I downloaded PISA 2012 Mathematics data for Italian regions: https://docs.google.com/spreadsheets/d/1CRQPn1KdgaiLxdxqxuJeiqLFUH5jnaaoqlDeQTuuVlU/edit?usp=sharing
The data for Northern Italy confirm that Piemonte, the region which received massive immigration from southern Italy, has the lowest scores (although not low) among the northern regions.
[The average score for northern Italy (minus Piemonte): Lombardia, Emilia Romagna, Valle D'Aosta, Veneto, Friuli Venezia Giulia, Trentino, Bolzano (Sud Tirol) is 512, which compared to a British score of 494, is equivalent to a Greenwhich IQ of 102.93.
Piemonte= 499
Northern Italy-Piemonte= 512-499=13.
13= 2.2 IQ points.

So about 30% southern admixture in Piemonte lowered the IQ by 2.2 points. Thus the genotypic difference between Northern and Southern Italy is 2.2/0.3= 7.33 IQ points.

Note that this calculation assumes that there was zero southern admixture in the rest of North Italy. This is false, as other northern regions received their share of southern migrants, about 10%. Thus the real N-S difference would be even larger.
Of course this is based only on Pisa Math. More accurate results would be given by using other scores as well.

The IQ of Northern Italians is really around 101 if we average scores on PISA tests. This is only 1 point higher than UK, Germany, France Native populations.

But 4-5 points higher than Sweden and Denmark natives according to your new PISA "IQ" data. That's very fishy indeed.

As the genetic components reflect mostly migrations dating to before Classical Antiquity

I'm still waiting for your evidence that those migrations of civilized groups could lower people's IQs. You don't answer because you have none.

You also fail to say if you really believe those Eastern Europeans have IQs of 82-94 as your "data" claims. You obviously don't but you won't admit it.

Piedmont (Piemonte) is the region with the lowest IQ in "Padania" (the north proper), scoring lower than Lombardy, Trentino, Friuli, Veneto. This is likely due to its massive admixture with southern populations, due to the massive migrations from Sicily, Puglia, Campania to Piemonte after the Great War.

No. The lowest scoring Northern regions on the PISA tests are Liguria, Valle D'Aosta and Emilia-Romagna. And Lombardy has as many Southern migrants as Piedmont but it scores almost as high as Trento, Veneto and Friuli.

I downloaded PISA 2012 Mathematics data for Italian regions: https://docs.google.com/spreadsheets/d/1...sp=sharing
The data for Northern Italy confirm that Piemonte, the region which received massive immigration from southern Italy, has the lowest scores

Still no. Liguria and Valle D'Aosta score lower than Piedmont in Mathematics. And again Lombardy, with all its Southern migrants, scores almost as high as the other Northern regions. Give up, you've got nothing.

You also fail to say if you really believe those Eastern Europeans have IQs of 82-94 as your "data" claims. You obviously don't but you won't admit it.

No, their IQ is probably higher. I never denied the importance of environment. However, it's unlikely to be as high as that of Western Europe. For a final verdict, I wait for genetic data coming from these populations.

Piedmont (Piemonte) is the region with the lowest IQ in "Padania" (the north proper), scoring lower than Lombardy, Trentino, Friuli, Veneto. This is likely due to its massive admixture with southern populations, due to the massive migrations from Sicily, Puglia, Campania to Piemonte after the Great War.

No. The lowest scoring Northern regions on the PISA tests are Liguria, Valle D'Aosta and Emilia-Romagna. And Lombardy has as many Southern migrants as Piedmont but it scores almost as high as Trento, Veneto and Friuli.

