Yesterday, we looked at how the European Union functions on a model of "democracy" that is based on very large constituency sizes, thus reducing access to policymakers, and reducing the ability of ordinary taxpayers to influence the lawmaking process. (The more residents there are for each elected official, the larger the constituency size.)
An analysis of constituency size showed that there are far, far more constituents per elected official in the EU Parliament than is typically the case for national legislatures in Western Europe.
Constituency size has been shown to correlate to government spending in many cases, as shown in research by Mark Thornton, George S. Ford, and Marc Ulrich. See here and here.
As Thornton et al. conclude:
[T]he evidence is very suggestive that constituency size provides an explanation for much of the trend, or upward drift in government spending, because of the fixed-sized nature of most legislatures. Potentially, constituency size could be adjusted to control the growth of government.
They note the reason for this appears to be diminished engagement between elected officials and constituents when constituency sizes are very large.
Also key in this analysis is the problem of "asymmetry" of interests between ordinary voters and minority interest groups. For example, a group of farmers interested in maintaining a subsidy will go to a lot of effort in order to lobby elected officials. This will be true even if the opportunity cost of access to elected officials is very high — the potential gains for lobbying success are very high. For voters who don't want to pay for the farmers' subsidies, however, the calculus is very different. Each farmer might receive thousands of dollars in subsidies. But each taxpayer only pays a small fraction of that to keep the subsidies flowing. A taxpayer may wish to lobby his elected official against the subsidy, but if meeting with an elected official is costly in terms of time and money, the taxpayer will quickly find it's not worth the trouble. Thus, lowered access to elected officials is more likely to prevent lobbying from individual voters than from established and well-funded interests.
Other factors mentioned by Thornton, et al. and others include:
- Large constituencies increase the cost of running campaigns, and thus require greater reliance on large wealth interests for media buys and access to mass media. The cost of running a statewide campaign in California, for example, is considerably larger than the cost of running a statewide campaign in Vermont. Constituencies spread across several media markets are especially costly.
- Elected officials, unable to engage a sizable portion of their constituencies rely on large interest groups claiming to be representative of constituents.
- Voters disengage because they realize their vote is worth less in larger constituent groups.
- Voters disengage because they are not able to meet the candidate personally.
- Voters disengage because elections in larger constituencies are less likely to focus on issues that are of personal, local interest to many of the voters.
- The ability to schedule a personal meeting with an elected official is far more difficult in a large constituency than a small one.
- Elected officials recognize that a single voter is of minimal importance in a large constituency, so candidates prefer to rely on mass media rather than personal interaction with voters.
- Larger constituent groups are more religiously, ethnically, culturally, ideologically, and economically diverse. This means elected officials from that constituent group are less likely to share social class, ethnic group, and other characteristics with a sizable number of their constituents.
- Larger constituencies often mean the candidate is more physically remote, even when the candidate is at "home" and not at a distant parliament or congress. This further reduces access.
On other words, there is compelling reason to believe that smaller is better.
The Case of the United States
In response to yesterday's article looking at EU representation, some astute readers noted that the US suffers from a similar problem.
This is true. In fact, among countries with elected national legislatures, the United States has one of the largest constituency sizes of all. As the first graph shows, the US is similar to Russia, Brazil, and Pakistan, but is far outside the trend even for those countries. Only India, with constituency sizes of more than a million people, has larger constituency sizes than the US in our sample.
Our second graph shows constituency sizes calculated using the total number of representatives in the national legislature. These include only elected representatives, and thus excludes the UK House of Lords, for example. In bicameral legislatures, totals from each house are added together. Thus, the US is 535 (100 Senators + 435 Representatives):
The United States is very unlike Western European national legislatures in this regard, and is even an outlier by Latin American standards. Overall, legislatures outside of Western Europe tend to have larger constituency sizes in general.
As we can see, in the United States, the average constituency size (or "district" size in the case of the US) is nearly 600,000. This is comparable to representation for member states of the EU in many cases. This total can also vary considerable from state to state, with the largest states having the largest district sizes. This is based on the population of each state divided by the number of members of Congress in each state (using current Census data):
In California, a district size, on average includes 677,000 people, which means each constituent must compete with 677,000 other people — not to mention large interest groups — for access to elected officials.
The picture is less stark at the state level. While Congress today manages the minutiae of daily life from wages to home loans, state governments also produce significant amounts of regulation and legislation.
In most cases, however, it is easier for constituents to gain access to state legislators. California remains, by far, the state with the largest constituency sizes, while smaller states, such as New Hampshire, Wyoming, Maine, and Alaska, have far smaller constituency sizes:
Changes Over Time
In most modern democracies, legislative size tends to increase at a much slower pace than population growth in general. In the United States, the size of the House of Representatives has been locked at 435 since 1929 with the Permanent Reapportionment Act of 1929. Ever since then, Congressional representation has been reapportioned based on relative population growth across the states, and not on growth in population.
When fixed at 435 Representatives (plus 96 Senators from 48 states), the population in the United States was 123,000,000. So, at the time, the average number of constituents per member of Congress was 231,000. That's a constituency size smaller than all but three states today. The size has grown about 156 percent from 1930 to 2010.
