IQ, the odd obsession of the Alt-Right
Note: This article has been edited at the request of some people in the Alt-Right, all paragraphs added have been marked at the start with (Alt-Right requested)
With the election of 2016 and the rise of the Alt-Right one of the most pernicious political discussions has once again come out of the dark corners of the web, the race and IQ question, in particular the discussion about the heritability of IQ and how America’s shifting demographics may affect its politics. As it turns out I have my own misgivings about the validity of IQ as a measurement but for the purposes of this article I will assume its validity, and I will analyze IQ in different states of the union, and how it may impact different aspects of political life. To summarize the findings, the best predictor of a state’s low IQ is not its homicide rate, its unemployment rate, its GDP per capita or the percentage of black people living in it, it is whether or not they approve of Donald Trump, which does not bode well for the Alt-Right.
Obtaining IQ estimates for each state
There is no country wide IQ tests, which is why for arriving at IQ scores for large populations you need to estimate it from other tests such as SAT or the NAEP, for the purpose of obtaining a IQ score for each state, I found two published papers [1][2]. Michael McDaniels 2006 paper didn’t control for school dropout or failure to take the test, and given that one of the strongest findings of IQ tests is that low IQ correlates to school dropout this struck me as a major methodological flaw, and its use of the NAEP even though the SAT has shown greater correlation with IQ was another point against it, which is why I choose to go with Satoshi Kanazawa, 2006, IQ and the Wealth of States.
Satoshi estimated IQ, using the SAT and by taking into account the dropout rate from schools, for the purposes of replicating Richard Lynn’s calculations from the Wealth of Nations (that correlated IQ with a Nation’s GDP) except with a correlation of statewide IQ compared with GSP, he arrived at similar results. A table containing all the states IQ was included in his paper, this are the IQ scores used in this article.
Correlations with a state’s median IQ
One often cited correlations between IQ comes from the previously mentioned book the Wealth of Nations by Richard Lynn, which links IQ to GDP. I managed to mostly replicate Kanazawa’s findings but with 2016 GSP data. I arrived at a r of 0.4058(approximating his r of 0.3217), a r2 =0.165 and a p<0.001, this replication demonstrates the validity of using this data to correlate older IQ scores with newer data, see Figure 1 for a scatter point display of the findings.
Annexed with this document there is a link to a google Spreadsheet with the table of raw data, and all the scatter points I created with all of the data points of relevance including state unemployment, homicide rate, GDP but the one variable low IQ scores had the strongest correlation with was the state’s net job approval of Donald Trump in 2017 (job approval as provided by 538[3]) with a r of -0.65 and a r2 of 0.417. That meaning the lower the average IQ of a state the higher the likelihood of that state approving of Donald Trump,see figure 2 for a scatter point visualization of the data.
Against what people in the Alt-Right presume, there is almost no correlation between a state’s percentage of blacks in its population and its IQ, I arrived at a r of -0.061 and a r2 of 0.006. With a correlation that is an order of magnitude smaller to the correlation of Trump approval and IQ, and a r2 soo small it rends the correlation meaningless(r2 typically signals the noise in the data, the lesser the r2 the more noise you can find in the linear regression), see Figure 3 for scatter point visualization.
(Alt-Right requested)After removing hispanics from the white population, using the Kaiser Family Foundation data, the correlation statistics for the relative white population to IQ shows only one thing: the white-ethno-state is gonna be pretty dumb, after controlling for hispanics with a new set of data, I arrived at a r of -0.2 and a r2 of 0.033, even trough the data is pretty noisy, there does seem to have a correlation downwards, so as the population of whites decreasses the IQ increasses, whether this is because diversity increasses IQ or because white people are genetically inferior is a topic for further study. See figure 4 for scattering data visual.
Similarly to Trump’s approval rating there is a strong correlation between a state history of voting republican since 1980 to 2016 and lower IQ scores, but with a less significant correlation a r of -0,44 and a r2 of 0,195. Which are significantly less strong than Trump’s approval rating relationship to IQ, see Figure 5 scatter point visualization.
Conclusion
Unlike what race-realists would have you believe, to the extent that IQ estimates intelligence, it also shows it to be inversely correlated with it being found near the Alt-Right. This finding is not new [4] typically people in the right and specially in the far right have lower IQ scores, which is what makes it odd they find the topic so politically pressiant. In the early 20th century highly intelligent people misguidedly used IQ scores to promote ethno-nationalism to exclude other ethnic groups from entering the United States, but now in a ironic turn of events, the dumbest among us are doing the same, I wonder if they understand how easily this can come back to bite them in the ass.
Table of data https://docs.google.com/spreadsheets/d/1j-k8rpcss_Hw819HiAHHYyD2jLbgBA_3k8SJXm62hYE/edit?usp=sharing
References:
[2]http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.372.7594&rep=rep1&type=pdf
[3]https://fivethirtyeight.com/features/trumps-popularity-has-dipped-most-in-red-states/
[4]http://personal.lse.ac.uk/kanazawa/pdfs/spq2010.pdf
All none directly linked data comes from government sources