Sort of Sorted but Definitely Cold

18 May

By now students of American Politics have all become accustomed to seeing graphs of DW-NOMINATE scores showing ideological polarization in Congress. Here are the equivalent graphs (we assume two dimensions) at the mass-level.

Data are from the 2004 ANES. Social and Cultural Preferences are from Confirmatory Factor Analysis over relevant items.





Here’s how to interpret the graphs:

1) There is a large overlap in preference profiles of Rs and Ds.

2) Conditional on same preferences, there is a large gap in thermometer ratings. Without partisan bias – same-preferences should yield about the same R-D thermometer ratings. And this gap is not particularly responsive to change in preferences within parties.

On (Modest) Differences In Racial Distribution of Voting Eligible Population and Registered Voters in California

13 Apr

Each election cycle, many hands are waved and spit is launched in air, when the topic of registration rates of Latinos (and other minorities) comes up. And indeed registration rates of Latinos substantially lag those of Whites. In California, percent eligible Latinos who are registered is 62.8%, whereas percent eligible Whites registered to vote is approximately 72.9%.

This somewhat large difference in registration rates doesn’t automatically translate to (equally) wide distortions in racial distribution of the eligible population and the registered voter population. For example, while self-identified Whites constitute 62.8% of the VEP, they constitute marginally more – 64.2% of the voting eligible respondents who self-identify as having registered to vote.

Here’s the math:

Assume VEP Pop. = 100
Whites = 63/100; of these 72% register = 45
Latinos = 23/100; of these 62% register = 14
Rest = 14/100; of these 62% register = 9
New Registered Population = 45 + 14 + 9 = 68
Registered: Whites = 66.2; Latinos = 20.6

Source: PPIC Survey (September 2010).
Note: CPS 2008, Secretary of State data confirm this. Voting day population estimates from Exit Poll also show no large distortions.

Some simple math:
For a two category case, say proportion category a = pa
Proportion category b = 1 - pa

Assume response rates for category a = qa, and for category b = qb = c*qa

Initial Ratio = pa/(1 -pa)
Final Ratio = pa*qa/(1-pa)*qb

Or between time 1 and 2, ratio changes by qa/qb or 1/c

T1 Diff. = pa - (1- pa) = 2pa - 1
T2 Diff. = (pa*qa - qb + pa*qb)/(pa*qa + (1-pa)*qb)
= (pa(qa + qb) - qb)/(pa(qa - qb) + qb)
= [pa*qa (1 + c) - c*qa]/[pa*qa(1-c) + c*qa]

T2 Diff. - T1 Diff. = [pa*qa (1 + c) - c*qa]/[pa*qa(1-c) + c*qa] - (2pa -1)
= [pa*qa (1 + c) - c*qa + pa*qa(1-c) + c*qa - 2pa (pa*qa(1-c) + c*qa)]/[pa*qa(1-c) + c*qa]
= [pa*qa + pa*qa*c - c*qa + pa*qa - pa*qa*c + c*qa - 2pa*pa*qa + 2pa*pa*qa*c - 2pa*c*qa]/[pa*qa(1-c) + c*qa]
= [2pa*qa - 2pa*pa*qa + 2pa*pa*qa*c - 2pa*c*qa]/[pa*qa(1-c) + c*qa]
= [2pa*qa(1- pa + pa*c -c)]/[pa*qa(1-c) + c*qa]
= [2pa*qa((1- c) - pa(1-c))]/[pa*qa(1-c) + c*qa]
= [2pa*qa(1-pa)(1-c)]/[pa*qa(1-c) + c*qa]

Diff. in response rates = qa - qb

When will diff. in response rates be greater than T2 - T1 Diff. -
qa - qb > [2pa*qa(1-pa)(1-c)]/(pa*qa - pa*qac + cqa)
qa(1-c)(pa*qa - pa*qac + cqa) > 2pa*qa(1-pa)(1-c)
qa(1-c)(pa*qa - pa*qa*c + c*qa) - 2pa*qa(1-pa)(1-c) > 0
(1-c)qa [pa*qa - pa*qa*c + c*qa - 2pa(1 -pa)] > 0
(1-c)qa[pa*qa -pa*qa*c + - 2pa + 2pa*pa] > 0
(1-c)qa[pa(qa - qa*c -2 + 2pa) -] > 0
(1- c) and qa are always greater than 0. Lets take them out. - - 2pa + - > 0
qa - qa.c - 2 + 2pa - > 0 [ dividing by pa]
qa + 2pa - + 1/pa) > 0
qa + 2pa > + 1/pa)
(qa + 2pa)/[qa(1 + 1/pa)] > c
[pa*(qa + 2pa)]/[(pa + 1)qa] > c

