Gideon Lewis-Kraus wrote a long piece in New York Times which is quite critical of the ancient DNA lab at Harvard led by David Reich. It has generated a lot of discussion and comments, themselves ranging from mildly to very critical – by Razib Khan, Steve Sailor, Greg Cochran, and others. The NYT article is not particularly well-written — too long and rambling, a lot of unnecessary details, flowery prose. Here’s how it starts:

A faint aura of destiny seems to hover over Teouma Bay. It’s not so much the landscape, with its ravishing if boilerplate tropical splendor — banana and mango trees, coconut and pandanus palms, bougainvillea, the apprehensive trill of the gray-eared honeyeater — as it is the shape of the harbor itself, which betrays, in the midst of such organic profusion, an aspect of the unnatural. The bay, on the island of Efate in the South Pacific nation Vanuatu, is long, symmetrical and briskly rectangular. In the expected place of wavelets is a blue so calm and unbroken that the sea doesn’t so much crash on the land as neatly abut it. From above, it looks as though a safe harbor had been engraved in the shoreline by some celestial engineer.

Nevertheless, I read the whole thing because it does raise a number of important issues. I’ll talk about two: the tension between general and particular and the issue of”oligopoly.”

On the first issue, Lewis-Kraus doesn’t hide where his sympathies lie. Here’s how he characterizes the two alternative intellectual attitudes:

those bewitched by grand historical narratives, who believe that there is something both detailed and definitive to say about the very largest questions, and those who wearily warn that such adventures rarely end well.

I am obviously on the other side of the barricades, as I am keenly interested in understanding general principles explaining the evolution and dynamics of human societies. But, more broadly, I argue that the whole dichotomy is false. Human societies are very diverse, and in any particular instance — state formation in Java, imperial collapse in Italy — such events are a result of an interaction between particular factors, specific to the locality and time, and general processes of state formation and dissolution. Both the diversity and general principles are interesting and worth of study. In fact, the Seshat project does both. We document the differences between past societies in different places and different times, and then capitalize on this variability to test theories about general principles underlying the evolution of complex societies.

Perpetuating this false dichotomy, in fact, harms our collective enterprise of studying and understanding history, which is best done by a collaboration between specialists and generalists (as we practice in the Seshat project).

On the second issue, oligopoly, my take is less critical. A good nuanced discussion is by Razib Khan. See also this twitter thread. I agree that it has a potential to become a real problem. But whatever we do, we don’t want to harm the good thing we already have. Thanks to the mega-labs, like Reich’s, we are making extraordinary progress in making ancient DNA data available for the students of the past.

I am on the “data consumer” side of the equation, and so I want the data to be published as soon as possible, and to be as easily available, as they can be. Full disclosure: I am currently in discussion with several members of the Reich Lab that may lead to a collaboration, which will combine our cultural data in Seshat with aDNA data (I’ll write about the project in this blog soon). Last summer I wrote to David and his colleagues and they quickly and positively responded. The Harvard group appears to be genuinely interested in expanding their research focus from just genetic data to how aDNA results fit with other kinds of data.

What we need is not to nurture resentment (which the NYT article seems to do), but to work together. Because there are so many gaps in data about past societies, whether it’s genes, pots, languages, or institutions, we need a “totalizing” approach to history, which brings together everything we know about past societies. Only with such an approach we will be able to interpolate between the small islands of light and figure out what happened in the rest of the dark ocean of human history (see my previous post on a similar problem).

Image credit: A DNA analyst sands a bone sample to remove surface contaminants before grinding it into a fine powder and extracting its DNA.

What was the quality of life for people living in historical and prehistoric societies? One particularly important dimension of quality of life is freedom from violent death. How high was the probability of being murdered by another person? Modern statistics that express violent death rates per 100,000 people per year don’t extend very far back in time. Obtaining good numbers for even well-known historical societies, such as Medieval European ones, or Classical Rome and Athens, is very hard. Once we get into prehistory, it becomes even more difficult.

Social scientists of various kinds have very different opinions about whether life in the past was more violent or more peaceful than today. My favorite example of what the anthropologist Lawrence Keeley calls the “pacification of the past” is early Mayan archaeology. The first archaeologists who studied the Maya imagined them as peaceful and wise farmers practicing low-intensity agriculture in the “jungle”, ruled by priestly elites. At particular seasons, early scholars thought, these people would congregate around the temples and perform solemn rites to express their awe of the mysterious universe.

