Am I Missing Something or Is Malcolm Gladwell a Fraud?

October 24, 2013
Another Jonah Lehrer?

Another Jonah Lehrer?

In an interview at Tuesday, writer Malcolm Gladwell is on The Daily Show plugging his new book David and Goliath. It deals with the notion of “desirable difficulties”, how you can overcome seemingly debilitating conditions and turn them into assets. One such condition, supposedly, is dyslexia, a disorder characterized by difficulties in basic language skills, like reading and writing.  So how is that desirable?

Gladwell mentions the phenomenon of successful and dyslectic entrepreneurs saying, “they didn’t succeed despite of their dyslexia, but because of it” and that their childhood problems with this condition “forced them to learn all kinds of strategies that ended up being more important.” More generally, he remarks that when asked about the reasons for their success they often speak of all sorts of problems as a source of their success. He seems blissfully unaware of the idea that this may be how people enjoy thinking about themselves – as the guy who beat the odds with his guts and determination. I’ve at least never heard anyone attribute their success to having a rich and powerful dad. And yet isn’t that how George W Bush became president, a man who many believe have dyslexia or some similar condition?

Or maybe Gladwell is well aware that this type of success stories sell, especially if they’re made to look as if they also tell the reader something new and interesting about human nature. Am I being cynical here? Let’s look at the evidence.

The only real, non-anecdotal evidence Gladwell presents is that a disproportionate number of successful entrepreneurs are dyslectic. This claim seems to be based on a study by Julie Logan, a professor of entrepreneurship at Cass Business School in London. In it Logan says,

This research set out to find an explanation for the high proportion of dyslexics among successful entrepreneurs…

This is a sneaky way of putting it because the word “high” suggests that it would be disproportionate, as Gladwell claims. But the study found no such thing. In fact, it didn’t even attempt to, since it has a sample size of ten (10) individuals who were selected because they are dyslectic and successful entrepreneurs! It’s interesting to note in the clip from The Daily Show when Gladwell mentions four such dyslectic entrepreneurs and and the host Jon Stewart jokingly wonders if they are only four and Gladwell laughs and says, “there is significantly more.” This way it appears as if he had a few dozen, but from what I’ve gathered there are only the ten from Logan’s study. Are an additional six really significantly more?

Anyway, where did Gladwell come up with this idea that dyslectics are overrepresented among prominent entrepreneurs? I haven’t read the book but in the Barnes & Noble Book Blog review, Amy Wilkinson states that it contains the claim that some 30 percent of successful entrepreneurs are dyslectic. But Logan’s study doesn’t state this anywhere. However, in her review of her own previous research she mentions that she found that 35 percent of all American entrepreneurs – successful or otherwise – have dyslexia. So it seems that when Logan cautiously stretched the truth with her “high proportion”, Gladwell took this one step further and fabricated the idea of a disproportionate amount of dyslectics in this group.

I guess I’ll have to read the book to be sure, but this looks very suspicious.

Update: I’ve now read the parts of the book that concern dyslexia. His claim is found in the quote,

An extraordinarily high number of successful entrepreneurs are dyslexic. A recent study by Julie Logan at City University London puts the number somewhere around a third. The list includes many of the most famous innovators of the past few decades.

And as I said above, Logan only claims that roughly a third of all American entrepreneurs are dyslectic. She pushes the same idea as Gladwell, that they are successful but all she can provide is self-rated competence and the rate at which the dyslectics grow their businesses – neither of which is evidence of actual success.

On The Daily Show he makes the same claim,

If you look at groups of very successful entrepreneurs or professionals you will find a much greater than expected number of dyslectics among their ranks.

Now he examplifies with people like Richard Branson, Charles Schwab etc giving the impression that dyslectics are disproportionate among the entrepreneurial stars.  But the research he is basing his claim on is still made on a sample of  all dyslectic entrepreneurs. It’s not a subset as Gladwell implies. And their success has not been established. This is just a simple case of having an idea, looking for the evidence, not finding it – and then making it up.


The Personality and Geography of the Entrepreneur

May 10, 2013

The Entrepreneurial Personality Profile

Creating new jobs is one of the top priorities for all politicians these days. This issue always ends up being discussed in terms of taxes and spending, sometimes in terms of education, but never ever in terms of people. This is ultimately the old Enlightenment view that human behavior, good or bad, is all about arranging external factors. People are assumed to respond to whatever plans or reforms the political leaders come up with.

And yet, if we look at indicators relevant to job creation it becomes clear that people do what they don’t respond the way they’re supposed to. Take self-employment for example. It has been very stable around 10-15 percent for many years now, seemingly unaffected by shifting policies. Before that it was a bit higher due to more people working in agriculture. A reason for this can be found in entrepreneur research, which shows that certain personality traits make for a good entrepreneur – just like other traits make for a good police, a good nurse etc. These traits are fairly stable over time explaining the relatively constant percentage of entrepreneurs in the American population (and most likely other populations as well).

The psychological research on people with this profession suggests that, in terms of the Big Five model, the entrepreneur profile is high on Extraversion, Openness and Conscientiousness, and low on Agreeableness and Neuroticism. The reasons for this particular mix may seem obvious, except perhaps the low Agreeableness. This is probably because an entrepreneur must be tough and not back down in negotiations. A lot of people complain about how nerve-wrecking it is to negotiate how much you should be paid with your employer, but an entrepreneur has to make similar negotiations on a regular basis, so it makes sense that they need thicker skin than the rest of us.

