Quotes and comments: Knowledge about ignorance: New directions in the study of political information (Ilya Somin)

Knowledge About Ignorance New Directions in the Study of Political Information

Quotes from


If political ignorance is rational and most voters choose not to learn

much about politics for that reason, widespread ignorance is a phenom-

enon that democracies will probably have to live with for the foresee-

able future. The challenge for democracy is to find a way to minimize

the harm that political ignorance can cause.

Assuming that we shud keep a democratic government form, then yes, this is correct. Altho, giving people more power shud, according to this theory, also result in them taking more time to educate themselves. That’s an interesting and optimistic implication.

Here, I want to emphasize a different shortcoming of shortcuts, one

that was partly anticipated in Converse’ s #$%! paper. Both empirical ev-

idence and the theory of rational ignorance suggest that most voters

acquire political knowledge not primarily for the purpose of casting a

more informed vote, but for entertainment purposes or to satisfy other

psychological needs. If this is so, the shortcuts they use might likewise

be chosen to serve nonvoting purposes rather than to cast a “better”

ballot. Such voters could rationally choose not to evaluate the political

information they have in an objective way: a form of “rational irra-

tionality” (Caplan &””#). Again, such a choice need not involve precise,

conscious calculations about the costs and benefits of evaluating politi-

cal information objectively . As with the decision to vote and the deci-

sion not to spend much time acquiring political information, the

choice not to put much effort into analyzing political information ob-

jectively could simply be the result of an intuitive sense that there is lit-

tle or no benefit to engaging in such analysis.On the other hand, voters

can easily recognize that extensive knowledge acquisition imposes sub-

stantial potential costs in terms of time and emotional stress.Thus, a de-

cision not to analyze political information rigorously could be an ex-

ample of “satisficing” behavior (Simon #$)’), where individuals make

rational decisions but do not necessarily engage in rigorous calculation.


Such dynamics might often lead voters to use shortcuts that mislead

rather than inform. For example, the use of party-label and ideological

shortcuts led both voters and even many sophisticated political elites to

misperceive President Richard Nixon’ s policies as conservative (Hoff

#$$!). Nixon presided over an unprecedented expansion of the welfare

state, established affirmative action, created the Environmental Protec-

tion Agency , proposed a guaranteed annual income and national health

insurance, and established closer relations with communist China and

the USSR. But he was still widely perceived as a right-winger. Simi-

larly , liberals rallied around President Bill Clinton, while conservatives

rushed to condemn him, despite his endorsement of conservative poli-

cies on free trade, welfare reform, crime control, and other important

issues. Liberals defended Clinton and conservatives attacked him in

large part because of what he represented on a symbolic level as a “draft

dodger” and philanderer, rather than on the basis of his substantive

policies (Posner #$$$). In both the Nixon and Clinton cases, the desire

of liberal and conservative “fans” to rally around their leader or con-

demn a perceived ideological adversary blinded them to important as-

pects of the president’s policies—despite the fact that information

about these policies was readily available.


Yes, those odd #”¤(“)# symbols seem to be a bug in the OCR’ing of the paper. Apparently, the OCR misinterprets numerals. Odd. Perhaps deliberate?

Recent evidence confirms the possibility that even the most knowl-

edgeable ideologues might systematically pick ideological shortcuts that

mislead more than they inform. A study of experts in politics and in-

ternational relations finds that their predictions of political events are

usually no more accurate than would be produced by random chance

(Tetlock &””(). Of greater interest for present purposes is the finding

that the most inaccurate experts are those that tend to make their pre-

dictions on the basis of broad generalizations—that is, experts who rely

the most on ideological shortcuts (ibid., chs. *–().7 This result could be

interpreted as an indication that the experts in question are irrational.


However, most social-science experts are rewarded not for the accuracy

of their predictions but on the basis of the originality and apparent so-

phistication of their scholarship. Similarly, pundits and other public in-

tellectuals are rewarded for their popularity with readers and viewers,

not their prescience (Posner &””&). Few, if any, Conversean “ideo-

logues” can increase either their incomes or their professional standing

by improving the accuracy of the ideological shortcuts they use. As a

result, they , like ordinary voters, often have little incentive to use short-

cuts effectively , and considerable incentive to stick with shortcuts that

are often inaccurate.


Interesting. Also reminds me that i really shud get around to reading the book Wrong: Why experts* keep failing us–and how to know when not to trust them, of which i unfortunately do not have an electronic copy… which means that i likely wont be reading it any time soon. Time has a review of it here.