The reality is shockingly different. Maya lowlands were densely settled (only now are we learning just how densely, thanks to new technologies employed by archaeologists). They were ruled by a warlike and rather bloodthirsty elite who fought incessant wars against each other (and were often killed or captured and then sacrificed by the victors). The peaceful Maya of the Classic Period are a fantasy.

A reconstruction of a Maya painted mural from the late 8th-century CE Temple of the Murals at Bonampak, Mexico (Room 2). Detail showing a battle scene. © Heather Hurst, Bonampak Documentation Project, Yale University / Bridgeman Images.

Lawrence Keeley’s book War Before Civilization was very important in turning the tide. Suddenly archaeologists started to see evidence of violence where previously their eyes slid over it (or at least, they didn’t deem it worthwhile to publish such data). An arrowhead embedded in a rib here, a massacre site there—the evidence piles up.

In The Better Angels of Our Nature Steven Pinker popularized this new willingness among scholars to study violence in the past. He proposed that the past was extremely violent, but that as civilization evolved (most importantly, as the ideas associated with the Enlightenment emerged), violence declined.

Unlike Keeley’s book, which was “broadly known in narrow circles,” Pinker’s message resonated with the general public, but also triggered a violent reaction from many anthropologists. I’ve written about this before (see The War over War).

So who is right? Now, thanks to a new project that the Seshat Databank has begun with the Institute for Economics and Peace (see here), we will be able to answer questions like this.

Archaeologists and historians have collected a lot of information about violence in past societies, but it is very patchy. These bits of data are small islands of light floating in a dark ocean. The general idea of the approach we use is to collect as many different kinds of knowledge as possible and use them in statistical analyses that utilize what data are available, without being hampered by the gaps.

The variables that we are interested in coding come in two clusters. First, there are direct measures (e.g., battle casualty statistics) and indirect proxies (e.g., skeletal trauma) for death rates. Second, there are predictor variables that may have explanatory power to indicate which societies are more violent and which are less. One hypothesis with a good chance of yielding reasonable predictive power is that various aspects of social complexity affect rates of violence. For example, larger societies (with large populations and territories) may have a lower death rate due to interstate warfare simply because they remove most of such conflicts to the frontiers. Or better governed societies (higher on the Seshat governance sophistication scale) may have lower violence rates because they more effectively maintain internal peace and order.

British soldier on the Western Front, France, during the First World War, surrounded by mountains of empty shell cases. Taken by Tom Aitken in 1917. © National Library of Scotland.

Because different kinds of violence—due to interpersonal conflict, political violence and internal warfare, or interstate warfare—have different drivers, our approach will investigate them separately.

Ideally we aim to estimate the overall violent death rate resulting from all the different kinds of violence. This includes many types and causes of violence: from inter-polity warfare to the homicide rate from interpersonal conflict. Following the usual approach to quantifying homicide rate, we define the main variable of interest as the number of people killed by other people per year per 100,000 population. It is understood that because our knowledge about past societies has many gaps, any estimates we obtain will have much uncertainty associated with them. Thus, the secondary goal of the analysis is to quantify this uncertainty so that we can answer the question: how well do we know what we think we know?

We already started data collection last fall and plan to analyze the results in late spring–summer. So we should have interesting results to report by this fall.

When I became interested in what eventually became Cliodynamics, initially I thought that I would just play with some mathematical models of historical processes, because I had read many times before that there is very little quantitative data in history. In fact, the opposite is true. It turned out that history has massive amounts of data with which we can test our theories, and this empirical corpus continues to grow. The main source of new quantitative data is the clever use of “proxies” — indirect quantitative indicators of various processes of interest in historical dynamics. I see articles publishing such new data almost weekly.

The latest one is Linking European building activity with plague history by a team of dendrochronologists based in many labs across Central and Northern Europe, led by Fredrik Ljunqvist. Dendrochronology is the method of dating tree rings, and it allows us to pinpoint the date when a tree was felled down to a year. Ljungqvist’s group went around Europe taking cores from wooden beams in old houses and then dated them. So far they have collected nearly 50,000 such dates for building construction.

The inscription on this house in Lübeck claims it was built in 1535. I wonder, would a dendrochronological date agree? (Photo by the author)

Because new houses are usually built to accommodate extra population, the distribution of dated building years provides us with a great quantitative proxy for population increases (see panel A in the chart below):

From Figure 1 of Ljungkvist et al. 2018

One thing to keep in mind is that because old houses are constantly destroyed, the probability that a 13th century house would survive to the present is much smaller than such probability for a 17th century house. Thus, the felling dates data should be detrended. If one does so, then two great oscillations become even more apparent, with slowdowns in building new houses reflecting the two periods of general crisis in Europe: the Late Medieval crisis of the 14th and 15th century, and the General Crisis of the Seventeenth Century. You can read more about these cyclic crises in our book Secular Cycles.

