Should Howard County elect council members at large?

This is my fifth and final post in Dividing Howard week on my blog, as I discuss some topics related to my new book on the history of county council redistricting in Howard County, Maryland, and the broader events of Howard County politics from 1960 on. Previous posts discussed the role of Columbia in spurring creation of a county council, the struggles of Howard County Republicans under the council district system, the problems with gerrymandering of council districts, and whether it’s possible to make redistricting less political. In today’s post I consider whether it would be preferable to go back to the previous method of electing council members at large.

As I noted in my discussion of selectorate theory, as a general principle it makes sense to broaden as much as possible the pool of voters who can meaningfully participate in electing leaders, so that those leaders will need to put together winning coalitions that are a significant fraction of the total voting population. Once elected such leaders would then be more likely as a general matter to pursue policies of benefit to everyone and not just to a relatively small band of supporters.

That general principle would lead us to require that a council member attract votes from people all over the county, and not just from those living in a relatively small district. As described in the early chapters of Dividing Howard, Howard County’s first charter required council members to be elected at large (in fact, no other scheme was permitted at the time by Maryland’s constitution), and the first five county council elections (1969, 1970, 1974, 1978, and 1982) were at-large elections.

Why not revert to the original system? There are multiple objections I can think of that need to be addressed. The first objection is that districts are needed to ensure diversity of the council, usually interpreted as racial diversity. This is the same argument recently used in support of the proposal to elect Howard County school board members by districts instead of at large. It’s motivated by the fact that at-large elections have historically been used in many jurisdictions to dilute minority voting power, in particular to ensure (in combination with white bloc voting for white candidates) that no African-American candidates are elected to at-large positions even where African-Americans form a significant portion of the voting population. The question of whether this argument is relevant to Howard County has both a practical and a legal dimension.

Practically speaking I don’t believe that an at-large system would necessarily be disadvantageous to African-American or other minority candidates. C. Vernon Gray was elected as the first African-American council member in an at-large election in 1982, and today I have no doubt that someone like Calvin Ball would be able to win election to the council on an at-large basis. Maybe I’ve missed something, but in modern times Howard County just doesn’t appear to have had the type of racially-motivated bloc voting, especially white voters voting as a bloc to reject black candidates, that has been characteristic of many other jurisdictions.

As discussed in chapter 23 of Dividing Howard, in 2001 African-Americans were only 23% of the Council District 2 population, yet local activists saw that as no barrier to electing an African-American council member to replace C. Vernon Gray; as Jared Thornton noted at he time, “A lot of things about Columbia seem to be different from any other place. We don’t need a super-majority in Howard County.”1

However whether (re)introducing at-large council elections would pass legal muster is an entirely different question. Changes made to the Voting Rights Act in 1982 (coincidentally, the year of Howard County’s last at-large council election) tightened up the criteria under which at-large schemes could be deemed discriminatory, and in particular did not require actual intention of discrimination. Thus even if racial motivations were not behind an effort to change council elections to be on an at-large basis, such a change could still face a legal challenge on racial grounds.

However in 2009 in Bartlett v. Strickland the U.S. Supreme Court held that the relevant provision of the Voting Rights Act affecting by-district vs. at-large elections did not apply unless minorities constituted an actual majority in the area in question. Since this is not the case in Council District 2 (or indeed in any council district in Howard County) it may be that a change back to at-large council elections would be relatively immune to legal challenges.2

Beyond its affect on racial and ethnic minorities, another issue with moving to an at-large election system is its effect on the balance between the Democratic and Republican parties in Howard County. It’s a common complaint today that Republicans are under-represented on the county council relative to their share of registered voters: As of the 2010 general election Democrats were about 48% of registered voters, Republicans 31%, and independents 21%, with Democrats thus having a 1.56-1 registration advantage over Republicans, equivalent to 61% and 39% shares respectively of voters registering with the two major parties.3

Looking at election data instead of registration data, in the 2010 general election Democratic council candidates collectively received about 56% of all votes cast for council candidates compared to 44% cast for GOP council candidates, with Democratic council candidates under-performing a bit based on the Democratic registration advantage. If the county council reflected this division then we should expect the GOP to have two seats instead of their current one.4

Would electing council members at large correct this situation? I suspect that it would not, unless it were combined with additional changes to the voting system. In the type of at-large elections held in Howard County, both in the past for county council and at present for the Board of Education, the top set of vote-getters (e.g., top five for county council) are elected. (Political scientists refer to this as a multi-member district plurality system.) In such a system parties can run slates of candidates, and if voters select candidates along party lines then it is possible that all candidates selected in an at-large election would be of a single party.