No. Liguria is not in "Padania" as it's south of the Appennini. Lombardia doesn't have the same percentage of southern immigrants as Piemonte, it's got half as much. Emilia Romagna culturally and ethnically belongs more to central than to northern Italy. You need to compare Piemonte to the regions at a similar latitude, and if you do this, you see that Piemonte scores lower than Lombardia, Trentino, Veneto, Friuli-Venezia Giulia, all regions that received far less southern immigrants.
Also keep in mind that northern regions have a lot more immigrants than southern regions (I've got the official statistics for student population) and immigrants have lower PISA scores.So the true difference between North and South is bigger.
Finally, in your blog you admitted that southern Italians have lower IQ "And while Southern Italians are likely to be a few points lower than Northern Italians — as the Irish and Scottish are a few points lower than the English — there's absolutely no reason to believe that North and South would be separated at their extremes by almost a full standard deviation."http://italianthro.blogspot.it/search/label/IQ
So we basically agree. We just disagree on how big this difference is. I think that the phenotypic difference is bigger but that the genotypic difference is about 0.5 SD (5-10 IQ points) and you think it's a bit lower.

No, their IQ is probably higher. I never denied the importance of environment. However, it's unlikely to be as high as that of Western Europe. For a final verdict, I wait for genetic data coming from these populations.

Then stop pretending that PISA scores are a good proxy for IQ. They're obviously not. We know from legitimate test results that no Europeans have IQs below ~98.

There's already genetic data on those populations, but it won't help you since you have no evidence that admixture from Neolithic farmers or anyone else causes lowering of IQs.

Liguria is not in "Padania" as it's south of the Appennini. Emilia Romagna culturally and ethnically belongs more to central than to northern Italy.

Liguria and Emilia-Romagna are both geographically in Northern Italy and politically claimed as part of "Padania" by the Lega Nord. It's the same with Valle D'Aosta, which is even more "northern". You just ignore them all because their PISA scores don't fit with your anti-Southern agenda.

Lombardia doesn't have the same percentage of southern immigrants as Piemonte, it's got half as much.

I don't think the difference is that big. Milan was as much of a destination for migrants as Torino. Do you have data?

Besides, Lombardy also has more foreign immigrants than Piedmont, Liguria and Valle D'Aosta but it scores higher on PISA tests anyway -- almost as high as Friuli, Veneto and Trentino that have a lot less immigration (both foreign and Southern).

http://images.treccani.it/enc/media/share/images/orig//system/galleries/Atlante_Geopolitico/immigrazione_fig_vol1_000100_009.jpg

So we basically agree. We just disagree on how big this difference is. I think that the phenotypic difference is bigger but that the genotypic difference is about 0.5 SD (5-10 IQ points) and you think it's a bit lower.

No, we don't agree. There's no evidence that the North-South difference in IQ is anywhere near as big as you and Lynn claim. And there's no evidence that any of it "genetic".

I don't think the difference is that big. Milan was as much of a destination for migrants as Torino. Do you have data?

I didn't find official data but I did my own calculations based on % southern surnames in Piemonte and Lombardia.

Besides, Lombardy also has more foreign immigrants than Piedmont, Liguria and Valle D'Aosta but it scores higher on PISA tests anyway -- almost as high as Friuli, Veneto and Trentino that have a lot less immigration (both foreign and Southern).

http://images.treccani.it/enc/media/share/images/orig//system/galleries/Atlante_Geopolitico/immigrazione_fig_vol1_000100_009.jpg

I compared the IQ of Piemonte with the average of all regions in northern Italy, including Emilia Romagna and Valle d'Aosta and it's lower than it by 2.5 IQ points.
You have just shot yourself in the foot or as we say in Italian, "Ti stai dando la zappa sul piede". If there are more foreign immigrants in Lombardia, and if foreign immigrants have lower PISA scores (this is a fact, just check the PISA scores by immigrant status), this means that the score difference between Lombardia and Piemontese NATIVES (which now includes also south/north admixed people) is even bigger than it appears by looking at all-inclusive data (no distinction between immigrants and natives). And if Lombardia has less admixture with southern Italians, then the effects of admixture with southern Italians in Piemonte are even bigger than I had estimated.

I didn't find official data but I did my own calculations based on % southern surnames in Piemonte and Lombardia.

You and Lynn really like to make claims based on insufficient (or no) evidence, don't you?

I can count surnames too, and if you look up the most common Southern Italian surnames (Russo, Esposito, Greco, De Luca, Rizzo, Santoro, Caruso), there are usually ~2x as many in Lombardy as Piedmont. And since Piedmont's population is ~1/2 of Lombardy's, that suggests the percentage of Southerners is about the same in both regions.