The second chart (panel B) in the figure shows the distribution of plague outbreaks, recorded in historical sources, from a well-known compilation based on the pioneering work by the French demographer Jean Nöel Biraben in the 1970s. As the article title indicates, the main factor, potentially explaining building slowdowns, which Ljunqvist et al looked at, is mortality resulting from epidemics. Here I must temper my praise for the article with some criticism. While the data themselves are wonderful, I am not sure that relating them to plague outbreaks is the most interesting thing one could do with them.

An old house in Toulouse (photo by the author)

As the authors themselves acknowledge, European population, and thus building activity, started to decline well before the Black Death of 1347-1352. The causes of this decline are actually well known, and are due to worsening structural-demographic conditions in Western Europe. Furthermore, and even more interestingly, population in most countries did not start growing immediately after the cessation of plague outbreaks. In our book Secular Cycles we discuss the possible reasons for late medieval England and France, and come to the conclusion that the factor that held back population growth was incessant socio-political instability (the Hundred Years War in France, and the Wars of the Roses in England).

Ljungqvist and co-authors propose a similar explanation for the Seventeenth Century population decline in Germany (the Thirty Years War), but they treat it as a singular event. In fact, the Thirty Years War was part of the General Crisis of the Seventeenth Century that affected most of Eurasia, from England to Japan.

Some years ago Walter Scheidel and I proposed that we can use the frequency of coin hoards as a quantitative proxy for internal warfare (see our article, Turchin P, Scheidel W. 2009. Coin Hoards Speak of Population Declines in Ancient Rome. PNAS 106: 17276-17279.). One could use a similar approach in the analysis of the data collected by Ljungqvist et al. As an example, here’s what coin hoard data for Bohemia (modern Czech Republic) look like:

You can clearly see civil wars associated with the Late Medieval and the Seventeenth Century crises (the Hussite Wars and the Thirty Years War). Unfortunately, Czech building dates are not yet publicly available. It will be very interesting to see whether coin hoard data provide a better predictor for the cessation of building activity, compared to the plague data.

Fredrik Ljungqvist has graciously shared with me data on another region, Central and Northern Europe (data were collected by Hans Hubert Leuschner of University of Goettingen). The coin hoard data I have for this region has fairly crude resolution (50-year intervals), but here’s what happens if we plot these two proxies together:

The blue curve is detrended building activity, and the red bars are coin hoard per half-century. It would be better to resolve coin hoards at shorter time intervals (which is something the Seshat project will soon do), but you get the rough idea—peaks of coin hoards are associated with troughs in building activity.

More generally, there are many more variables whose effects could be investigated. We provide a lot of such proxies in Secular Cycles. For example, we show the dynamics of temple or church building activity for several secular cycles (these buildings can be well dated from historical records). It would be interesting to see whether building activity by the elites and the state (who are primarily responsible for temples and churches) is paralleled by building activity driven by more humble classes.

In any case, despite my critique, it is a very welcome development that we have a new proxy for historical dynamics. We may be at the point where traditional sources, such as historical chronicles, are being mined out, but I expect that quantitative historical data will continue to flow in, thanks to new clever ways of looking at history quantitatively.

An old house in Aarhus, Denmark

Note added 17.XII.2018. Fredrik Ljungqvist clarifies: “The description that ‘Ljungqvist’s group went around Europe taking cores from wooden beams in old houses and then dated them’ does not feel fully accurate as we compiled data already available from decades of archaeological dendro-dating work in Central Europe and reused it for our study. (All the data contributors were included as co-authors.) We did not do any new dating work for this study.”

“To defeat populism, America needs its own Macron–a charismatic leader who can make centrism cool” wrote Max Boot in June 2017, one month after Emmanuel Macron was elected president in France. I always felt that calling Macron a “centrist” is a remarkable misrepresentation of his hard neo-liberal ideology. But the extent of his radical refashioning of France’s social fabric is becoming clear only now in the wake of the Yellow Vest Rebellion, at least for those who don’t follow French politics closely, like myself.

