3. If there are no elections between the date a seat is opened and the date the seat expires, then there will be no special election. This may be seen as a flaw, but is actually an advantage. We seem quite worried about incumbency advantage here, but Clark is now going into 2014 as an incumbent after gaining only 40,000 votes. By contrast, Governor Patrick who received about 140,000 votes (very rough estimate) in the 2010 election from the communities that now compose MA-5, would be appointing the interim Congressman. Which is the undemocratic process?
It was a good idea to change the rule in 2004 - an appointed choice should not get a four-year term as Senator. But a one-year term is preferable to an election that features poor participation. A democracy is only as good as its rules, and the rules in MA-5 have produces a result that is insufficiently democratic.