I compared the IQ of Piemonte with the average of all regions in northern Italy, including Emilia Romagna and Valle d'Aosta and it's lower than it by 2.5 IQ points.

No, you didn't compare any IQs because PISA scores are not IQs. But it's a fact that Liguria and Valle D'Aosta have less Southern and foreign immigration than Piedmont but score lower on the PISA tests. And it's a fact that Lombardy has more Southern and foreign immigration than Friuli, Trentino and Veneto but scores almost as high as they do. It's also a fact that Northern Italy, with all its Southerners and foreigners, scores higher on the CPS than the natives of every other European country (and much higher than Sweden and Denmark natives). All of these facts refute every part of you "argument".

If there are more foreign immigrants in Lombardia [...] And if Lombardia has less admixture with southern Italians

There's less than a 2% difference in foreigners between Lombardy and Piedmont, and you haven't proven that Lombardy has less Southern admixture than Piedmont. So basically, you have no argument.

Maybe you should spend less time "estimating" things and more time dealing with reality...

I'm still waiting for you to acknowledge the Ravens data from D'Amico et al. showing no significant North-South difference in Italian IQ.

I'm still waiting for your evidence that admixture from advanced and civilized groups like Neolithic farmers, Phoenicians etc. could lower people's IQs.

I'm still waiting for you to admit that PISA scores are bad proxies for IQ since you've admitted Eastern Europeans have higher IQs than your "estimates".

And I'm still waiting for you to acknowledge the real IQ data showing all Europeans to have IQs of ~98 or higher.

I think it would be better if you phrased it more along "I don't think there is any strong evidence that ..." instead of "there is no evidence", which is of course false since this evidence is what you are discussing. The first way of phrasing it is less hostile.

My interest in this I that I don't want to waste my time moderating a flame war.

Fine. But I think it would also be better if Duxide stopped ignoring facts and evidence, avoiding questions, and making up his own data.

From the March–April 2015 issue of Intelligence, Two Italies? Genes, Intelligence and the Italian North–South Economic Divide:

The thesis that socio-economic disparities between Southern and Northern Italian regions are explained by genetic differences in the average IQ is examined (Lynn, 2010a, 2012). Historical data on income, infant mortality and life expectancy, offer scant support to a possible nexus between IQ differences and socio-economic development. The ancient history of Southern Italy is also inconsistent with a supposed Phoenician and Arab adverse genetic impact on the average IQ of Southern populations. The paper proposes that regional IQ differences reflect North–South disparities in education and socio-economic development levels. The significant increases in mean scholastic achievement tests, registered in the Italian South in the period 2003–2012, support this conclusion.

[b][...][/b]

From the March–April 2015 issue of Intelligence, Two Italies? Genes, Intelligence and the Italian North–South Economic Divide:

That was a nonsense paper, with the primary argument being an Ignoratio Elenchi. It was argued that biological indexes did not predict SES centuries ago -- and therefore, biological differences can not, by way of cognitive differences, explain current SES differences. As if the relation between cognitive ability and SES could not have shifted across time, with ability becoming more important in post-industrialized economies. Of course, the author found that regional differences in education centuries ago predicted differences in cognitive ability. And if the author bothered to look, it would have been found that numeracy (from at least the 1400s on) does also. So the historical data is consistent with Lynn's hypothesis properly understood i.e., cognitive differences across time, now causing SES differences owing to the current importance of cognitive ability.

From the March–April 2015 issue of Intelligence, Two Italies? Genes, Intelligence and the Italian North–South Economic Divide:

That was a nonsense paper, with the primary argument being an Ignoratio Elenchi. It was argued that biological indexes did not predict SES centuries ago -- and therefore, biological differences can not, by way of cognitive differences, explain current SES differences. As if the relation between cognitive ability and SES could not have shifted across time, with ability becoming more important in post-industrialized economies. Of course, the author found that regional differences in education centuries ago predicted differences in cognitive ability. And if the author bothered to look, it would have been found that numeracy (from at least the 1400s on) does also. So the historical data is consistent with Lynn's hypothesis properly understood i.e., cognitive differences across time, now causing SES differences owing to the current importance of cognitive ability.

