How Much of a Problem is the Senate?

by Michael C. Dorf

In the last week, various liberal law professors and others in whose circles I move have taken to using the midterm election results to decry the US Senate. They point out -- correctly -- that nine million more people voted for Democratic Senators than for Republican Senators; yet the Republicans probably gained at least one seat and at least held their edge. That's not exactly a fair comparison (for reasons described here), but it does capture the bigger picture: If we look at all three classes of Senators, we find that Republicans have more Senators, even though the Democrats represent more people.

Is that a problem? Well, if one is a Democrat (as I am) of course it's a problem. Republicans will continue to confirm very conservative judges and justices; and when there's a Democratic president again, Republicans' advantage in the Senate may enable them to block Democratic appointees (again). Meanwhile, should the Republican edge hold into the next Democratic administration (and even if it does not, absent abolition of the filibuster for ordinary legislation), it will permit Senators representing a minority of the country to block legislation favored by a Democratic House majority and a Democratic president. Thus, I share the dismay of many of my fellow Democrats at the impact of the Senate on the laws we have and how they are interpreted.

But I do want to raise a few questions about the current bout of Senate skepticism that rests on first principles.

Let's begin with the theoretical. A bicameral legislature in which one house is apportioned more or less based on population and the other based on political units seems like a reasonable design for a federal system. To be sure, one might not want a federal system. A federal system in which the various units represent distinct ethno-cultural groups may tend to civil war (Yugoslavia), breakup (Czechoslovakia), or resentment by the minority groups of the majority (Belgium, Canada). Meanwhile, where differences between states are modest, one might conclude that a federal system is unnecessary. One could still have decentralization of various responsibilities within a national unitary system.

However, there are reasonably well-functioning federal systems: for example, Australia, Canada (notwithstanding Quebec), Germany, Switzerland, etc. Where there are good reasons to have a federal system, then something like the Senate as one of the houses of the national government is, if not the only plausible approach, at least defensible.

Are there good reasons for the US to be a federal system? Maybe not today, but as a historical matter, the US as we know it would not have come into existence as a national unitary system. Indeed, it almost didn't come into existence even as a federal system, because even the structure with which we ended up was too centralized for a sizable chunk of the electorate.

Maybe the world would have been better off had the US been smothered in its cradle. In The Federalist, Publius warned that if the US fractured, its parts would be swallowed up by the European powers. That might have been for the better. For one thing, slavery might have ended substantially earlier than it did. But we can't possibly know how the world would have unfolded over the last 24 decades if the US did not survive as a unit. At the very least, we can pretty confidently say that, given butterfly effects, neither I nor any of the people reading this blog would exist in such an alternative world.

Moreover, even viewed from a presentist perspective, federalism has virtues. If Matt Whitaker terminates or hobbles the Mueller investigation, federalism allows newly elected NY AG Tish James to use state authority to investigate crimes that Trump and his minions may have committed. More broadly, federalism works as a hedge. For me, living in a very blue town in a blue state substantially mitigates the impact of national policies.

My point is not that American federalism is the best of all possible systems or even all-things-considered justified. My point is simply that American federalism has benefits as well as costs, and insofar as the Senate is an accommodation for federalism, it accommodates the benefits as well as the costs.

To my mind, there are two main problems with the US Senate. First, the ratios between the largest and smallest states are just too large to justify exactly equal representation. And second, Article V makes it effectively impossible to adjust those ratios to give small states less, but still disproportionate, representation. Ordinary amendments to the Constitution are extremely difficult. Changing the Senate is effectively impossible, because Article V requires every state's consent. The small states that would lose any of their substantial over-representation will not consent to an amendment that reapportioned the Senate on anything closer to a population basis.

Is there a way around that? In recent years, judicial review of constitutional amendments has spread around the globe, so that constitutional courts in various countries have developed doctrines that allow them to say that a provision amending the constitution is invalid because it deviates too far from core principles. I am extremely dubious of the prospect of the US Supreme Court (or even a single justice) adopting such principles, but even if I thought that possible, it's very far-fetched to believe that any justices would say that a central provision of the original Constitution (such as the Senate or the Equal Suffrage Clause) is itself unconstitutional.

