3. Book Review: Human Accomplishment, by Charles Murray.

This is the third article in a series on innovation and public policy. See also, story titled "Nature and Causes Innovation, and the Implications for Public Policy", November 18, 2003, and story titled "Bush Says Liberty Creates Innovation Which Creates Wealth", November 18, 2003.

A. Introduction.
B. What Murray Did With His Data.
C. What Murray Did Not Do With His Data.
D. Decline of Accomplishment.
E. Religion and Aristotelean Ideals.
F. Conclusions.

November 19, 2003. One common denominator of many of the recent speeches, papers, and reports about innovation and public policy is that the speakers and writers often refer to historical support for their arguments. However, while it is common to state that history supports this or that proposition, very few cite any historical facts, let alone historical works.

Now, along comes veteran conservative controversialist Charles Murray of the American Enterprise Institute (AEI) with a book that is a history of innovators. It is titled Human Accomplishment: the Pursuit of Excellence in the Arts and Sciences, 800 B.C. to 1950. The book was released on October 21. It is 688 pages, with a substantial index, bibliography, endnotes, and appendices. Statistically minded readers will revel in his tables, graphs, and multivariate regression analysis results.

A. Introduction. Human Accomplishment is quantitative study. He identifies the leading 4,000 innovators over the last three thousand years, and assigns scores indicating the extent of their contributions. He also collected data on such things as when they lived, where they lived, and what were some of the surrounding circumstances in which they lived.

Murray (at right) finds that "At irregular times and in scattered settings, human beings have achieved great things. They have discovered truths about the workings of the physical world, invented wondrous devices, combined sounds and colors in ways that touch our deepest emotions, and arranged words in ways that illuminate the mysteries of the human condition."

Having found that innovation occurs at irregular times and in scattered settings, Murray then makes some significant strides, through empirical analysis, towards identifying some conditions that have been conducive to innovation and accomplishment, and others that have not. Since some of these conditions, such as individual freedom and the presence of elite universities, can be either advanced or limited by governments, the book has public policy implications. Although, Murray asserted at a lecture at the AEI on October 27 that the book "has no apparent public policy ramifications".

Murray's methodology is to quantify human accomplishment in many of the sciences, technology, art, and literature over a nearly three millennium period (800 B.C. through 1950) by resorting to what others have written about these fields. He developed a huge database of significant figures (scientists, inventors, artists, writers, and others) and significant events. He did not compile the lists based on his own subjective assessments. Rather, he used a reputational approach. That is, he collected leading encyclopedias, histories and surveys, and looked to see who got mentioned, and by how many sources. He also considered how many pages were devoted to each innovator or creator (he calls them "significant figures") in these sources. Based on this data, he calculated scores, on a scale of 1 to 100, with 100 being the maximum possible.

For example, Isaac Newton and Albert Einstein both attain the maximum possible score of 100 in the field of physics. Leonhard Euler scores 100 in mathematics, James Watt and Thomas Edison both score 100 in technology.

He has tables in the book listing the top twenty in each field. The rest of the significant figures are listed in appendices at the end of the book. Since his study only examines significant figures through 1950, most information technology innovators are missing from the book. However, Johan Gutenberg, famous for his printing press, just misses the top twenty in technology with a score of 23, and has to settle for a listing in the appendix. Charles Babbage, who worked on his "Difference Engine", makes the top twenty in technology, at 11th place, with a score of 33 out of 100. Guglielmo Marconi is 6th in the field of technology with a score of 50. John Von Neumann, who developed the structure of digital computers, nearly makes Murray's top twenty lists in both mathematics and technology. Vannevar Bush, author of "As We May Think", scores a respectable 16 in technology.

Readers may look at the lists in this book and conclude that some people are overrated, while others are underrated. Nevertheless, these scores and rankings reflect combined opinions of a large number of encyclopediaists and historians who have studied the various fields covered by the book. For example, Murray consulted 17 sources on technology.