Source

Until recently, France was one of the few major Western countries that resisted the world-wide trend to growing economic inequality. Not anymore. Macron’s government has enacted a comprehensive program of reforms seemingly (or, more likely, intentionally) designed to make the poor poorer, and the rich richer. Pensions have been frozen (but solidarity taxes on pensioners were increased), minimal wages were frozen, but at the same time taxes on the very rich were reduced. Increasing the taxes on gasoline, which triggered the rebellion, hits the poor much more than the rich. It also hits the rural areas more, and by most accounts the majority of Yellow Vests are provincials.

The assault on France’s social state is remarkably comprehensive, and includes measures affecting the health system, education, and transport infrastructure (see this article by Diana Johnstone for more detail). The result of these policies has been graphically plotted in this BBC graph:

There are few more striking illustrations of the Matthew Principle (which says, ICYMI, that the rich will get richer, while the poor will become poorer; more on this in Chapter 10 of my War and Peace and War). Thanks to Macron, in 2018 France gained the dubious distinction of having the second highest number of new millionaires in the world. The first one is, of course, the United States. But proportionally speaking, France beats US in the rate of growth: 14 percent per year (compared to 5 percent in the US)! Macron really earned the distinction of being the president for the rich.

For any student of history, the Yellow Vest Rebellion immediately brings to mind the Yellow Turban Rebellion that ended the corrupt Eastern Han Dynasty in the second century AD:

Source

That peasant rebellion, like most popular uprisings, was eventually suppressed. What will be the ultimate fate of the Yellow Vests? The future will tell.

Over the past weekend France has been rocked by “the worst urban riot in 50 years” (see reporting by Reuters, DW).

Indeed the previous comparable outbreak of political violence in France was in May 1968, almost precisely 50 years ago. (As a reminder, I define “political violence” as internal collective conflict that occupies the middle territory between individual-on-individual violence (crime) and interstate wars.)

Although most historians disagree with the idea that there are cycles in history (at best, it rhymes), our cliodynamic research has identified a number of periodic processes in historical dynamics. And one of them is the 50-year cycle in political violence (see a previous post on this topic). This cycle doesn’t need to be very precise—in historical data the cycle periods can vary anywhere between 40 and 60 years. But sometimes it strikes with eerie precision, like what we see today in Paris.

In the United States we also see this cycle, which resulted in spikes of political violence spaced almost precisely 50 years apart: late 1960s–early 1970s, circa 1920, and in the 1860s–early 1870s (see my book Ages of Discord for details). It is one of the reasons for my prediction that we will experience a peak of political violence in the early 2020s. (But not the only one: even more important are such factors as intraelite conflict and popular immiseration. Also, that prediction was made 10 years ago, and the way political unrest has been developing here hints that we may have this spike arrive “before it is scheduled”).

Returning to the political turmoil in France, if my reading of the situation is correct, we haven’t seen the peak yet. It’s interesting that a Reuters article, published last Spring on the 50-year anniversary of the May 1968 riots, concluded that French mood [is] far from revolutionary despite lingering May ’68 spirit. Strangely enough, the article argued that the current economic malaise affecting France (for example, expressed in high unemployment rate) was an argument against the willingness of the French to protest against the government. In the Structural-Demographic Theory, however, popular immiseration is instead one of the factors driving mass-mobilization potential, and therefore the social pressures for instability.

A very interesting question is what level the structural-demographic pressures for instability have reached in France. Unlike with the US, I haven’t run the numbers, so I can’t answer this question currently (and not for a while, as I will be wrapped up with the analysis of Seshat data and with models based on these data). But that will really determine whether political instability spreads, or dies out.

Last year I had an interesting conversation with someone I’ll call the Washington Insider. She asked me why my structural-demographic model predicted rising instability in the USA, probably peaking with a major outbreak of political violence in the 2020s. I started giving the explanation based on the three main forces: popular immiseration, intra-elite competition, and state fragility. But I didn’t get far because she asked me, what immiseration? What are you talking about? We’ve never lived better than today. Global poverty is declining, child mortality is declining, violence is declining. We have access to the level of technology that is miraculous compared to what previous generations had. Just look at the massive data gathered together by Max Rosen, or read Steven Pinker’s books to be impressed with how good things are.

There are three biases that help sustain this rosy view. First, the focus on global issues. But the decrease of poverty in China (which is what drives declining global poverty, because Chinese population is so huge), or the drop in child mortality in Africa, is irrelevant to the working America. People everywhere compare themselves not to some distant places, but to the standard of living they experienced in their parents home. And the majority of American population sees that in many important ways they are worse off than their parents (as we will see below).

