Some Thoughts on Representation
How well do single-seat districts lead to representation? (And of what?)
The discussion of single-seat districts and pending redistricting cause me to think about an objection that is often put forward in opposition to multi-seat districts, (as opposed to our current system of single-seat districts): who would my representative be? After all, right now we all know that if I live in California’s 12th District then I am represented in the House by Nancy Pelosi and if I live in Texas’ 1st District then I am represented by Louie Gohmert.
But in terms of “representation” how true is this, really?
There are conservative Republicans in CA12 who, almost certainly, feel better represented in the Congress by Gohmert than they do by Pelosi (11.9% of CA12 voted for Trump). The converse is true for folks in TX01 (where 26.2% voted for Biden).
Apart from the fact that yes, she is the Representative of California’s 12th Congressional District, in what real way does Nancy Pelosi represent a hardcore conservative constituent? The dedicated libertarian? And how often over the years has there been a vote wherein her CA12ness was more important than her Democratic POV?
I recognize there is an important linkage between place and legislative representation, but the reality is that the ideological/philosophical/policy outcome kind of representation is also important (arguably, more so). Both single-seat and multi-seat districts represent both place and ideology (which also have linkages) but single-seat districts emphasize place over philosophy while multi-seat PR emphasizes philosophical representation.
Ultimately, however, what matters to you as a citizen? That a vote was favored by a majority of persons in your district or that legislation you favor gets passed? Is it “good” for your district for a law to pass with 51% support of the district if your representative voted for that bill? What about the 49% who opposed?
For all the energy we spend on talking about how important place is (whether we are talking about states or just huge tracks of land we color blue or red within those states) for most national issues is it really the case that one’s political preferences are really represented by one’s US Representative?
If one is a QAnon devotee, one may be far better represented in Congress by MTG than by one’s actual Rep. Likewise, AOC may be far more representative of one’s views that than the fellow one actually voted for (or against, as the case may be).
The point being that if, as per the example in James’ post on redistricting, there is a population with 60% preferences for one party and 40% the other, then in a winner-take-all system, 60% are represented in the way I am describing above, and 40% are not, even if the shared geography of the two groups has a representative in Congress. So, sure, if there are issues that are purely about my district and not about ideology, I am represented, but how often does that actually happen?
We are taught that geography-based representation is central, but the reality is that most issues are more likely to be denominated by philosophical and other variables more than being reducible to location-based preferences.
The fundamental lesson here is that a given district does not have one interest. Far from it.
As I frequently note, people have interests, land does not. We should want a legislature that represents those interests and proportional representation, that usually requires multi-seat districts (although MMP, my top preference, does not). I do think that we Americans are so engrained with the notion of location-based representation and convinced of place-based interests being paramount that we often have a hard time understanding the limitation of that kind of thinking as well as the options that can overcome those limitations.
Some other related observations.
First, the real source for local interests to be contested is local government (which we all tend to ignore). And yes, those suffer from the effects of single-seat districts as well (although there is some variation in local elections across the US). While national legislation can have local effects, they are diffused. Real change to issues that affect localities is handled by state and local government (e.g., law enforcement, education, sanitation, utilities, and so forth).
Second, some worry that multi-seat districts would mean they would not have their own representative, which would affect constituency services. But, of course, everyone has two Senators to choose from to seek help with problems, and rather than causing confusion, it simply increases the number of entry points for a person seeking help. As such, living in a district with five members instead of one just means more possible routes for representation and aid, which does not strike me as a problem.
Beyond that, how many people actually solicit their Representative for services? While a real and important thing, I also think this concern is overblown.
Third, an illustration of place v. interest: one of the reasons I think that some “small” state would gladly do away with the Electoral College if given the chance is that the majority of citizens and politicians in, say, Vermont, would prefer for the popular vote to prevail because they heavily prefer Democratic presidents and the EC is more likely to help a Republican get elected than a Democrat. We are so taught to think it terms of interests being linked to real estate that we often don’t stop to think about whether that is the best way to think about it.
(The fact that a number of small states have voted for the National Vote Compact lends credence to this conclusion).