Murray also collected data for a large number of other variables, such place of birth, date of birth, where these individuals grew up, where they worked, and the dates when they flourished. He also collected data on such factors as the political system of the various locations of these individuals, economic growth, whether there was an elite or other university nearby, whether there was war or domestic turmoil, and other variables.

In short, whatever the shortcomings of his methodology might be, he has amassed an amazing body of data that one might use test various hypotheses regarding what leads to innovation, and what does not.

He then goes on to conduct some useful statistical analysis. Unfortunately, he leaves untested many testable hypotheses that might be relevant to many public policy debates today.

Then midway through the book he argues that the rate of accomplishment has been on a steady decline for over century, including in science and technology. Actually, his raw data does not demonstrate this -- he manipulates his data. Finally, he argues that the reason for this decline has to do with religion, transcendental goods, and the values espoused by Aristotle in Nicomachean Ethics. Whatever the merits of this section of the book might be, it is of little relevance public policy debates regarding promoting innovation.

B. What Murray Did With His Data. Murray found that accomplishment has been scattered, rather than uniformly distributed. There has been, at times, a concentration in some (but not all) Italian Renaissance cities, in France, the low countries, England, certain German cities, and elsewhere. One can view his maps to see scatter plots of where innovation has occurred at various times in various fields.

For example, in Italian Renaissance, accomplishment was concentrated in Florence and Venice, but not Naples, Genoa, or Milan. In the British Isles, accomplishment centered around London, the industrial north, and lowland Scotland.

Murray is always controversial. He does not defy expectations in this book. Thus, he writes that most innovation and creation has been accomplished by men, not women. It has been accomplished by Europeans, not other ethnic and cultural groups. It has not even included Americans.

He finds that freedom is important. He writes that "Streams of accomplishment are fostered by political regimes that give de facto freedom of action to their potential artists and scholars". Although, his notion of freedom is limited. It basically means freedom to express and create to the innovative classes. It does not entail democracy. Moreover, he finds that totalitarianism suppresses innovation.

He finds, based on statistical analysis, that war and civil unrest has not disrupted innovation.

He finds, based on his data, that there is a relationship between elite universities and human accomplishment, but not between all other universities and accomplishment. That is, he found innovation is more likely to occur when there is an elite university nearby.

He finds, again based on data analysis, that innovation is self reinforcing. That is, where there has been innovation, there is likely to be more.

He also has a section that supports the argument that religious liberty promotes innovation. He looked at innovation by Jews. He found that Jews had "sparse representation in European arts and sciences through the beginning of the 19C", but that within a century Jews were disproportionately represented among the lists of innovative people in almost all fields (except astronomy). He writes that this coincided with the emancipation of Jews. Previously there had been laws denying legal rights, access to universities, and public office.

C. What Murray Did Not Do With His Data. Murray assembled a database with a large number of variables and observations, that could be used to test various hypotheses regarding what leads to innovation. But, he makes no attempt to examine many important issues.

First, Murray has nothing in his book about the impact of intellectual property rights (IPR) protections. He leaves out any discussion of patent or copyright. He could have examined data from the 19th and 20th Centuries -- this would still be a very large sample size -- to see if more innovation took place in locations with IPR protections.

The acknowledgements at the beginning of the book, and the praise on the back cover, are noteworthy. Murray devotes a paragraph in his acknowledgements to Michael Novak. He is the author of The Fire of Invention, the Fuel of Interest: On Intellectual Property. Murray credits Judge Richard Posner with reading an entire draft of the book. Posner's web site at the University of Chicago states that he will be the co-author, with William Landes, of the book titled The Economic Structure of Intellectual Property Law (Harvard University Press forthcoming). On the back cover, Paul Johnson is the first to praise Murray's book.

Both Murray and Johnson (who is also a former AEI scholar) argue that there was a large output of innovative activity in England and Scotland during the early industrial revolution. Johnson gives great credit to the British patent system.

Two of these three find patent protection to be important. One (Posner) gave a lecture at the AEI criticizing patent protection on the grounds that it curtails innovation while providing monopoly rents.

