Friday, September 5, 2014

The Completely Bogus Economic Study Being Used To Claim that Science and Religion Are Enemies

Chris Mooney would seem to be trying to undo whatever good his previous work in pointing out that science and religion are not in an inevitable conflict was.  And it would seem that his latest column in his new campaign is pretty weak cannon fodder in that effort.   He entitles it "Study: Science And Religion Really Are Enemies After All" but the "study" he bases his column on doesn't really even address science, it addresses the number of patents produced in different geographic locations correlated with alleged measures of "religiosity" in those locations.  The study by Princeton economist Roland Bénabou and two colleagues, as is typical of the genre of economics, these days, "filters" the data taken from PEW and a number of other compilers of data and what would more honestly be called "alleged data" and subjects them to a number of "filters" to "correct" for various factors.  Some of the filters are rather odd.  Here is how Mooney describes the process and its results.

"Places with higher levels of religiosity have lower rates of scientific and technical innovation, as measured by patents per capita," comments Bénabou. He adds that the pattern persists "when controlling for differences in income per capita, population, and rates of higher education."

That's the most salient finding from the paper by Bénabou and his colleagues, which uses an economic model to explore how scientific innovation, religiosity, and the power of the state interact to form different "regimes." The three kinds of regimes that they identify: a secular, European-style regime in which religion has very little policy influence and science garners great support; a repressive, theocratic regime in which the state and religion merge to suppress science; and a more intermediate, American-style regime in which religion and science both thrive, with the state supporting science and religions (mostly) trying to accommodate themselves to its findings.

It is in the process of this inquiry on the relationship between science, religion, and the state that the researchers dive into an analysis of patents, both in the United States and across the globe. And the results are pretty striking.First, the researchers looked at the raw data on patents per capita (taken from the World Intellectual Property Organization's data) and religiosity (based on the following question from the World Values Survey: "Independently of whether you go to church or not, would you say you are: a religious person, not a religious person, a convinced atheist, don't know"). And they found a "strong negative relationship" between the two. In other words, for countries around the world, more religion was tied to fewer patents per individual residing in the country.

Those data aren't shown here, however, because in many ways, that would be too simplistic of an analysis. It is clear that many other factors than just religion (wealth, education, and so on) influence a country's number of patents per capita. What's striking, however, is that after the authors controlled for no less than five other standard variables related to innovation (population, levels of economic development, levels of foreign investment, educational levels, and intellectual property protections) the relationship still persisted.

This study is most useful because it demonstrates how even very absurd and attenuated efforts to generate buzz by promoting materialist-atheist nonsense can get published in academia and be echoed in the popular media, serving as atheist click bait among those who believe themselves to be beacons of scientific rationalism when they would seem to lack even an inclination to check facts. And even the most minor of effort to investigate the claims on which the methodology is based exposes the absurdity of it.

The fact is that the economists don't establish that holders of patents are in any way typical of the residents in a given geographic location, which would be vital for their claims to be valid.  For all they know individual patent holders could, when subjected to this same kind of dodgy simulated averaging, be more religious than the general population of the place and that that could vary from one country or state to another.  I haven't found any survey like that done.

One thing that is certain, holders of patents are not typical of residents of any country but are a tiny minority who are, so, atypical.  They are often not even residents of the place where a patent is issued.  Not to mention that most patents in some places, such as the United States, are held by "persons" who are not people at all.

The first questions that came to me were how many patents are issued in a country, to start with, how many holders of patents hold more than one of those. The number of patents issued in the United States are fewer than half a million a year, for example in 2012 that number is 468,960 in which year the population of the country is given as 313,910,000.  So, we start with a good reason to doubt that patent holders would be typical of the general population.  And while I wasn't able to quickly locate how many of the patent holders held more than one patent, I was able to find out some rather interesting facts, for example identifying the person granted a patent in a geographic location is no guarantee that that is where the person is from.

In 2013, 51 percent of the 303,000 patents filed in the U.S. were of foreign origin, according to the USPTO. That's a decrease of one percentage point compared to 2012, but about equal to the percentage of foreign patents granted every year for the past decade. To get some perspective, in 1963, only 18 percent of patents originated from foreign sources.

I don't know of what kind of "filter" the economists could use that would repair the damage that fact would do to their research method of comparing the already atypical patent holder to the imaginary religion-averaged typical resident of a geographic area.   I'm not certain but I don't think the PEW numbers they used would have included foreign residents in the geographic area for which they report their numbers.  I would like to hear how they dealt with that fact, if they didn't, merely, as it seems, ignore that inconvenient truth.   And there is worse news for Bénabou's methodology in that same article those figures were taken from.

The force of foreign innovation is not only felt in patent creation, it's also in the number of startups foreigners create in the U.S. The two are frequently related, as the company usually commercializes the patented idea or product.
Additionally, more than half of startups in Silicon Valley were founded by foreign-born entrepreneurs, according to Wadhwa and the Kauffman Foundation. (Kauffman's most recent index, released on Wednesday, also indicates that immigrant entrepreneurs are currently starting businesses at a rate roughly twice that of native-born business owners.)