Fourth, a comparative side note: it is possible to have a legislature elected in one, country-wide multi-seat district, as is in the case with the Israeli Knesset and the Colombian Senate. (Israel is the purer case as their legislature is unicameral, while Colombia’s is bicameral, with its first chamber being elected in geography-based multi-seat districts). In these cases, the philosophical variable is paramount, although it is possible in such a system for regional parties to draw their votes from location-based constituencies if they are sufficiently organized and cohesive.
If a geographic, single point of contact for constituent services is desired, simply assign a congress critter to a region of the state for that purpose.
The opposition to multi seat districts 1. Entrenched interests and 2. Fear of change.
The problem seems to be that politicians like some sort of proportional representation when they’re out of power, but quickly see the advantages of first past the post (ie single-seat districts) once they’re in power (the main advantage is that its the system that put them into power).
That’s not a particularly American phenomena, it seems to be true in many parts of the world. Even parties that campaign on replacing first past the post (like Trudeau’s Liberal Party in Canada two elections ago) change their mind when in power (he received a majority in Parliament with 40% of the vote, and decided that was better for the country than his having only 40% of Parliament seats).
To change the system you need the support of the very people put in power by the old system. Its achievable, but hard to do. And on top of that, there are quite a few variations of proportional representation, and the people who want to replace first past the post disagree on the specifics of what to replace it with.
@Northerner: It is absolutely the case that politicians favor the system that provides them with power. This is, without any doubt, a major obstacle to reform.
I’ve been “represented” by what would become Tea Party Republicans since I moved to Cincinnati 40 years ago. Rob Portman claimed to represent me for a while, now I have FL GOP Greg Steube.
@Steven L. Taylor: Rank and file GOPs still talk about term limits as some sort of panacea. I don’t hear GOP pols and pros talk about it much after they attained a majority.
My experience with multi-seat districts at the local level in the Philly suburbs seems very different than what is implied here. For township supervisor and school board, we have multi-seat elections, but a straight ticket of one party or the other has won nearly every race in the 20+ years I have lived here. The 2019 school board races were an exception, as a multi-party slate (including some sitting board members elected as Democrats) ran against a ticket endorsed by the Democratic Party. The multi-party slate won (with my support, since if the Democratic-endorsed slate’s main argument is that their opponents aren’t all Democrats, that doesn’t say much in their favor).
I think part of the push-back for multi-seat districts is that many states have a large urban vs. rural divide, and the rural areas are afraid of a straight ticket from the urban areas winning all the seats. So you need more than multi-seat districts, you also need proportionate representation.
The United States is 473 times the size of Israel–which is 14% smaller than DFW. Israel is all one location.
Colombia is 1/3 smaller than Alaska.
While I may be closer to Nancy Pelosi philosophically than I am to Ron Johnson, I don’t believe that Pelosi would be looking out for Wisconsin in any way. At the very least, Johnson is accountable directly to Wisconsin.
Location in the US is widely more important that in most countries in the world. The needs of California, Texas, Vermont, and Wisconsin are very different. They have different climates (dictating different energy needs), different geography (dictating such things as water rights), and different industries. Even with the same broad industry–say agriculture–farmers in CA, TX, WI and NE have different concerns.
I’m not saying that the system we have is the best it could be, but it’s disingenuous to suggest that geographic location isn’t important in representation in the US.
@Moosebreath: I would have to know the electoral rules to be able to comment. I am speaking here of multi-seat districts in large legislative bodies elected via some version of PR.
@Mu Yixiao:
And I was not suggesting such a system for the US, although the Colombia-to-AK comparison misses the population differential.
Well, as noted, geography has a role. But I think that “looking out for Wisconsin” is almost certainly an overblown notion inflated by the dynamics I describe in the post.
Taxation levels, health care governance, military actions, and so forth, i.e., a lot of federal policy, is really about national policy goals over which we differ, and had less to do with state-specific interests.
And look, my point is not that there are zero location-based interests, but there are assuredly fewer of them than the national dialog (and mindset) tends to suggest.
Well, that wasn’t what I said, so luckily not disengenous!
@Steven L. Taylor:
And all of those policies get tied back to where the products are manufactured, where the jobs go, where the bases are, etc. Those are location-based concerns that affect how representatives approach the national policies (though not always for the better–The reason we have the SLS rather than something competitive with SpaceX is because one senator was looking out for his state).