So then, if incenting creation is as the center of public policy debates regarding intellectual property law, and Murray's conservative colleagues are writing on the subject, why did he ignore patent and copyright laws when he sought to explain why innovation occurs in some places, but not others?

TLJ spoke with Murray on October 27, and asked him why did he not examine the affect of patent protection? He responded that "Now, in the case of patent law, you obviously have a problem, because you can only go back a little while, to the point at which patent law began. So, in an empirical sense you would have to truncate your analysis a whole lot."

Second, Murray does not examine the effect of government spending, or spending by princes and prelates, on art, architecture, research, or science. Today, many governments spend large amounts on research and development, universities, and patronage of arts and culture. Previously, many monarchs, princes, and popes spent lavishly.

Many people who speak about promoting innovation today advocate increasing government spending on innovation related activities.

TLJ asked Murray why he did not examine this subject in his book. He responded that "things like that again are restricted to a relatively relatively recent past, and to relatively few places".

Third, Murray does not examine the impact of education upon innovation. How much education have innovators had? What kind of education did they receive? He was asked about this at a lecture at the Cato Institute on October 28. He said, "Do I have any coherent thoughts on how education affects excellence -- no."

Fourth, he does not examine immigration  in detail. He could have, since he collected data on places of birth, and places of work, of his 4,000 significant figures. He does some analysis of his data. But, it is brief, and he applies some overly restrictive definitions of immigration. For example, he states in a footnote that he did not treat as immigrants people who "were moving within language ... settings". Thus, while Renaissance Italy was divided into many different jurisdictions, with their own rulers, laws and customs, he treats this as one entity. In Murray's analysis, someone who emigrated from Naples to Florence did not emigrate. Nor did he count movement within the German speaking areas, even though until near the end of his study, there were a great multitude of states with German speaking populations.

He also applies a narrow view of immigration. For example, in discussing immigration to the U.S., he writes that he only counted "people who grew up in a foreign country and then moved to the United States to conduct most of their important work ..." Thus, established scientists who fled Hitler were not immigrants under Murray's methodology. But arguably, they contributed to the innovative capacity of the U.S.

Murray could have done much more with his data. He could have examined whether people were moving from places with less IPR protection to more IPR protection, from places without universities to places with universities, from places without religious freedom to places with religious freedom, from places without market economies to places with market economies, and so forth.

D. Decline of Accomplishment. Murray makes the argument that while the pattern varies some from category to category, "the overall story is one of recent decline, usually starting sometime in the 19C". No other writer or speaker who is covered in this series, makes this argument.

For example, Bruce Mehlman, the Department of Commerce's Assistant Secretary for Technology Policy, has given a speech titled "21st Century Policy Challenges for American Innovation Leadership" in which he argues that innovation is expanding rapidly. He states that "It has been estimated that about 90% of all scientific knowledge has been generated over just the last 30 years, by about 90% of all scientists and engineers who have ever lived ... and now are living and working. Both the rate of generation of new technology, and the number of workers globally engaged in S&T may double again over the next 15 years. With computer processing power and genomic data also doubling every 18 months or faster, it is no exaggeration to predict that there will be more change in the next 30 years than we saw in all of the last 100."

So, how does Murray reach his conclusion. First he resorts to his huge collection of data. When he first provides graphs showing the number of significant persons (the vertical axis) by year (horizontal axis) the line remains low and roughly horizontal from ancient times through around 1400. Then, it takes off. Whether one looks at the graph for scientific persons, or all significant persons, it appears that as the date approaches 1950, the number of significant persons approaches infinity. That is, his raw data shows that innovation has been on the rise for half a millennium, and is not letting up.

Murray did not stop here. First, he transformed his raw data by dividing, for example, the total number of significant persons at any point in time by the population, thus providing a per capita innovation statistic. But, this still does not give Murray a record of recent decline. So, he then performs a second data transformation. Rather than dividing by the actual population, he divides by what he terms the "de facto population". Basically, by this he means people with education, prosperity, urban residence and "exposure to the larger culture". He estimates that in 1400, this only included 10 percent of the actual population, while it grew to 75 percent by 1950. (He only counts Europeans in this manipulation.) This, then, produces a downward sloping innovation per (de facto) capita line.