When will diff. in response rates + initial diff. > T2 diff.
qa - qa*c + 2pa - 1 > [pa*qa (1 + c) - c*qa]/[pa*qa(1-c) + c*qa]
[pa*qa(1-c) + c*qa][qa - qa*c + 2pa - 1] - [pa*qa (1 + c) - c*qa] > 0
- pa*qa + pa*qa*c - c*qa + [pa*qa(1-c) + c*qa][qa - qa*c + 2pa] - pa*qa - pa*qa*c + c*qa > 0
-2pa*qa + [pa*qa(1-c) + c*qa][qa - qa*c + 2pa] > 0
-2pa*qa + [pa*qa - pa*qa*c + c*qa][qa - qa*c + 2pa] > 0
-2pa*qa + pa*qa[qa - qa*c + 2pa] - pa*qa*c[qa - qa*c + 2pa] + c*qa[qa - qa*c + 2pa] > 0
-2pa*qa + pa*qa*qa - pa*qa*qa*c + 2pa*qa*pa - pa*qa*c*qa + pa*qa*c*qa*c + 2pa*qa*c*pa + c*qa*qa - c*qa*qa*c + 2pa*c*qa> 0
-2pa*qa + pa*qa^2 - 2c*pa*qa^2 + 2qa*pa^2 + pa*c^2*qa^2 + 2pa^2*c*qa + c*qa^2 + c^2*qa^2 + 2pa*c*qa > 0
-2pa*qa + 2qa*pa^2 + 2pa*c*qa + 2pa^2*c*qa + pa*qa^2 - 2c*pa*qa^2 + pa*c^2*qa^2 + c*qa^2 + c^2*qa^2 > 0
2qa*pa(-1 + c + pa + pa*c) + pa*qa^2 (1 - 2c + c^2) + c*qa^2(1 + c) > 0
2qa*pa(-1 + c + pa(1+c)) + pa*qa^2 (1 - c)^2 + c*qa^2(1 + c) > 0
two of the terms are always 0 or more.
2qa*pa(-1 + c + pa(1+c)) > 0
-1 + c + pa(1+c) > 0
pa > (1-c)/(1 +c)

Measuring Partisan Affect Coldly

24 Mar

Outside of the variety of ways of explicitly asking people how they feel about another group — feeling thermometers, like/dislike scales, favorability ratings — explicit measures asked using mechanisms designed to overcome or attenuate social desirability concerns — bogus pipeline, ACASI — and a plethora of implicit measures — affect misattribution, IAT — there exist a few other interesting ways of measuring affect:

  • Games as measures – Jeremy Weinstein uses games like the dictator game to measure (inter-ethnic) affect. One can use prisoner’s dilemma, among other games, to do the same.
  • Systematic bias in responding to factual questions when ignorant about the correct answer. For example, most presidential elections years since 1988, ANES has posed a variety of retrospective evaluative and factual questions including assessments of the state of the economy, whether the inflation/unemployment/crime rose, remained the same, or declined in the past year (or some other time frame). Analyses of these questions have revealed significant ‘partisan bias’, but these questions have yet to be used as a measure of ‘partisan affect’ that is the likely cause of the observed ‘bias’.

Some Potential Negatives of Elite Polarization

18 Feb

Growing ideological distance between the parties has produced clearer choices. This added clarity has resulted in improved propensity among voters to make ideologically consistent choices (Levendusky 2010). This is seen as a positive.

However, there may be some negative normative implications as well. If parties have moved away from the center, and if most people are near the center (as data shows), two things follow –
1) The average distance of each of those people from either of the parties has increased. So people’s choices have become impoverished.
2) The penalty of misclassification – for a leaner to mistakenly vote for the wrong party – has increased substantially. It may well be that while propensity of misclassification has decreased, the penalty has increased, leaving aggregate utility slightly worse off.

Secondly, if the government is at least partly in the business of providing public goods that require collective action (distributed costs), the split nature of constituencies and constituency-based entrenched positions may very likely lead to an under-provision of public goods.

Thirdly, and something that is covered in the first point, clearer choices don’t mean the choices that are the best, or even what people want. One would hope that the choices on offer are optimal but we know that monopolies or duopolies under conditions where start-up costs are high to get into the sector have a sparse record to providing something like that.

Fourthly, it also follows that given we have firm partisans, parties will stop broadening their constituencies beyond a certain point due to the law of really rapidly diminishing returns under sorted electorate conditions. This will mean that the policy buckets shrink, and they will have larger incentives to cater to their bases.

Fifthly, the legitimacy of the government is likely to be reduced among the losing camp which has reasons to believe that the ruling coalition doesn’t represent it.

Institutional Distrust

25 Jan

It is sometimes assumed that high levels of institutional distrust in America are peculiar to it. So much so that a variety of theories have been offered to ‘explain’ this peculiarity including, but not limited to, elite polarization, income inequality, polarized media, etc. Empirical support for the ‘American exceptionalism’ however is somewhat less clear – across some major Western democracies (outside of the perennially ‘sunny’ Danes; one may talk about Scandinavian Exceptionalism perhaps), percent who ‘tend not to trust’ [pay attention to the y-axis] national government, national parliament, and political parties is alarmingly high. These high levels raise concerns about the legitimacy of the system.