I’ll bet if you dug into it, you’d find that the choice for the new Navy destroyers was a big issue for Wisconsin representatives–they’re being built in Sturgeon Bay. And for those from Virginia–they’re not being built in Norfolk.
That sounds like you’re saying location isn’t important. I disagree.
@Mu Yixiao:
To say that one thing is, on balance, more important than another is not the same things are saying that something “isn’t important.”
I actually don’t think most issues are reducible to location-based preferences, do you?
That’s cool, of course. I just think you are being more didactic in your interpretation of my position than warranted, but that’s cool as well.
@Mu Yixiao: And keep in mind: the underlying argument would be for geographically-delimited multi-seat districts that would still have state-specific representation as part of the mix, just not single-seat districts.
Perhaps this was not clear.
@Steven L. Taylor:
“I would have to know the electoral rules to be able to comment. I am speaking here of multi-seat districts in large legislative bodies elected via some version of PR.”
That’s sort-of my point. The original post spends much more time on the virtues of multi-seat districts than it does on the virtues of proportional representation. Multi-seat districts without PR don’t seem to end up much different than single seat districts.
@Moosebreath: I was assuming PR in my discussion, which clearly was not clear.
The point was supposed to be more conceptual in any event.
IN fairness, I did say in the OP, “but single-seat districts emphasize place over philosophy while multi-seat PR emphasizes philosophical representation.”
I think your ideas have fallen victim to the curse of “stream of consciousness” (i.e., “blogging”). 🙂
You really can’t cleanly compare single-seat “winner-takes-all” districts with multi-seat PR districts. You’re bringing in too many variables at once. Each of those variables is worth discussing, but they need to be separated for the purpose of discussion.
@Mu Yixiao:
Well, my academic discipline says otherwise. 😉
I can see the vague outlines of a path from this sentiment to “money is speech” — where someone donates to the candidate who will best represent their interests, and who will represent you, and give you access, because of that donation.
I agree with what you wrote. Don’t get me wrong. I just see that destination and it’s late and I’m tired and I don’t see how we don’t get there in America.
I hate they “money is speech” conclusion, and I think proportional representation is the better end destination, but way harder to get to in the US than “money is speech”.
@Steven L. Taylor: SNAP!
@Gustopher: In my attempt to make a point about the fact that representation is more than geography, I think I have gotten readers to focus too much on the decoupling rather than simply getting people to think about how representation is simply more than just being from the same place.
(And definitely not advocating “money is speech” here).
This is anecdotal and on a much smaller scale, but here goes. My city (pop about 80k) moved from a ward based system for councillors 20 years ago to city wide election for the councillors. We have been through 5 election cycles now. What seems to happen is the person who finished 8th in the prior election (we have 8 councillors) gets defeated in the next election and the top 7 stay. The infighting and corruption never seem to stop. The city has spent a fortune on lawyers to investigate and prepare cases involving disputes between councillors for outside adjudication. The council is dominated by local hotel owners, who have permitted a special tourist tax that the hotels are allowed to keep.
Our region (about 450 k) operates on a ward system. The previous council also had serious problems with disputes and corruption involving improper hiring of friends and expenses. In the last election the wards where the worst offenders were based voted each out individually. I am not sure if this would have been possible when voting for 16 people on a ballot listing several dozen candidates.
Again, just anecdotes, and I am not sure if these observations are scalable to a whole country.
@Steven L. Taylor:
Representation has always been more than geography. Absent other changes, such as a true multi-party system, diminishing the importance of geography to the extent your suggesting is something I’m very skeptical of. Candidates would be incentivized to only seek to represent those with a similar ideology which, in our current political culture, means that candidates would focus more on culture-war issues. It seems to me that all it would do is encourage more polarization and dysfunction.
Here in my own state of Colorado, we have a wide diversity of interests that are driven in large part by geography. Tourism and resource-based interests in the western mountains, agriculture on the easter plains, tech, services, higher-education and urban concerns along the front range. It’s already difficult enough to cram those diverse interests into a two-party binary system that is already overly-focused on culture-war issues, but removing geography from representation in this state would be a disaster IMO. It wouldn’t help, for example, liberals who live in more conservative mountain communities who may share the social and cultural values of progressives in Denver, but don’t share the bread-and-butter priorities that are necessarily different between the mountain communities and the front range.