I heartily recommend you read that article which goes into quite a bit of detail that gives legal and economic reasons that a lot of the patents issued in the United States are issued to foreign individuals and companies started by them, which, in itself is an insurmountable hurdle for the study Chris Mooney depends on.   The number of patents issued, not to people who are the ones who report their level of religiosity in the PEW surveys, but to corporations is rather large.

Inventors who work for private companies or the Federal Government commonly assign ownership of their patents to their employers; self-employed or independent inventors typically retain ownership of their patents. Therefore, examining patent data by the owner's sector of employment can provide a good picture of a sector's inventive work. Corporations owned 82 percent of patents granted to U.S. entities (including other U.S. organizations, the Federal Government, and independent U.S. resident inventors) in 2001.[21] This percentage has gradually increased over time. From 1987 to 1997, corporate-owned patents accounted for between 77 and 79 percent of total U.S.-owned patents. Since 1997, corporations have generally increased their share of total patents, rising to 80 percent in 1999, 81 percent in 2000, and 82 percent in 2001.

Individuals (independent inventors) are the second-largest group of U.S. patent owners. Before 1988, individuals owned, on average, 23 percent of all patents granted to U.S. entities.[22] This figure has trended downward since then, to a low of 17 percent in 2001. The Federal Government's share of patents averaged 3 percent from 1963 to 1987, eventually falling to 1.1 percent in 1999.[23] Its share remained at about 1 percent in 2000 and 2001.[24]

The methodology of the study could not overcome the fact that their proposed comparison is between apples and watermelons.   I am certain that the PEW surveys do not include the "religion" of start ups and public corporations, though perhaps they should, considering the recent and appalling Supreme Court rulings in that area.

Chris Mooney's equation of the issuance of a patent as representing "science" is also highly problematic.  Patents don't necessarily represent new science, new applications of science or even any science.  Many are issued for things that could be called "science" only by the most tortured of stretches.  In my reading I came across the curious fact that Jamie Lee Curtis is the holder of a patent for a disposable diaper which has a pocket in it to hold baby wipes.  The pocket is water proof, which is more than you can say for his article and the study it is based on.

I also remembered that I have a relative who holds a patent, one on a minor variation of a design for a drain board for a sink.  He needed to patent it because he could be accused of violating a design patent unless he protected his innovation which was done with no science happening, whatsoever.  He's a Catholic.

Update:  While shaving, it occurred to me that since patent laws are different in different countries and states,  there is not a uniform entity that is a "patent" and that the primary reason for patents being issued is not the promotion of new science but financial and, in their most allegedly idealistic explanation, a promotion of consumer product innovation.   To use the issuing of patents as a universal measurement of scientific activity and support is simply illogical on that basis alone.

Update 2:  HA!

Update 3:  In the author's graph of American states, Delaware is listed high up in the level of innovation along with Idaho.  I don't know about Idaho but the fact is that the position of Delaware is likely due to the well known fact that it is incredibly favorable to corporations incorporated under its law and so is the official location of a large percentage of U.S. corporations.  If the percentage of patent holding by corporations holds for them, that state would be expected to have a disproportionate representation in the issuance of patents.  Which would have nothing to do with the religious character of its human residents or its support for scientific activity.  I suspect there are many such factors that would skew the results for reasons not covered in the claims.

4 comments:

  1. So: do we point out that LeMaitre was a Jesuit, or Mendel a monk? That Galileo and Copernicus never rejected the teachings of the Church? That most of the thinkers of the Enlightenment, Catholic and Protestant, were religious (lay if not clergy or of orders)?

    Or do we point out that patents have little to do with "science"? Or that, given the majority of patents in America go to corporations, that discerning a pattern based on religious beliefs in the community is pretty much nonsense? Or point out that issuance of patents is more a reflection of the legal system (i.e., patent law) of the country involved, than anything else?

    Or do we just treat this "study" as the arrant nonsense it is?

    Hard to know what to do, isn't it?

    ReplyDelete
  2. Considering the complexity and variability of the two things alleged to be compared, the "level of religiosity" in the population of a geographic location and the issuance of patents, I doubt any social scientific methodology purporting to combine those could not generate enormous and insurmountable problems. I think whenever soc-sci types claim to do that their results will not hold up when held up against real life. Other soc-sci guys will have an incentive to cut them down, too. If they aren't too busy generating their own bogus studies.

    Chris Mooney is a real disapointment, I used to hold him in the number of rational, unbigoted atheists but he's clearly joined the bigots. I think it was when he went back to CFI that marked the turning point, though he's broken with them a bit.

    ReplyDelete
  3. This "study" might as well be Dawkins' remark that more European (read "white") countries have snagged Nobel prizes and other science awards, than have African or Asian (read, again, non-white) countries.

    Because science is our new measure of human worth, humanity in general, cultural value/superiority, etc., etc., etc. Which is all, as Charlie Pierce would say, "my balls."

    ReplyDelete
  4. Einstein patented a fridge with Szilard. Wonder what Simmsy would think about that.

    ReplyDelete