And the apparent narrowing in gaps between 2003 and 2012 is probably solely due to immigration. Census data show that from 2001 to 2011 the population in the south decreased, whereas that in the north increased significantly and this is due mostly to immigrants moving to the wealthy north.

Back in September 2010 Italianthro wrote in Refuting Richard Lynn's Italian "IQ" Study:

[b][...][/b]

Lynn goes on to attempt to correlate his fake IQs with achievement. The primary measure he uses is per capita income, which is double in the North what it is in the South. His source is the Italian Statistical Office, but he should be aware that figures for the South's economic performance are greatly underestimated because the official statistics fail to take into account a large underground economy there, according to Burnett and Vaccara (1999):

But the third factor, somewhat alleviating the second, is the existence of a far vaster private sector than ever shows up in the economic statistics. The size of the lavoro nero sector and the black market in the South clearly exceeds that of any other EU region.... In Calabria, with its dire employment figures, 84 percent of the families own their own home. What such anomalies must mean is that real income in Calabria is far higher than what is "on the books." Many among the vast numbers of officially unemployed are, in fact, partly or fully employed. They are earning no social benefits, but they are earning the daily lire that keep their families afloat. [...] A very large part of the South's hidden labor is made up of entrepreneurs, sometimes also employing black labor, and existing themselves outside official recognition, taxation, protection, control, or counting. A recent analysis concludes that "there exists in several zones of the Mezzogiorno [Southern Italy] a whole fabric of small and very small businesses that escape every census, but that work and make profits, share among themselves a serious level of production, export to other regions [of Italy] and abroad." [...] This massive sector skews all the statistics. It means that the GDP for the Italian South (and for Italy as a whole) is far from accurate. And the unemployment figures do not reflect reality.

[b][...][/b]

I was recently reminded of this when reading Raufhon Salahodjaev's article from the March–April 2015 issue of Intelligence, Intelligence and Shadow Economy: A Cross-Country Empirical Assessment:

This paper empirically assesses the influence of intelligence on a shadow economy, using data from 158 countries, over the period 1999–2007. The results provide strong evidence for the claim that intelligence is negatively associated with an underground economy. This paper establishes that, on average, a one standard deviation increase in IQ is associated with an 8.5 percentage point reduction in a shadow economy relative to GDP. The negative effect of intelligence remains intact when controlled for conventional antecedents of a shadow economy.

[b][...][/b]

The above argument is ridiculous. Italianthro attempts to reject a quantitative argument by resorting to a qualitative argument. Lynn said that per capita income in North is double that in South. And Italianthro said that the contribution of underground economy refutes the above stats. But he and the referred authors seemed incapable to give us the % of the difference attributed purely to underground economies.

What he has been able to show, is that there are potential confoundings, but how much they affect the result is not given. And yet, he "knows" that Lynn's stats is completely unreliable. How presumptuous ! And how wonderful !

Also, I don't get his point on unemployment. It's not as if IQ causes unemployment after all.

Concerning the study of Raufhon Salahodjaev, I think i'm being skeptical. I have been reminded recently of another curiosity in economics. A study between-country finding a positive correlation (between government spending and growth) and an older one within-country finding a null correlation (even slightly negative). And that has reminded me that this complexity has been illustrated many times in economics in many different subjects of analysis, but also in behavior genetics recently (e.g., the apparent lower heritability of IQ with SES and yet the equal heritability of IQ between blacks and whites). This has reminded me of one criticism of Wilkison & Pickett's book on inequality, that they are committing the so-called ecological fallacy (i.e., inferring individual characteristics from an analysis on groups).

I generally trust more within-country analyses than between-country analyses, because the correlation can easily be spurious due to some particular features of the countries. So, the study of Salahodjaev probably needs to be replicated using a different approach. In particular, the point made by Italianthro is that of the difference between southern and northern Italy. Different regions but same country. So, this limits somewhat the generalizeability and applicability of Salahodjaev's study.