Second, the Washington Insider talks to other members of the 1 percent, and to some in the top 10 percent. The top-income segments of the American population have done fabulously in the last decades, thank you very much.

Third, many economic statistics have to be taken with a grain of salt. Government agencies are often under substantial political pressure to put a positive spin on the statistics they publish. Many economists work hard to please the economic elites and other powers-that-be, because that’s how you get ahead in that profession. Fortunately, there are enough “heterodox” economists who provide us with alternative views. This all doesn’t mean that statistics are worse than “damn lies”; on the contrary, one cannot make sense about where we are headed without statistics. The point here is that one needs to understand why different statistics may give us different answers.

So what has been happening with the well-being of common, non-elite Americans? In my work I use three broad measures of well-being: economic, biological (health), and social.

The most common statistics one sees about economic well-being is the trend in per-capita household incomes. This is not a particularly good way to measure economic well-being for two reasons. First, as households became smaller (because Americans have fewer children), the same wage of the primary breadwinner gets divided by a fewer heads, and that yields an illusion of things getting better. Second, as a result of massive entry of women into the labor force, the typical household today has two bread-winners, compared to a single-wage household of fifty years ago. Furthermore, many households today have even more than two wage-earners, because adult children don’t move away. As a result of both of these factors, the time trajectory of household income yields an overly optimistic view of how well Americans are doing economically.

The best way to see the state of working America is to focus on typical wages of non-elite workers (which means excluding CEOs, high earning corporate lawyers, top athletes, and rock stars ). Here’s what real (adjusted for inflation) wages of two typical non-elite groups tell us:

The pattern is unmistakable: rapid, almost linear growth to the late 1970s, stagnation and decline (especially for unskilled labor) thereafter. Here’s a more detailed breakdown of men’s wages since 1979, broken down by wage percentile (10th is the poorest, 95th is the richest):

Source: State of Working America

Why did this happen? I answer this question in a series of posts, Why Real Wages Stopped Growing (see it in Popular Blogs and Series). The TL;DR answer is that it was a combination of immigration, loss of manufacturing jobs overseas, massive entry of women into the labor force (thus, this factor both inflated household income and, perversely, depressed wages for men), and changing attitudes towards labor. A model incorporating these influences does a pretty decent job of capturing both the turning point of the 1970s and fluctuations afterwards:

Source: Ages of Discord

One potential problem with this statistic, real wages, is that one needs to adjust nominal wages for inflation, and a lot of trickery can happen at that stage. This is a big topic (perhaps for a future post). In my work I have sidestepped this issue by focusing on the relative wage, which is the nominal wage divided by GDP per capita, also expressed in nominal dollars. Here what this statistic tells us about the state of working America for the whole history of the US:

Source: Ages of Discord  Relative wage has been declining since the 1960s.

Another important indicator is availability of jobs. The jobless rate published by government agencies is not a very useful statistic, because it tells us about short-term fluctuations, and excludes people who gave up on the job market. A better measure is the labor participation curve, especially for men:

(source)

Dividing men by educational attainment is a way to check that declining labor participation rate is indeed due to decreasing demand for labor (because high school drop-outs have fewer job prospects than college-educated men). The trend is uniformly down, but it is worst for less-educated men.

An amusing way to spin this bad news was pointed out by one commenter on my previous post. An NBER article by Mark Aguiar and Erik Hurst, “Measuring Trends in Leisure”, optimistically concluded that between 1965 and 2003 “leisure for men increased by 6-8 hours per week” and that “this increase in leisure corresponds to roughly an additional 5 to 10 weeks of vacation per year.” A closer reading of the article, however, shows that this “leisure increase” was driven by a decline in “market work hours”. In other words, all those extra 10 percent of men with higher school or less, who dropped out of the work force since 1970, are simply enjoying their “vacations.”

Economic measures of well-being only tell a part of the story. One of the most shocking, to me, developments was that the immiseration since 1970 has affected biological measures of well-being”. Here’s a graph of trends in average stature (height):

Source: Ages of Discord

Panel (a) shows that average stature of native-born Americans grew rapidly until the 1970s, and then stagnated. A real shocker is that for some segments of the population (Black women) it actually declined in absolute terms. Panel (b) shows that there is a clear relationship between economic and biological measures of well-being (it’s further explained in Ages of Discord).