Murray explains his "de facto population" manipulation. He does not, for example, offer empirical support for the propositions that poor, humble or rural people cannot go on to innovate. Paul Johnson, whose writings are the subject the next article in this series, has examined this subject. He finds that numerous scientists and technologists (many of whom are listed in Murray's book) have come from humble backgrounds, had little or no formal education, and/or started work as laborers.

Nor does Murray explain why per capita measurement should be used in the context of innovation, which largely results in non-rivalrous goods. That is, one person's ability to understand, or consume, an idea, is not precluded by another person's consumption of that idea. Per capita measurements are typically applied in the context of rivalrous goods, and income.

In short, Murray's statistical methods are thorough and objective up to the point where he argues that innovation is on the decline. Then, he manipulates his data.

He evens writes defensively about his conclusion that scientific and technological innovation is on the decline. He writes that "It may seem absurd to argue that science and technology declined in any sense during any part of the 19C or 20C. Those two centuries saw more progress in the well-being of mankind, by orders of magnitude, than any other two centuries in the history of mankind, and it was made possible by unprecedented strides in science and technology."

He suggests that his reputational method of quantifying innovation tends to rate higher the first people to work in their fields, while subsequent innovators tend not to be given nearly as large a place by historians.

E. Religion and Aristotelean Ideals. The final chapters of Murray's book assumes as true the premise that innovation is on the decline, and then offers an explanation for why this is the case. This section of the book is based on a faulty premise, is largely a polemic essay based upon subjective opinion, and has few if any public policy implications.

At this point he abandons empirical analysis. He writes that "At this point I am no longer dealing with ideas and hypotheses that lend themselves to clear cut tests, though quantitative data can inform them". Later he concedes, "I cannot supply quantitative measures ....", and the analysis is "less quantitative, more speculative, and definitely more opinionated."

For whatever it is worth, this is his argument. "I proceed from the view that accomplishment in the arts and sciences is one manifestation of a characteristic of human nature discussed at length by Aristotle in books seven and ten of the Nicomachean Ethics." Murray focuses on four things: purpose, autonomy, organizing structure, and transcendental goods.

First, regarding purpose, he argues that "A major stream of human accomplishment is fostered by a culture in which the most talented people believe that life has a purpose and that the function of life is to fulfill that purpose."

Second, regarding autonomy, he argues that "A major stream of human accomplishment is fostered by a culture that encourages the belief that individuals can act efficaciously as individuals, and enables them to do so."

Then, he argues that the ancient Greeks, Renaissance Catholics, and Reformation Protestants possessed purpose and autonomy. Would be innovators today are a bunch of post-Darwin, post-Freud, secularists and nihilists. The former were creative because of their Aristotelean ideals. The later are not creative because they lack these ideals.

Then, regarding "organizing structure", he argues that "The magnitude and content of a stream of accomplishment in a given domain varies according to the richness and age of the organizing structure." Finally, he argues that "A major stream of accomplishment in any domain requires a well-articulated vision of, and use of, the transcendental goods relevant to that domain." These may include a sense of goodness, truth and beauty. Truth, of course, is the transcendental good most important to science, writes Murray.

Murray argues that these Aristotelean virtues, possessed by the Greeks, and the Renaissance and Reformation Christians, but in large part lacking elsewhere, account for patterns of accomplishment.

F. Conclusions. Murray has devised and implemented a methodology for quantifying innovative and creative activity over the centuries. When data of this sort is combined with other data about these innovative and creative people, their surroundings, and the laws and policies in place when they lived, it provides the opportunity to statistically test a number of hypotheses that are relevant to some of the public policy debates regarding promoting innovation.

Murray developed this methodology, built a database of 4,000 people, and conducted some useful statistical analysis. However, he left unexamined many issues. And unfortunately, his book detoured into a polemic essay of little relevance to public policy.