Even when you control for various confoundings, I think it can be very approximative, especially when you are having an "economic variable" which is the case in Salahodjaev's study. But since controlling for various economic variables (e.g., public policies) is almost impossible to do optimally, my (relative) skepticism is justified. Again, I'm not saying the study is wrong, but I see these contradictory findings so often recently that I want to be careful now.

The above argument is ridiculous. Italianthro attempts to reject a quantitative argument by resorting to a qualitative argument. Lynn said that per capita income in North is double that in South. And Italianthro said that the contribution of underground economy refutes the above stats. But he and the referred authors seemed incapable to give us the % of the difference attributed purely to underground economies.

What he has been able to show, is that there are potential confoundings, but how much they affect the result is not given. And yet, he "knows" that Lynn's stats is completely unreliable. How presumptuous ! And how wonderful !

Also, I don't get his point on unemployment. It's not as if IQ causes unemployment after all.

Concerning the study of Raufhon Salahodjaev, I think i'm being skeptical. I have been reminded recently of another curiosity in economics. A study between-country finding a positive correlation (between government spending and growth) and an older one within-country finding a null correlation (even slightly negative). And that has reminded me that this complexity has been illustrated many times in economics in many different subjects of analysis, but also in behavior genetics recently (e.g., the apparent lower heritability of IQ with SES and yet the equal heritability of IQ between blacks and whites). This has reminded me of one criticism of Wilkison & Pickett's book on inequality, that they are committing the so-called ecological fallacy (i.e., inferring individual characteristics from an analysis on groups).

I generally trust more within-country analyses than between-country analyses, because the correlation can easily be spurious due to some particular features of the countries. So, the study of Salahodjaev probably needs to be replicated using a different approach. In particular, the point made by Italianthro is that of the difference between southern and northern Italy. Different regions but same country. So, this limits somewhat the generalizeability and applicability of Salahodjaev's study.

Even when you control for various confoundings, I think it can be very approximative, especially when you are having an "economic variable" which is the case in Salahodjaev's study. But since controlling for various economic variables (e.g., public policies) is almost impossible to do optimally, my (relative) skepticism is justified. Again, I'm not saying the study is wrong, but I see these contradictory findings so often recently that I want to be careful now.

Dear guys,
I am the author of the Intelligence and Shadow Economy paper. I do agree with the limitation of cross-national studies. However, there is also important reason in favor of using simple mean ols. The intelligence, measured by IQ scores, does not vary substantially over the short time span. Furthermore, it would be very costly to collect annual data on national IQ scores :)

Hi, thanks for the answer (I'm surprised to see you here). I agree that what I'm asking may be too much. The best alternative, then, would be to use proxies for IQ, e.g., GMAT scores available for each successive years from 2001 to 2012. I have, with Chuck, collected some data available here :
http://humanvarieties.org/2014/02/02/quick-post-lvs-national-iqs-predicts-gmat-scores-across-173-nations/

Also, if you don't want to let IQ vary over time, why not making the IQ variable stationary (see Cole & Maxwell (2003), "Testing Mediational Models With Longitudinal Data: Questions and Tips in the Use of Structural Equation Modeling") ? If you use path analyses and SEM, your model would look something like this :



where x is the independent var, m the mediator, y the dependent var. (graph taking from Cole & Maxwell 2003)

Personally, I think it's doable.

Also, it important to admit that some of the IQ scores in L&V dataset are decade or more old, this makes another huge limitation for potential reserch. I think me and my co-authors might have to use GMAT/GRE or other student assesment test results to validate robustness of our findings. What you suggested on your website, is very good observation, but I am not sure if the GMAT scores are available for longer time spans (like 1990 - 2010).

Another way to use IQ in panel setting maybe to use pooled OLS with time and country dummes, this way we have pseudo fixed effects setting, and STATA will not drop the IQ in the regression.