A Rare Commodity

By all accounts, the traits that make up the entrepreneurial profile are highly inheritable and stable from early adulthood. This means that the number of entrepreneurs (or job creators) will be limited by the amount of people who have the profile and the strength of its relation to entrepreneurial activity. So when politicians say they will create new jobs they appear to indulge in a fantasy. More realistically, they can help entrepreneurs to create jobs. Like a secretary can help an architect or a doctor to do his job more efficient.

I don’t want to sound all Ayn Rand here; I understand that the world would be pretty unpleasant without polices and nurses (and very inefficient without secretaries). We shouldn’t worship these people like some libertarians do, but at the same time we have to accept that a lot of people can be nurses and polices but few can be entrepreneurs – and entrepreneurs are more important. While other research has shown how intelligence affects the economic development, on both an individual and national level, it may be that entrepreneurship is even more important. The intelligent person can maintain the order of things, but entrepreneur is the driving force in technological development. He or she is the reason you’re reading this on a computer (or similar device), why you have fresh food stored in your fridge and why you can travel to the other side of the planet in less than two days.

So while libertarianism may be little more than a personality disorder, Rand still has a point. These individuals have a profound effect on society and to understand the world around us we need to understand them better.

Drawing the Map

One important question regarding entrepreneurship is whether all countries and regions are equally blessed with this commodity or if it’s unevenly distributed. Obviously some countries are more entrepreneurial than others, but that may be for political reasons, because while you can’t create entrepreneurs, you can suppress them. There are probably people in North Korea who have ideas and want to start businesses too.

But if we look at the personality side of this issue, international comparisons suggest that personality traits are unevenly distributed, so if two culturally similar nations had the exact same policies implemented one would still be more innovative than the other for purely genetic reasons. These international studies have been criticized since they sometimes make little sense – Japan scoring lower than Nigeria on Conscientiousness, Italians being introverted and things like that. But even with these clear mistakes, the overall results from the biggest international study of this kind are validated by culture level measures.

But we might get a more reliable test of the importance of the entrepreneurial personality by looking within nations. As I mentioned in my previous post, the data compiled by psychologist Peter Jason Rentfrow, shows an uneven distribution of Big Five factors within the United States. This means that the average levels of some states will be more similar to the entrepreneurial profile than others. But does this also mean that average levels of traits characterizing an entrepreneur on the individual level will translate to more entrepreneurial activity on the regional level?

A case for this is made in a recent study by German psychologists Martin Obschonka and colleagues, published in the Journal of Personality and Social Psychology earlier this year. They used Rentfrow’s data to create a measure of Entrepreneur-prone personality profile for the different states in America. This is what they came up with,

State-level map of the entrepreneur-prone personality

State-level map of the entrepreneur-prone personality, the darker the more entrepreneurial.

As you can see the East-West divide of Neuroticism I discussed in my previous post is present here too, although less pronounced since the other traits weigh in as well. But it’s clear that the distribution is uneven and clustered.

Obschonka then compared this ranking with measures of entrepreneurship, such as the Kauffman Index (measuring business startups) and self-employment rates. They found that even when controlling for state-level measures of prosperity, race, age and gender distribution, they got correlations around 0.35 between the personality profile and the Kauffman Index. Here is the map of entrepreneurial activity for comparison,

State-level map of entrepreneurial activity, the darker, the busier.

State-level map of entrepreneurial activity, the darker, the busier.

As expected from the correlations you can see a resemblance in that there is a cluster in the West. There are some states that break the pattern in a conspicuous way: Mississippi and Maine have the wrong profile but perform well while Nevada and Washington are the other way around. This may be due to the fact that both these measures used are fairly crude. As far as I can tell, The Kauffman Index counts all business startups – hotdog stands and high tech ventures alike. And the personality profile is built on the assumption that all traits are equally important which, if true, seems like a happy accident. (Rentfrow was kind enough to send me his data in an Excel file so when I have the time, I’ll be looking to see if a weighted measure could give a better picture of the entrepreneurial personality as well as comparing it to lots of other interesting variables).

Obschonka and his colleagues then went on the make regional-level studies of American Metropolitan areas, the UK and Germany. They got the same result with similar sized correlations but couldn’t reach statistical significance due to the small number of regions. So while preliminary, these findings suggest that their results hold not just for America but within other countries too, and most likely between countries as well.

Implications

This study is the first of its kind and the positive result no doubt means it will be followed by others. The results, should they hold up, can explain things like why people leave California for Utah or why Japan leveled out, and maybe why China isn’t the future either, but perhaps why Chile might be. It’s also interesting to note that entrepreneurial hotspots are not clearly linked to the large metropolitan areas, that are often hyped as being at the forefront of just about everything. Judging by both personality and activity it seems like states like Colorado and New Mexico show more promise than states containing larger cities, like California and New York. Then we have the fact that these clusters were most likely in part created by migrations and can be changed by future migrations.

It’s a whole new field of research opening up here – doesn’t even have a name yet but I’m sure it will soon enough – and it will be highly interesting to see what comes out of it in the future.


%d bloggers like this: