Descriptive Representation

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 7.4

Perhaps the most widely referenced work on representation is the previously mentioned book by Hanna Pitkin, The Concept of Representation. Representation can be symbolic, as a king (or a flag) “represents” the nation as a whole. A representative might be an authorized agent, such as a lawyer who speaks or acts as an expert on behalf of a client, as Burke favored. Elections promote this authorized agent concept with Manin’s principle of distinction. Representatives might instead be descriptively representative – as in Adams’s simile of an “exact portrait” – or Mirabeau’s of a map – describing the population in terms of gender, race, class, ideology, interests, and so on.

Jane Mansbridge, of the Kennedy School of Government at Harvard, notes that

“When the U.S. constitution was being debated in 1787-88, and the Federalists advanced their principle of distinction, the Anti-Federalists opposed that principle, arguing that representation required ‘likeness’ and ‘resemblance’ to constituents so that representatives could ‘possess their sentiments and feelings.’”

The New York state Anti-Federalist using the pen-name Brutus (likely Melancton Smith, a former member of the Continental Congress), wrote:

 “The idea that naturally suggests itself to our minds, when we speak of representatives, is that they resemble those they represent; they should be a true picture of the people: possess the knowledge of their circumstances and their wants; sympathize in all their distresses, and be disposed to seek their true interests.”

He also wrote:

“The very term, representative, implies, that the person or body chosen for this purpose, should resemble those who appoint them – a representation of the people of America, if it be a true one, must be like the people. …In this assembly, the farmer, merchant, mecanick. and other various orders of people, ought to be represented according to their respective weight and numbers.”

Alexander Hamilton expressly challenged the notion that the legislature should come from all classes in Federalist Paper No. 35. He wrote:

“It is said to be necessary, that all classes of citizens should have some of their own number in the representative body, in order that their feelings and interests may be the better understood and attended to. But we have seen that this will never happen under any arrangement that leaves the votes of the people free. Where this is the case, the representative body, with too few exceptions to have any influence on the spirit of the government, will be composed of landholders, merchants, and men of the learned professions. But where is the danger that the interests and feelings of the different classes of citizens will not be understood or attended to by these three descriptions of men?”

This assertion that the elite can be expected to fairly represent the interests of the lower classes could theoretically be true, but has not been borne out by reality. Back in chapter 2.5 (the post entitled “An Un-Representative Congress”) I discussed the extensive research of Martin Gilens. In a subsequent book, Affluence and Influence: Economic Inequality and Political Power in America, he neatly summarized his findings that Congressional

“responsiveness to the affluent remains strong while responsiveness to the poor and middle class is completely absent.”

Descriptive representation can have a direct impact on how well representatives perform in terms of knowing the true interests of their constituents and evaluating what action the represented population would take if they had all of the information possessed by the representative. It is natural to perceive and evaluate information in a way that is consistent with one’s own prior biases, class, racial, and other interests. While specific representatives in specific cases may rise above this natural bias, it is undoubtedly the norm. Thus the failure of descriptive representativeness is likely to result in failures of substantive policy representation as well.

Studies have also found that ethical beliefs vary according to wealth. For example, a 2023 study by University of Michigan researchers found that people in the top 5% income bracket are significantly more likely to accept societal income inequality as fair, than are the lower 95% of income earners. One of the researchers, Alain Cohn, said 

“In other words, wealthy individuals are more likely to consider unequal incomes as fair even if the inequality arises from chance.”

The fact that elected representatives skew strongly towards the wealthy likely has an impact on what they perceive to be problems that need addressing, and the policies they consider.

As an example of how the lack of descriptive representation can distort the public agenda (deciding which issues need addressing), a series of studies by Michael Kraus and others at Yale found that Americans dramatically overestimate the progress made in racial economic equality since World War II. There have clearly been improvements in statutory discrimination. However, in terms of wages for college-educated Blacks vs. college-educated Whites, high school-educated Blacks vs. high school-educated Whites, and accumulated assets (wealth), the gap between Blacks and Whites has actually increased today compared to 1945. The wealth gap can be credited to such things as historic low levels of assets among Blacks from “red-lining” policies, and other discrimination, that prevented them from getting mortgages, etc., such that even if no further discrimination occurred, wealth passed intergenerationally and compounded means the wealth gap has actually grown. This misperception of progress having been made that is “wildly discrepant from reality” isn’t evenly or randomly distributed among Americans. The researchers found that it is hugely greater among high-income Whites than for any other group. Thus, the significant over-representation of high income Whites elected to legislatures exacerbates the distortion of agenda setting and policy. The authors note:

“One implication of the present findings is that policy discussions about the detrimental effects of economic inequality, as well as potential ways to reduce it, are unlikely to contend with the racial economic gaps that are contributing to these overarching economic trends. That is, solutions for persistent racial economic inequality might never enter discussion if policymakers and the public believe these racial divides are already closing.”

