In recent weeks, a slew of think tanks and journalists have published reports or articles arguing that the American political system is in dire need of “democratization.” These pieces make a variety of claims, but their basic thrust is simple: America’s political institutions, as currently constituted, do not reflect core democratic principles like political equality and majority rule. While they point to many democratic deficiencies, the authors of these pieces give special attention to the U.S. Senate, considered to be an especially egregious anti-majoritarian institution due to the fact that the fifty states, regardless of their very different population sizes, enjoy the same level of representation within it.
One might assume, based on reading these articles, that the U.S. Senate is an unusually anti-majoritarian legislative chamber (and, by extension, that the U.S. Congress is an unusually anti-majoritarian legislature) when compared to other national legislative bodies. After all, as many of these articles take pains to point out, the U.S. Congress was designed over 230 years ago by wealthy white men who did not share our modern views about what popular representation should look like. From this perspective, the Congress (and particularly the Senate) looks like the undemocratic relic of a long bygone era. Other countries tend to have far younger parliaments, most of which were designed during the 20th century, when a consensus on basic principles like citizen equality and majority rule was emerging in much of the world. It stands to reason, therefore, that the world’s other democracies are likely to feature legislative chambers that look much more small-d democratic than the Senate. But is this in fact true?
To answer this question, we need to examine three features of modern national legislatures: how many chambers they have, how upper-chamber legislators are selected, and (if a country has a two-chamber legislature) how the powers of the upper chamber compare to the powers of the lower chamber.
Number of Chambers
The United States has a bicameral (i.e., two-chamber) legislature, something that is very common in the rest of the world. According to a fairly recent study, 77 (or 41%) of the world’s 189 national parliaments are bicameral. More importantly, double-chambered legislatures are much more common than single-chambered legislatures among highly-populated countries like the U.S. Nearly all of the world’s most populous democracies have bicameral legislatures in which both chambers hold important powers. In this regard, then, the U.S. is clearly within the global norm.
How Upper-Chamber Legislators Are Selected
Selection procedures for national legislatures vary substantially across the world, but at least among the world’s democracies with bicameral legislatures, lower chambers (including the U.S. House of Representatives) are almost always selected via popular elections according to a population-equality standard (i.e., either the districts from which legislators are elected have equal populations, or the number of seats per district is proportional to district population). Thus, the selection procedures used for nearly all lower chambers are broadly consistent with the basic principles of democracy. Some might argue that gerrymandering of lower-chamber legislative districts, in the U.S. as well as elsewhere, is inconsistent with democratic principles (indeed, the U.S. Supreme Court has said as much on multiple occasions), but that topic is beyond the scope of this essay.
Unlike lower chambers, upper chambers are often selected via mechanisms designed to facilitate the representation of subnational units of government (i.e., states, provinces, or administrative subunits). This means that upper-chamber selection procedures are more likely to be in tension with modern democratic norms regarding popular rule and citizen equality. However, the exact process by which upper-chamber legislators are selected, and the extent to which facilitating jurisdictional representation leads to deviations from the population-equality standard, varies tremendously across countries.
The table below compares selection procedures for the upper legislative chambers of eight of the world’s most prominent democracies, including the United States. The table’s second column briefly describes the actual selection process, the third column indicates whether a country’s subnational units are equally represented in the chamber, and the final column shows my calculations (based on the most recent data I could find) of a statistic called the Loosemore-Hanby Index, which political scientists use to measure how malapportioned a legislative chamber is (basically, how much representation within it deviates from a population-equality standard). The index ranges from 0 to 1, with 0 indicating a chamber where representation is perfectly proportionate to population and 1 being a hypothetical chamber where the only districts receiving representation have no people living in them at all.
Let’s begin with the second column. Examining it, one can see that the United States Senate uses a selection process (popular election by the people of each state) that is actually more democratic than do the upper chambers of France, Germany, and India. In those countries, political elites rather than ordinary citizens select upper-chamber legislators. Indeed, this is the way the U.S. used to do it: prior to the ratification of the 17th Amendment in 1913, state legislators (rather than citizen voters) selected U.S. senators.
A few years ago, a movement was afoot among conservatives to repeal the 17th Amendment and return to state legislative selection of U.S. senators. This movement was widely ridiculed by the mainstream media as a silly effort to return the U.S. to the undemocratic dark ages, but it is worth noting that France, Germany, and India all adopted their current constitutions in the mid-20th century, long after the U.S. ratified the 17th Amendment. The framers of the French, German, and Indian constitutions were modern 20th-century leaders, but they nonetheless believed in the value of giving local jurisdictions a direct voice in the national government by having officials in local governments choose upper-chamber legislators. The examples of these countries suggest that, even in the modern world, there may be competing principles besides securing a democratic process that should be given at least some consideration in the construction of legislative bodies.