Also, it important to admit that some of the IQ scores in L&V dataset are decade or more old, this makes another huge limitation for potential reserch. I think me and my co-authors might have to use GMAT/GRE or other student assesment test results to validate robustness of our findings. What you suggested on your website, is very good observation, but I am not sure if the GMAT scores are available for longer time spans (like 1990 - 2010)

Here is the link. You have to register now:http://www.gmac.com/market-intelligence-and-research/research-library/gmat-test-taker-data/profile-documents/1986-87-to-1990-91-profile-of-gmat-candidates.aspx

The scores go back to 1984-1985, so for some countries you will have almost 30 years of data. Emil Kierkegaard, who runs this site, wrote a program which extracts data from tables in a file e.g., pdf -- so compiling the scores need not be as laborious for you as it was for us. If you want earlier measures, you can use "age heaping" scores. See:

http://humanvarieties.org/2014/02/11/lvs-2012-national-iqs-predict-2011-2012-gre-scores-for-114-citizenship-groups-2010-2012-toefl-scores-for-157-citizenship-groups-pisa-scores-of-migrants-from-62-nations-of-origin-across-1/

These give good data for 1890 to 1940 birth cohorts (so, given the average age of the samples, about 1930 to 1980). In a unpublished analysis, I found that the correlation between age heaping (numeracy) and NIQ was mediated by years of education. The age heaping scores can be downloaded at: clio infra: https://www.clio-infra.eu/datasets/select/indicator/411

I have Lynn and Meisenberg 2015 achievement scores, by the way. I am using them for another project. You could use these to validate other data sets.

Also, it important to admit that some of the IQ scores in L&V dataset are decade or more old, this makes another huge limitation for potential reserch. I think me and my co-authors might have to use GMAT/GRE or other student assesment test results to validate robustness of our findings. What you suggested on your website, is very good observation, but I am not sure if the GMAT scores are available for longer time spans (like 1990 - 2010)

Here is the link. You have to register now:http://www.gmac.com/market-intelligence-and-research/research-library/gmat-test-taker-data/profile-documents/1986-87-to-1990-91-profile-of-gmat-candidates.aspx

The scores go back to 1984-1985, so for some countries you will have almost 30 years of data. Emil Kierkegaard, who runs this site, wrote a program which extracts data from tables in a file e.g., pdf -- so compiling the scores need not be as laborious for you as it was for us. If you want earlier measures, you can use "age heaping" scores. See:

http://humanvarieties.org/2014/02/11/lvs-2012-national-iqs-predict-2011-2012-gre-scores-for-114-citizenship-groups-2010-2012-toefl-scores-for-157-citizenship-groups-pisa-scores-of-migrants-from-62-nations-of-origin-across-1/

These give good data for 1890 to 1940 birth cohorts (so, given the average age of the samples, about 1930 to 1980). In a unpublished analysis, I found that the correlation between age heaping (numeracy) and NIQ was mediated by years of education. The age heaping scores can be downloaded at: clio infra: https://www.clio-infra.eu/datasets/select/indicator/411

I have Lynn and Meisenberg 2015 achievement scores, by the way. I am using them for another project. You could use these to validate other data sets.

Thank you very much for such detailed review. Chuck, I was wondering you run this interesting website, that I have enjoyed reading. Why dont you publish in mainstream journals, such as Intelligence, Personality and Individual Differences, Evolution and Human Biology?

Also, is the Lynn (2015) paper is published? Cant find the link :(

Chuck, I was wondering you run this interesting website, that I have enjoyed reading. Why dont you publish in mainstream journals, such as Intelligence, Personality and Individual Differences, Evolution and Human Biology?

Also, is the Lynn (2015) paper is published? Cant find the link :(

I have been somewhat ill. I would like to publish some of the results discussed at HV along with many others that I have not got around to posting on, but I literally do not have the energy. And no one else expressed interest in writing up or building off of the analyses. Apparently, the subject matter is too taboo.

(I publish here because it is relatively easy to. And the reviewers are very helpful. I make a lot of typos and language errors -- as sometimes it hard to focus (both optically and mentally).

As for Lynn's new scores they will not be published until the end of the year. The NIQ scores are not done, but the achievement scores (NACHQ) (e.g., PISA, etc.) are. We are using them in this analysis, which has been slowly progressing for the aforesaid reason.