Representation can be any mixture of Pitkin’s concepts, as well as being more or less substantive or accountable. But unlike the other concepts of representation, descriptive representation can only apply to the whole.  While an individual representative may be an authorized agent, no individual representative can be descriptive of a diverse population. But it is a mistake to think simplistically of the Jewish representative representing Jews, the Black representative representing Blacks, and the labor union representative representing workers. Citing various political theorists from the 18th to 21st centuries, political scientist Joel Parker in his doctoral dissertation points out that:

“An emphasis on resemblance does not hinge on the assumption that we can represent only those of our ‘own’ group, but the opposite. Just as a nation is more than the sum of its citizens, a representative body is more than the sum of its members. Since we are all affected by the actions taken by political bodies en toto, we should think of the entire body making decisions that affect us as being representative. The representative body as a whole must stand for the nation as a whole.”

It is not sufficient for descriptive representation to be nominal. Depending on the goal of those making the selection, it is possible to assemble a group of token representatives that is superficially descriptive. For example, it is possible to carefully select a group of African-American “representatives” who act in a manner diametrically opposed to the interests of the vast majority of African Americans. Hand-picked descriptive representatives offer no guarantee of comprehensive representation. But the use of scientific sampling can help overcome this failing. Random selection will generate a body that is not only roughly similar in appearance, or other identifiable demographic measures, but also in terms of invisible characteristics such as perspectives, interests, cognitive style, values, class, and political beliefs. With sortition, the visible similarity is essentially a “double check” for the overall representativeness of the sample. It is necessary to move beyond seeing an individual delegate as one’s representative, to seeing the representative body as a whole as being representative (or failing to be representative) of the community as a whole. This jury model of  representation using sampling methods, or “representation by sample,” was proposed as far back as 1913 by A. Lawrence Lowell in his book Public Opinion and Popular Government.

But many people reject descriptive representativeness as a good thing. The competing vision is of average people being ruled by the best among them. This was certainly Plato’s and Madison’s view. Henry J. Ford (no relation to the motor company tycoon), an ally of President Wilson and also a former president of the American Political Science Association, wrote in 1924 that the notion of a reflection or resemblance is as ridiculous in politics as in corporate governance:

“No one would regard it as a correct principle in the organization of a joint stock company that the management should be such as would fairly reflect such knowledge as the shareholders might have about the business the company is to carry on. The principle universally relied on there is a grant of power to men of special knowledge and capacity, subject to responsibility for results.”¹

As discussed earlier, the opposing preferences for representation by mirror, or by merit were debated in both America and France at the time of their revolutions. But the overwhelming sentiment among the wealthy men who led both revolutions, was that representatives should be elite men like themselves. This debate about typical vs. superior, or the related clash between representativeness and expertness may actually be false dichotomies. Being a mirror that adds diversity can make the deliberative body, as a whole, superior to a body made up of individually “superior” members, and a descriptively and genuinely representative body may bring forth kinds of expertness that a body of individual experts couldn’t muster (and the randomly selected ordinary members can consult actual experts). I will examine this in great detail in subsequent chapters. 

1. The governance comparison to a stock corporation is inappropriate, since, unlike a corporation, in a community there is not a single measure of “results” (profits). If results are achieved through harmful externalities, such as pollution, or harm to a community, that is not a corporation’s concern. It is also noteworthy that he dismisses the idea of the governing board being like the shareholders, but he doesn’t even bother to dismiss the possibility that they could be like the workers (who likely know more in detail about the business than the “expert” board members), or like the community or their consumers, as those likenesses are inconceivable to him. He stresses the expertise priority for a stock corporation in a capitalist economy, but the analogy to society is far off the mark.