Turning to the table’s third column, we see that three other countries (Australia, Brazil, and Mexico) largely replicate the U.S. model of giving equal representation to their states in their upper chambers, regardless of state populations. Once again, all three of these countries adopted their constitutions long after the United States. Interestingly, the countries in which local officials choose upper-chamber legislators (France, Germany, and India) are also ones that do not grant equal representation to their localities. Only one country in my sample — Japan — has what some might consider to be a truly democratic upper chamber, i.e., one in which legislators are elected by ordinary citizens and jurisdictions with more people have more legislators.
Of course, giving equal representation to a country’s localities will have different effects depending on how unequal the distribution of the national population is across those localities. The table’s fourth column shows that upper-chamber malapportionment is consistently higher in those countries that grant equal representation to localities, but that it is especially high in the United States and Brazil, where the population differences between the most-populous and least-populous states are very large. That said, malapportionment is clearly also an issue in the Australian and Mexican Senates, where the differences in state populations are not quite as substantial.
One important difference between the U.S. and the other countries that grant their localities equal representation in their upper chambers is that the U.S. does not grant any upper-chamber representation to the people living in its national capital district. In Brazil and Mexico, the national capital districts encompassing Brasilia and Mexico City have exactly the same representation as the Brazilian and Mexican states in the national senate. In Australia, the two subnational units with the status of territories rather than states (including Australia’s Capital Territory) have, since 1975, been allotted two senators in the Australian senate (Australia’s states each have twelve).
While the U.S. policy of not giving the 700,000 residents of Washington, D.C. any representation in the Senate is clearly indefensible from a democracy perspective, whether the Australian approach (giving the capital district some representation, though less than that of states) or the Mexican/Brazilian approach (giving the capital district equal representation to the states) is more democratic is a complex subject. It is important to remember that the Australian Capital Territory is less populous than any of the Australian states, while the Mexican Federal District is actually more populous than nearly any Mexican state and the Brazilian Federal District roughly resembles a mid-sized Brazilian state in its population. The U.S. looks more like Australia than Brazil or Mexico in this regard: Washington, D.C. is much less populous than the vast majority of U.S. states (only Wyoming and Vermont have fewer people than D.C.). Consequently, if D.C. were added as the 51st state (as many Democrats want to do), this would actually exacerbate the U.S. Senate’s malapportionment problem: based on my calculations, the U.S. Senate’s score on the Loosemore-Hanby Index would increase from 0.365 to 0.372, just a notch below Brazil.
Given that the Australian approach (giving a national capital district some, though not full, representation) is not allowed by the U.S. Constitution, arguably the most democratic option available for rectifying the disenfranchisement of D.C. residents in the U.S. Congress would be to have Maryland or Virginia annex the vast majority of D.C. and make D.C. residents Maryland (or Virginia) residents. D.C. would then become a Maryland (or Virginia) municipality, would likely be entitled to a congressional district of its own, and its residents would have the right to vote in statewide U.S. senate elections. Of course, this approach would not advance the Democrats’ true goals in advocating for D.C. statehood (increasing the percentage of Democrats in the Senate), so it is unlikely that we will be hearing much about it.
The Powers of Upper Chambers
The procedures for selecting upper-chamber legislators discussed above only really matter if an upper chamber is more than just a figurehead chamber. For example, the British House of Lords (the British parliament’s nominal upper chamber) is selected via a manifestly undemocratic process, but because it has not had significant lawmaking powers since 1911, political observers have generally not viewed its existence as a profound antidemocratic defect within the British political system (though, interestingly, the chamber’s unexpected role in the Brexit saga has reinvigorated a longstanding debate about its status in British government).
While most bicameral legislatures do not feature upper chambers as powerless as the House of Lords, the general pattern among them is that lower chambers are more powerful than upper chambers. Where upper chambers are comparatively weak (as is the case in much of the democratic world), the upper-chamber selection process is less of an affront to democratic principles. In this regard, the U.S. is an important — though not a unique — exception to the rule.
Perhaps the simplest and most instructive way to measure the relative strength of chambers in bicameral legislatures is to examine what happens when the two chambers have an irreconcilable disagreement on a basic policy issue. In such a circumstance, does a country’s constitution provide a formal means for breaking the deadlock at hand? If so, does this procedure favor one chamber or the other? The table below provides some basic information on that question for the countries discussed in the previous section (it should be noted that each legislature features a complex relationship between the chambers that is described in highly simplified form here).
As can be seen, three of the parliaments described above resolve inter-chamber disputes (at least on some issues) by giving the lower chamber the upper hand. While the exact procedures vary across the countries, the outcome is the same: if the disagreement is irreconcilable, the lower chamber can impose its will on the upper chamber. Numerous additional bicameral parliaments that are not discussed in this essay use this basic approach as well. Importantly, it is not always procedurally easy for the lower chamber to impose its will, and in some cases it is rarely done, but the option is there.