For example, in Howard County if the 55% of voters who voted for Democratic council candidates in the 2010 general election were instead to vote as a bloc for a Democratic slate of five candidates in an at-large election, no Republican council members would be elected at all. As described in chapter 5 of Dividing Howard, this is pretty much what happened in the last at-large council election in 1982, with Democrats winning all five council seats and the only GOP candidate (Charles Feaga) being shut out.

Correcting this situation, so that the party composition of the council better reflects the party composition of the electorate, would require not just the abandonment of council districts but also the introduction of a voing scheme for proportional or semi-proportional representation. To go back to the 1982 Howard County council election, being the only GOP candidate didn’t help Charles Feaga: The best Republican voters could do was to vote for Feaga only, and no other candidate; however since Democrats significantly outnumbered Republicans this was insufficient to counter the effect of Democrats voting a straight ticket for five Democratic candidates.

One way to address this issue is to allow voters to cast multiple votes for one candidate, so that, for example, a GOP voter in 1982 could cast five votes for Charles Feaga instead of one. This so-called cumulative voting system was actually considered for use in Worcester County, Maryland, back in the 1990s to address the discriminatory effects of an at-large system on black voters. Other possible systems would have people vote for parties (not candidates) and then allot council seats on the basis of the total vote received by each party (a party list system) or allow users to express preferences between candidates (e.g., a single-transferable vote or STV system).

My overall point is that an at-large system in and of itself, especially like the one previously used in Howard County, would not necessarily address the complaints that Howard County Republicans have about the current district system. Neither would a system that combined, say, five council members elected by districts with two at-large members. Electing two members at large would help ensure expansion of the set of voters able to select a council majority (in line with my discussion of selectorate theory in a previous post), but given the Democratic advantage in voter preferences it’s quite likely that the two at-large members would always be Democrats, converting the present 4-1 Democratic council majority into a 6-1 majority.

This simple fact is that no voting system is perfect or can be perfect, in the sense of correctly reflecting all voters’ preferences and not producing results that seem to be contradictory to common sense. If people want to reform the way Howard County Council candidates are elected (a goal with which I’m sympathetic) then they’ll need to take the time to properly sift through the alternatives and (most important) build a strong case to the people of Howard County as to why such a change is necessary. Recall from Dividing Howard that it took at least five years to convince voters to move from the previous commissioner system to a county executive and county council, and over ten years to convince voters to replace the at-large system with council districts. For anyone interested in avoiding having another round of council redistricting in 2021, the time to start working is now.

One reason I decided to write my series of blog posts on Howard County council redistricting was to provide a sense of perspective about the issue that I think is badly needed. One reason I turned the blog series into the Dividing Howard ebook was to make it available for future readers who might be interested in the topic when the next round of redistricting occurs. If you haven’t already bought a copy, while wait until 2021? Dividing Howard is only $2.99 from Amazon or Barnes and Noble, and all royalties go to the local charity Voices for Children, which recruits and trains volunteer advocates to represent the best interests of abused and neglected children in Howard County courts.


1. Unfortunately I couldn’t find a current figure for the African-American population of Council District 2, but based on a quick check of Census data for census tracts within District 2 I suspect the proportion of African-Americans in the district is about the same as in 2001, about one quarter; the highest proportion in any one tract is 35%. (I took the figures from the 2010 Census interactive population map.) In comparison, the current African-American population of Howard County as a whole is 17.5% (American Fact Finder, Table DP-1, Profile of General Population and Housing Characteristics: 2010, 2010 Demographic Profile Data).

2. A good review of the legal issues around at-large elections, including the effect of the 1982 changes to the Voting Rights Act, is “At-Large Electoral Systems and Voting Rights” by Sidney Hemsley. Unfortunately however it does not discuss Bartlett v. Strickland.

3. Registration data is from my Howard County general election turnout spreadsheet. This in turn is based on data from the Maryland State Board of Elections for 1988, 1990, 1992 (for all voters, the Democratic, Republican, Libertarian, and Alliance parties, and unaffiliated voters), 1994, 1996 (for all voters, the Democratic, Republican, Libertarian, Reform, Natural-Law, and Taxpayers parties, and unaffiliated voters), 1998, 2000 (for all voters, the Democratic, Republican, Libertarian, Reform, Green, and Constitution parties and unaffliated voters), 2002, 2004 (for all voters and broken down by party), 2006, 2008, and 2010.

4. In the 2010 Howard County general election Democratic council candidates collectively received a total of 57,131 votes compared to a total of 45,590 for all GOP candidates combined. The totals are based on the official results for the 2010 general election as published by the Howard County Board of Elections.