This would all be different in a multiparty system where geographic interests could be indirectly represented much better. For example, you might have a party with candidates that specifically represent the particular interests of mountain communities, or urban areas, or the plains or suburbs or whatever. You don’t and can’t get that in a binary system where the two parties are primarily divided and motivated by culture-war issues.
So again, I think the flaw in your proposal is the assumption (wrong, in my view) that better party representation between Republicans and Democrats in terms of math at the state level is inherently better than at the district level. I can see how it would better for the party loyalists, culture war idealogues, but absent a diverse multiparty system, it wouldn’t be better at incentivizing the two parties to consider the diverse interests in this country – rather it would have the opposite effect in my view.
Another way to look at this is to take this idea to its logical conclusion – suppose geography was removed completely from House elections and all Representatives were elected nationally with the top 435 getting seats. If mathematically perfect partisan representation is your goal, then that would do it, but I think the downsides should be pretty obvious.
@Andy:
Well, first, I really wasn’t making a proposal–I was trying to make a conceptual point that I may or may not have successfully made.
Second, my point is that a multi-seat district proportionally allocated is, by definition, and in every way more representative than a single-seat district.
@Steven L. Taylor:
I don’t disagree. And if we made the United States into one big single 435-seat district, it would be even more representative. Which would be super-awesome, right?
But representative of what? Would it better represent the actual preferences of the population, or represent the artificial way those preferences are forced into two competing camps? Clearly it’s the latter.
My point is that some mathematical improvement in ideological partisan representation isn’t that important when “representation” is constrained by only having two choices. It’s not even clear that it would make our binary system operate any better – indeed, as I argued above, I think it would make it worse. But more important to my point is my fundamental view that our binary-choice system is inherently unrepresentative, and no amount of tinkering intended to make the partisan split more representative can change that.
@Andy: Honest clarification question: are you assuming I am arguing for better representation while keeping the only 2 parties?
Because if we went to multi-seat districts and PR we would no longer have a party duopoly.
I think this touches on something you said in a previous comment thread and thought I corrected then: while I talk about Rs and Ds (because that is what we have now), I am never arguing for just two parties when I talk about PR.
Yes, I talk about some reforms (the less good ones) that would improve representation under our current system (that would retain the parameters that create the duopoly), but that is assuming tweaks and not a significant overhaul.
@Steven L. Taylor:
Well, yes, that’s how you frequently frame the arguments, including in this post, in terms of partisan representation not matching the supposed partisan preference expressed by vote totals.
That said, I do know that you would prefer a multi-party system – and so would I.
Fair enough!
It just appears that your posts spend a lot of time and effort on systemic improvements where representativeness is measured in partisan terms, including discussions on PR. Perhaps I am overly focusing on that.
It’s probably clear that I have a pretty strong skepticism when it comes arguments that appear designed to change the system to match partisan preferences. I am very much against that sort of thing.
@Andy:
Well, when I am talking about existing partisan divides, I am not sure how else to frame the discussion.
I find a huge hurdle in any of these conversations is getting an American audience to think beyond that which exists here already.
This leads to trying to align understanding of broader concepts with known political reality.
To be honest, I think you read my posts on this subject far too narrowly. I will certainly have to take some responsibility for lack of clarity, but I would argue that I am pretty clear, especially to long-term readers, as to my preferences, goals, and intentions.
(And, of course, when I get to speculative/hypothetical, I am told I am engaging in fantasy).
All well and good, and I take the point. But, it is empirically true almost any likely reform is going to help one party and hurt the other, so the issue shouldn’t be “who does it help?” but “what good does it do (or not do)?”
Most possible pro-democracy reforms are also pro-Democratic Party. But the important part of that sentence is “pro-democracy” while “pro-Democratic Party” is ancillary to me.
Now, I know that it is not ancillary to Democratic partisans, but if that means they will further the cause of small-d democracy, then their partisan motivations are a good thing as far as I am concerned.
@Andy:
Yes, because the main way one measures the representativeness of an electoral system is the partisan results it helps produce. It is central.
While I understand that partisan results are not the entirety of the discussion, where else would you have me focus?