That the lower chambers of the French, German, and Japanese legislatures are the decisive chambers in much national policymaking does not mean that the upper chambers of these countries are as ineffectual as the British House of Lords. As chambers designed to represent their countries’ localities, these upper chambers assert themselves on legislation directly affecting regional governments. This is especially the case in Germany, where the Bundesrat (composed of officials selected by the German state governments) retains absolute veto power over any national legislation affecting the finances or administrations of the German states. Unlike the Bundesrat, the French Senate has traditionally been a weak upper chamber, but over the past several decades it has developed an identity as a defender of the interests of France’s regional governments, and in 2003 the French constitution was amended to give the Senate exclusive originating power over all legislation affecting French jurisdictions.
Scanning the rest of the table, we see that the Australian and Indian Parliaments offer another means of resolving inter-chamber policy disputes: a joint sitting of the entire national parliament (in Australia’s case, only after new elections have occurred and a new parliament is seated). Here too, the means of breaking inter-chamber deadlock ultimately favors the lower chambers; since they have more seats than upper chambers, their perspective is more likely to win out in a joint sitting.
The United States, unlike every country in my sample except Mexico, has a national legislature in which there is no formal way of breaking a deadlock between the two legislative chambers. The House and Senate can try to iron out their differences via a conference committee, but the conference committee (if successful) will produce a conference report that still has to be independently passed by both chambers. This gives the U.S. Senate a level of power in ordinary lawmaking that the vast bulk of national upper chambers do not have, and it makes the malapportionment of the Senate a far more significant issue.
What the Senate Conversation is Missing
When compared to other national upper chambers, is the United States Senate unusually undemocratic? The simple answer to this question is: it depends on how we understand the Senate’s purpose. If its purpose is the same as that of the House (i.e., to represent the needs and interests of ordinary Americans in national policymaking), then clearly the U.S. Senate is a very undemocratic body. As the malapportionment statistics I have presented conclusively show, some ordinary Americans (i.e., those from small states) have a much greater say in the Senate’s composition than others (i.e., those from large states), and these inequalities are greater than those present in other prominent democracies (like Australia and Mexico) whose upper chambers are set up in similar ways. Moreover, the fact that the residents of America’s capital district are unrepresented in the Senate creates a representational injustice that is unique among the world’s major democracies (though, if the goal is to ameliorate overall inequalities in Senate representation, granting DC statehood is not the best way to remedy this problem).
However, France, Germany, and India provide us with examples of modern upper chambers oriented around a different purpose: serving as a venue for the limited representation of regional governments in national policymaking. As alluded to earlier, this alternative purpose seems roughly consistent with the vision that the framers of the American Constitution had for the Senate way back in 1787. Not only did the framers design the Senate to be selected by state legislatures, but, as Sarah Binder has argued, it seems likely that they believed it would be a substantially less dynamic and more reactive chamber than the House.
It is possible that recovering the framers’ vision of the Senate (and modifying it somewhat to address modern circumstances) could be of tremendous benefit to the American political system. Numerous programs created by Congress are administered by the states, so giving the states a direct seat at the table in the development of these programs (like the German states enjoy through the Bundesrat) could result in more effective policymaking, a chief goal of many progressives. Giving the states a direct voice in the Senate could also prevent the unnecessary centralization of policy in DC, something conservatives generally desire.
If we conceive of the U.S. Senate’s purpose as being similar to that of the German Bundesrat or French Senate (i.e., it is meant chiefly to represent the states in national policymaking), then malapportionment is not as serious of a problem as are two other issues with the modern Senate. The first is that, ever since the 17th Amendment was ratified, state legislatures no longer select U.S. senators. The second, and equally important, problem is that the U.S. Senate, unlike those other chambers, has taken on a coequal role with the U.S. House in the policymaking process, including on matters that do not directly affect state governments. The first problem points to a potential solution that some conservatives might like (returning to state legislative selection of senators), but the second problem points to a solution that progressives might be fond of (reducing the Senate’s powers vis-à-vis those of the House). Of course, both solutions could only be implemented via constitutional changes, and that is no small feat given that the U.S. Constitution is exceedingly hard to amend.
Nevertheless, I hope I have demonstrated that the biggest difficulty with the current conversation regarding how to fix the Senate is our lack of clarity regarding what it is for. Because the Senate has evolved into a more popular chamber over the course of 230 years of American history, our tendency to evaluate it based on how well it facilitates popular rule has grown. But many upper chambers of advanced countries exist for other reasons, and they continue to command legitimacy in the modern democratic age. Thus, prior to deciding how we want to reform the Senate, we should think carefully about what (if anything) we want from it.
 Importantly, only subnational units enjoying some level of representation within the upper chambers were included in the calculations of the Loosemore-Hanby Index. At-large seats in Japan, India, and Mexico were factored into the calculations using a technique suggested by Richard Snyder and David Samuels. See David Samuels and Richard Snyder, “The value of a vote: malapportionment in comparative perspective,” British Journal of Political Science 31 (2001): 651–671.