Affectively Polarized? – Partisan Polarization Among the Masses

9 Jan

The shooting of Representative Giffords has reignited the debate about the extent to which the public is polarized. Some political scientists have answered the question by evaluating data on policy positions of people over the years. And the data are clear on the question—no, not really.

However, lack of ‘real’ differences hasn’t always meant a lack of perceived differences. Nor has it meant lack of negative affect. Affective dislike, conditional on similarity, is unsurprising and typical, as the history of racial and ethnic hatred will attest to.

So, we tested to see whether partisans dislike each other. We asked a representative sample of Americans whether certain traits described Republicans and Democrats. From the 18 traits (selfish, generous, close-minded, honest, etc.), we create a latent measure of partisan affect (ICC here. Here’s a plot of latent partisan affect by partisan self-identification.

The Unscientific Republican

16 Dec

Only 6% of scientists in a random sample of American Association for the Advancement of Science (AAAS) identify themselves as Republicans, according to a Pew 2009 Study. Assuming AAAS is not tremendously unrepresentative of scientists as a whole, what explains under-representation of Republicans in science?

While Republicans are under-represented among people with advanced degrees more generally, I here limit myself to offering some possible hypotheses for explaining under-representation of Republicans in science, refraining from any attempt to analyze their validity.

Science education or practice causes liberalism:

  1. Brainwashing: Science is taught by liberals, who brainwash their students into believing their preferred ideology.
  2. Diversity: Upper echelon of science today is racially and culturally? diverse. Interacting with a diverse set of people causes people to become liberal.
  3. Cynical: Science is funded by the government, and people in the field, like elsewhere, love the hand that feeds them.
  4. People practicing science come to see the value of policy guided by science. On some major issues – like climate change and evolution, Republican Party has taken ‘anti-science’ stances.
  5. Science causes people to question religious dogma, become atheists or even anti-theists, etc. causing them to choose a party less aligned with religion.


  1. Intelligence: Intelligence causes liberalism (Kanazawa). Intelligence is one of the things that affect the choice of the discipline and the number of years a person chooses to study. Intelligence is the confounding variable that predicts both.

Liberals select into science, and Republicans select out of it.

  1. Conservatives “have a need for cognitive closure” which is not amenable to open questions and scientific inquiry (Jost).
  2. Concept of “conservatism” is one of maintaining the status quo.
  3. Social dominance orientation: Republicans prefer going into status quo enhancing occupations like economics, business, police; Democrats less so (Sidanius).
  4. Religious people don’t go into science, which they see as anti-religion.

Some Notable Ideas

30 Jun

Education of a senator

While top colleges have been known to buck their admissions criteria for the wrong reasons, they do on average admit smarter students, and the academic training they provide is easily above average. So it may be instructive to know what percentage of Senators, the most elite of the political brass, have graced the hallowed corridors of top academic institutions, and if at all the proportion attending top colleges varies by party.

Whereas 5 of the 42 Republican senators have attended top colleges (in the top 20), 22 of the 57 Democratic Senators have done the same. The skew in numbers may be because New England is home to both top schools and a good many Democratic Senators. But privilege accorded by accident of geography is no less consequential than one afforded by some more equitable regime. As in if people of fundamentally similar caliber are there in Senate in either party, a belief cursory viewing of CSPAN would absolve one-off, some due to the happenstance of geography have been trained better than others.

Discounting elite education, one may simply want to look at the extent of education. There one finds that whereas 73% of the Democratic Senators have an advanced degree, only 64% of Republican Senators have the same. While again it is likely that going to top schools increases admission to good advanced degree programs, thereby making advanced education more attractive perhaps, there again one can conjecture that whatever the reason – some groups of people, due to mere privilege perhaps, have better skills than others.

Why do telephones have numbers?
Unique alphabetical addresses could always be uniquely coded using numbers or translated as signals. However technological limitations meant that such coding wasn’t widely practiced. This meant people were forced to memorize multiple long strings of numbers, a skill most people never prided themselves with, and when forced to learn it, they took to it like fish to land. Freedom from such bondage would surely be welcomed by many. And now, unsparing hands of technological change which have made such coding ubiquitous hope to deliver comeuppance to telephone numbers. Providing each phone with an alphabetical ID that maps to its IP address would be one smart way of implementing this change.

Misinformation and Agenda Setting
People actively learn from the environment, though they do so inexpertly. Given an average American consumes over 60 hrs of media each week, one prominent place from they learn is the media. Say if one were to watch a great many crime shows – merely as a result of the great many crime shows on television – one may impute wrongly that crime rate has been increasing. A similar, though more potent, example of the this may be the 70% of Americans who believe breast cancer is the number one reason for female mortality.