In addition to arguing for the utility of shortcuts, defenders of the view

that widespread political ignorance is not a serious problem have main-

tained that information problems can be overcome by means of the so-

called “miracle of aggregation” (Converse #$$”; Page and Shapiro #$$&;

Wittman #$$(). According to this theory , if ignorant voters’ errors are

randomly distributed, then the “incorrect” ballots cast for candidate A

will be canceled out by similar mistakes in favor of Candidate B, and

the votes of the relatively well informed will determine electoral out-



This argument has a number of flaws, including the fact that the

well-informed minority that determines electoral outcomes in this sce-

nario is likely to be highly unrepresentative of the electorate as a whole

(Delli Carpini and Keeter #$$%; Somin #$$)). On the other side of the

ledger, the danger that voters may rationally rely on inaccurate and mis-

leading shortcuts suggests a particularly powerful reason why their er-

rors are unlikely to be random. On many issues, ignorance shows sys-

tematic patterns of bias in one direction or another (see, e.g., Delli

Carpini and Keeter #$$%; Caplan &””&; and Althaus &””*). This is to be

expected if voters, including even many relatively knowledgeable “ide-

ologues, ” are relying on opinion leaders, ideologies, and other shortcuts

that have been selected for reasons other than accuracy.


Recent research suggests that even the most sophisticated and highly

rational voters may rely on shortcuts that have little relevance to politi-

cal candidates’ likely performance in office. For example, a recent study

of elections for the presidency of the American Economics Association

shows that the relative physical attractiveness of the rival candidates is a

powerful predictor of which candidate prevails in the voting (Hamer-

mesh &””(). The AEA electorate consists of academic economists who

are presumably knowledgeable about the functions of the AEA—and

presumably more committed to rational, maximizing behavior than is

the average voter in ordinary elections. If such voters nonetheless rely

on dubious information shortcuts, it is likely that voters in other elec-

tions are at least equally likely to do so.8


Interesting argument. Reminded my of wisdom of the crowds.


As for the influence of fysical attractiveness. Yeah.. I have for a long time been playing with the idea the politicians shud be anonymous for voting purposes. The idea is to get rid of such effects. I dont know how feasible that idea is. However, certainly, some improvements to the current situation can be made. For instance, outlawing election posters. They use the same effect/bias that ads also use, namely, the mere exposure effect. More rational TV-debates are also possible.

An important implication of the rational-ignorance hypothesis is that

voter knowledge is unlikely to increase very much merely as a result of

the greater availability of information. Even if information is readily

available at low cost, rationally ignorant voters have little or no incentive

to spend time learning it and weighing its implications. This inference is

borne out by empirical evidence showing little or no change in political

$knowledge levels over the last (” years, despite greatly increased educa-

tion levels and a parallel increase in the availability of information

through electronic and other media (e.g., Bennett #$)) and #$$%; Smith

#$)$; Delli Carpini and Keeter #$$%; and Althaus &””*). Thus, advocates

of ambitious theories of democratic participation cannot expect most

voters to reach the knowledge levels their theories require anytime soon.



In most modern democracies, government spending accounts for at

least a third of GDP , and the regulatory activities of the state extend to

almost all areas of life. In the United States, federal spending accounts

for &”.) percent of GDP , and state and local governments spend an ad-

ditional #*.’ percent.13 And the growth of government spending over

the last century has been matched by a parallel expansion of regulation

(Higgs #$)’).


Rationally ignorant voters are unable to keep track of more than a

tiny fraction of all this government activity. Indeed, they probably

would be unable to do so even with considerably greater knowledge

than most of them currently possess.Other things equal, the greater the

size and complexity of government, the greater the likelihood that

many of its activities will escape meaningful democratic control.14 This

result is troubling both for those scholars who regard democratic con-

trol of public policy as an intrinsic good (e.g., Pateman #$'” and Barber

#$)!), and those who value it for purely instrumental reasons such as

the need to curb abuses of power by political elites.


I agree with the last reason mentioned, that is, reason we need some kind of democracy is to avoid abuses of power i.e. nepotism and then like (read the link for many interesting examples).

In a federal system, citizens dissatisfied with government policy in their

state have the choice of either trying to use “voice” (traditional voting)

to address their grievances, or opting for “exit”: leaving for a jurisdic-

tion with more favorable policies (Hirschman #$'”).15 Those who

choose the exit option in effect “vote with their feet. ” Voice and exit

each have their respective strengths and weaknesses (ibid.). But one that

is largely ignored by most analysts is the comparative incentives they

create for knowledge acquisition.