As a practical matter and for the foreseeable future, we are stuck with the Senate. We may or may not also be stuck with the Electoral College, which allows minority vote-getters like G.W. Bush in 2000 and Trump in 2016 to win election, and uses a formula that overweights small states, albeit not nearly as much as the Senate. The EC can be eliminated by an "ordinary" constitutional amendment and perhaps circumvented via an Interstate Compact.

Meanwhile, does the continued existence of the Senate spell doom for Democrats? In a word, no.

The Senate over-represents small states. It does not--or at least not inherently--over-represent Republicans. For one thing, 4.5 of the ten smallest states by population (VT, DE, RI, 1/2 ME, NH) are represented by Democrats or Independents who caucus with Democrats (Bernie Sanders and Angus King). If we follow the Spinal Tap Rule and go to 11 (thus including Hawaii), we find an even split of 11 Democrats and 11 Republicans representing the 11 smallest states.

To be sure, it's true that the overall impact of the Senate favors Republicans, mostly because of how heavily concentrated Democrats are in California, Illinois, New York, and the mid-sized states of New England, whereas so many of the smallish but not tiny states are strongly Republican. But even that's not inevitable. To see why, let's think about some basic political science and economics.

Duverger's Law holds that a first-past-the-post electoral system will lead to two dominant parties. For much of the 20th century, the two parties were ideologically close to each other. That, in turn, is what one would expect from Hotelling's Law.

If this is unfamiliar to you, consider a fairly standard illustration of Hotelling. Imagine a beach that is 300 meters long and two mango vendors competing to sell bathers fresh mango snacks. To maximally serve customers, the vendors would be situated at the 100-meter and 200-meter points. However, that's not a stable equilibrium. Vendor 1 (at the 100-meter mark) will realize she can eat into Vendor 2's market by moving closer to Vendor 2. By doing so, she will still get most of the business from her side of the beach and she will now get some of the business from the other side of the beach. Likewise Vendor 2 will make the same realization. The result is that eventually the vendors will cluster at the 150-meter mark. Don't believe the theory? Visit the flower district, the diamond district, etc.

That's Hotelling in economics. Politics works roughly the same way, except that political "markets" are bifurcated. To sell your political product to the general public, you must first sell it to primary voters. In recent years, competitive primaries, especially in the Republican Party, have dragged the parties away from the general election midpoint. So has the tradeoff in the general electorate itself between appealing to independents (by going close to the midpoint) versus mobilizing the base for turnout. If the Democrats and Republicans were at 145 and 155 respectively in the mid-twentieth century, this new dynamic has increased the spread between them.

But here's the thing: The parties' ideological positions are dynamic. There was nothing inevitable about one party that is socially conservative and economically libertarian versus another party that is socially liberal and economically (modestly) redistributive. And needless to say, there are many more than two dimensions along which we could slice politics. Over time, we can expect both parties to adjust their positions in order to maximize their respective turnout of shifting populations of base voters and their appeal to independents/centrists. If Republicans are getting 70% of the vote in Kansas by standing at the 250-meter mark (but not at 300, see Kris Kobach!), expect Democrats to move to at least 150 and maybe something more like 175.

To be clear, I'm NOT taking a position on the battle within the Democratic Party between, for lack of better terms, the Clinton and Sanders wings. The formula for appealing to both activists and swing voters could be to move left on some key issues, partly because existing divisions may not map clearly onto left versus right. My point is not that there will be any particular change in positions but that there will be change. A party that consistently loses under the existing rules and cannot change the rules (as we cannot, per the Equal Suffrage Clause) will change what it stands for in order to attract new voters.

That's not necessarily a pretty prospect for a certain sort of liberal (like me). The Democratic Party that ends up successfully appealing to voters we now think of as solidly Republican will be truly populist (by contrast with Trump's mostly faux-populism) but also could be substantially more friendly to racism, xenophobia, and other ills than the current version of the Democratic Party that loses too many of those voters.

So yes, the Senate is a problem, but it is not a problem that systematically disfavors Democrats, except in the short run. Although it is common these days to refer to political attachments as tribal, and although there is no doubt some truth to that characterization, over the long run the parties are mostly just means of grouping and packaging issue platforms.