When looking at Congressional delegations at the state level, we find the largest constituency size in 1930 was in New York with 267,000, and the smallest was found in Nevada with a constituency size of only 30,000.
Constituency sizes have more than doubled over this time in most states, and in some states, the size has more than tripled. For the sake of a better scale, I have left Nevada off this list, since the increase in Nevada over this period was more than 1,200%:
Constituency sizes at the state level have been smaller. The Massachusetts House, for example, reached a total size of 635 in 1837. Given the population at the time, that works out to an average constituency size of a tiny 1,160 (excluding the state Senate). Even as recently at 1978, the Massachusetts House had 80 more members than it has today — with about 800,000 fewer people.
Out west, Colorado's General Assembly, for example, has always had 100 members. It had 100 members in 1876 when Colorado became a state, meaning that the average constituency size was about 1,900 people in 1880. Today, that has increased to around 53,000.
The Anti-Federalists Wanted Smaller Constituent Sizes
For a look at what was originally envisioned for constituency size in the United States, we can consult the Constitution itself, which mandates "the number of Representatives shall not exceed one for every thirty Thousand." Noticing that the Constitution did not actually mandate increases in the size of the congress to match population increases, the anti-federalists had attempted to amend the constitution to ensure the legislature grew with population (see "article the first"). The anti-federalist Cato, in his Letter No. 5 noted "that the number of representatives are two few" among several "evils that will attend the adoption" of the new constitution. As usual, the opponents of the 1787 Constitution have been proven correct.
Nevertheless, in 1790, members of congress represented on average 37,000 people. In Massachusetts in 1800, for example, there were 422,000 people sharing 16 members of Congress. Senators were not directly elected at the time, however, so if we include only members of the House, the average constituency size in Massachusetts in 1800 was 30,142. To reach a constituent size like this today, the US Congress would require 10,000 members. If this strikes us as "impractical" then it's likely that the United States is far too large to offer anything that we might seriously call "representative government." It is far more likely that the notion of a single nation-state with 318,000,000 people is what is impractical.
Many Argue for Making Constituent Size Even Larger
In recent decades, elected officials have pressed to increase the size of constituencies, even as population growth makes them bigger every year. As we have noted, Massachusetts significantly reduced the number of representatives in its state house in 1978, and politicians in Pennsylvania are currently attempting to reduce the size of the state legislature. Advocates of such measures usually claim that reductions in representative totals will save money, but the dollar amounts saved are minute compared to the size of the state budget. It is claimed that six million dollars will be saved if Pennsylvania reduces the size of its legislature. But, the chances that the state budget will be reduced by six million is approximately zero, and the prospect of reducing member salaries or the immense staffs maintained by members does not appear to be on the table. In practice, reductions in legislator totals merely impose greater costs on private taxpayers and voters who will pay more dearly in terms of opportunity cost and cash in efforts to communicate with and influence policymakers. Meanwhile, reductions in legislative size reduces costs for lobbying groups which benefit from having fewer legislators to keep track of.
Advocates also claim that "technology" now makes large legislatures obsolete since constituents can send emails and make phone calls. In practice, personal relationships are an enormous factor in influencing votes, and larger constituencies make these relationships more inaccessible for ordinary people.
Technological advances argue more in favor of making legislatures larger. It is now easier than ever for larger numbers of people to interact in a large legislature. Gone are the days when it was necessary for everyone to gather in one room for a single vote, or for physically counting hands or yeas and nays. Members of Congress need not even be in Washington, DC to vote. State legislatures could be composed of hundreds of people who exchange bills, bill amendments, and count votes instantaneously thanks to technological advances. In the days of yore, large legislatures quickly became unruly thanks to cumbersome record-keeping and the length of time needed to draft bills and amendments. Those days are gone.
Constituent Size Is Just One Strategy
As noted by Thornton, et al., reducing constituent size could be a helpful strategy in reducing government spending, but my purposes here are not to suggest that democracies composed of small constituencies are some type of ideal form of government. Nor do I claim that the US Constitution should be regarded as an unassailable authority on these matters. Constituent size by itself cannot be used to ensure that human rights are protected or that peaceful people be left alone. Certainly, ideology is a major independent factor in limiting state power: if people want a large intrusive government, they're going to get it regardless of constituent size. And, as Nathan Benefield of the Commonwealth Institute has noted, other important factors include the size of legislative staffs, the degree to which legislators are "professional legislators" rather than "citizen legislators," and the size of legislative salaries.
The change over time in constituent sizes does call into question how modern defenders of the US political system can so blithely insist that it is illegitimate to oppose the federal government in the form of nullification or secession. "You have representatives in Congress!" is the common refrain that is supposed to silence dissent. But, as we have seen, "representation" today looks nothing at all like it did even in the early 20th century. Legislators are increasingly inaccessible, physically distant, wealthy, and expensive to influence. In many cases, they spend virtually the entire year thousands of miles from the people they are supposedly representing.
If this is "representation," our definition of the word has become thoroughly flawed.
CORRECTION: An earlier version of this article mistakenly included the unelected upper house of the Canadian parliament in the article's calculations. The article has been adjusted to include only the elected lower house.