The effectiveness of voice is significantly constrained by rational ig-

norance. As we have seen, individual voters have little incentive to ac-

quire and effectively use relevant information about public policy . By

contrast, exit has the tremendous comparative advantage of creating

strong incentives for individuals to acquire the necessary information to

make decisions about where to live.16 A knowledgeable individual or

family can move to a more hospitable jurisdiction even if the neighbors

left behind remain ignorant. Thus, individuals are likely to put much

more effort into acquiring information about the best jurisdiction in

which to live than into acquiring knowledge about the right candidate

to vote for. Moreover, effective “foot voting” may require less detailed

information than ballot-box voting, since the former does not entail

knowing which officials are responsible for which policies. It also obvi-

ates the need to be able to separate out the impact of multiple govern-

ment policies from each other, and from the effects of background so-

cioeconomic conditions.17


Empirical evidence shows that even severely oppressed populations

with very low education levels can often acquire remarkably accurate

information about differences in conditions between jurisdictions and

then make the decision to vote with their feet. For example, in the

early twentieth century , millions of poor African-Americans in the Jim

Crow-era South were able to determine that conditions were relatively

better for them in the North (and sometimes in other parts of the

South) and make the necessary moves (Henri #$'(; Cohen #$)$; Bern-

stein #$$), ‘)&–)(). This achievement stands in sharp contrast to the

failure of many of today’ s much better educated (and certainly less op-

pressed) voters to acquire basic political knowledge.

In order for foot voting to be effective, however, political power must

be at least partly decentralized. In a unitary state in which all or most

important policies are set by the central government, there is no exit

option other than the very difficult and costly one of leaving the coun-

try entirely . Thus, the informational advantages of foot voting over bal-

lot-box voting provide an important argument in favor of political de-



Obviously , foot voting is not a panacea for all the shortcomings of

government policy. For example, it cannot protect immobile people and

assets, such as property rights in land. And it is far from the only con-

sideration that needs to be taken into account in determining the opti-

mal level of political decentralization.18 Nonetheless, the informational

advantages of foot voting deserve considerably greater attention from

students of federalism and institutional development.


Very interesting argument. So, we have been moving in the wrong direction in Denmark for some time now, it seems. Very interesting to wonder what wud happen if, for instance, drug laws were a matter of regional law not national. Certainly, this makes experimentation much easier. Experimentation obviously makes it easier to know what works, and what doesnt.


One note about such a form. Some central government is necessary (i suspect, havent done research), and it has the incentitive to try to acquire more power constantly by enacting new laws, precisely as we see it in the US.


One good thing tho. If we can avoid more centralization in the future, especially some kind of world government (basically, expanding the power of FN, EU and the like). Then, becus traveling costs becom progressivly smaller over time, foot voting will becom progressivly less costly. Yay, something to look forward to! When i talk with people living in the US, i often suggest to them that they simply MOVE out of the US. That country is beyond repair (its voting system is locked in a two party system, see various films by CCPGrey), and only a revolution can fix it.

It is no secret that majority opinion in the Arab world and in many

other Muslim countries is largely hostile to the United States. Some an-

alysts attribute this result to specific U.S. policies, such as support for Is-

rael and the Iraq War (e.g., Scheuer &””!), while others cite a “clash of

civilizations” between fundamentally opposed Western and Muslim

value systems (e.g., Huntington #$$)). Either or both of these explana-

tions may be valid. But it is also important to consider the possible con-

tribution of widespread political ignorance.


As the data in Table # show, a &””& Gallup Survey of public opinion

in Arab and Muslim nations found large majorities denying that the

September ## attacks were carried out by “groups of Arabs. ” For exam-

ple, )$ percent of Kuwaitis, ‘! percent of Indonesians, and )% percent

of Pakistanis were apparently ignorant of this basic fact. A &””& survey

conducted by the Egyptian newspaper Al Ahram (&””&) found that *$

percent of Egyptian respondents blamed the September ## attacks on

“Israeli intelligence/the Mossad, ” while only #$ percent said that “Al-

Qa’eda or other Islamic militants” were responsible.21 Both the Gallup

and Al Ahram polls were conducted well before the start of the Iraq

War, so the responses are not the products of anti-Americanism gener-

ated by that conflict.




Leave a Reply