For another health measure, life expectancy, we have a similar situation. Overall, America is losing ground in relative terms (for example, in comparison to robustly growing life expectancies in Western Europe). For some segments of the population the decrease is in absolute terms. Here’s a particularly revealing look at the data:

The correlation between red counties and those who voted for Trump in 2016 is rather obvious.

There has been a lot of discussion recently of factors that may be responsible for declining health of common Americans. I won’t go into it, for lack of space, but here’s one component:

Source

Finally, a good indicator for social well-being is the proportion of Americans married; or, alternatively, the age at which they marry:

Source: Ages of Discord

There is a long-term increase in the age of marriage driven by modernization (top panel), so we are interested in fluctuations around the trend (bottom panel). During the periods of increasing well-being (for example, between 1900 and 1960), average age of marriage tends to drop. Immiseration causes it to rise. In fact, an increasing proportion of people doesn’t marry at all. Many of them stay with their parents, and their earnings help to inflate household income statistics.

We know from the work of Jonathan Haidt and others that one of the most powerful factors explaining personal well-being is social embeddedness. Having a spouse is one of the most fundamental ways of being embedded. But a variety of other indicators, collected by Robert Putnam, shows that Americans are becoming increasingly less connected (I’ve written about it in another post).

In short: a variety of indicators show that well-being of common American has been declining in the last four decades. The technical term for this in the structural-demographic theory is immiseration.

===================================================

Notes

22.VIII.2017: Added the chart on men’s wages since 1979 broken down by wage percentile.

It will soon be two years since the US presidential elections of 2016, which should have made it clear to everybody that our society is in deep crisis. The technical term in the structural-demographic theory (SDT) for it is a revolutionary situation: when the established elites are still holding the levers of power, but the social pressures for crisis have built up to the point where something has to give.

What I found remarkable as we have lived through the past two years (indeed, the past eight years since I made my prediction of the impending crisis), is how precisely we today are following the trajectory into crisis that my colleagues and I saw in the historical societies we have studied. The explanation, probably, is that the three major mechanisms driving up social pressure for crisis in the SDT work in a mutually reinforcing way. The fundamental drive (a kind of a “pump” that drives up social pressure) is the oversupply of labor, which developed after the 1970s as a result of multiple interacting factors, and more recently was made acute by technological change driving automation and robotization. Oversupply of labor is the root cause for both popular immiseration and elite over-production/intra-elite competition. Both of those factors, then, contribute to the fiscal crisis of the state, because immiserated population can’t pay taxes, while the elites work to reduce the taxes on themselves.

We saw all those mechanisms operating in our current crisis. Immiseration of large swaths of the American population was what fueled the successful campaign of a counter-elite presidential candidate, Donald Trump. Intra-elite conflict has reached unprecedented heights (since the First American Civil War), as the established elites are using various means at their disposal to get rid of the counter-elite chief of state. At the same time, a weird coalition of Trump and the established elites (remember, laws must be approved by the Congress) legislates deep cuts into the taxes the elites will pay, bringing the fiscal crisis of the state much sooner. Political violence has also reached new heights, although thankfully mostly demonstrators and counter-demonstrators are beaten up, not killed (a major exception was Charlottesville a year ago).

Until last year I thought that we collectively have a decent chance of avoiding the crisis, but I now have abandoned this hope. A major reason for my pessimism is the resolute refusal by our ruling class (including its both Liberal and Conservative wings) to see the real causes of the crisis. They are internal, not external. As a result, the mid-term elections will be completely free of (largely mythical) Russian influence, but no attempt is made to address the deep structural-demographic causes. All these pressures continue to increase.

The major question on my mind now, instead, is how we could sail through the crisis without a major amount of bloodshed. This is where the “Ginkgo Model” may serve as a useful conceptual device. As I said earlier in this post, the trajectories of entry into structural-demographic crises are fairly narrowly channelized. But once the crisis breaks out, suddenly a much broader fan of possibilities opens up. It’s just like a Ginkgo leaf:

Some post-crisis trajectories go to a really dire territory: a bloody civil war, a revolution bringing an oppressive regime, or disintegration of the state into a number of territorial sections. Other post-crisis trajectories are less dire. In the best scenario, the elites manage to pull together and implement the reforms needed to defuse the pressures for crisis—reversing trends of immiseration and elite overproduction and restoring the fiscal health of the state.

However, unlike in the Ginkgo leaf analogy, the fan of probabilities of emerging from a crisis is heavily lopsided—and unfortunately in favor of really negative outcomes. Over the years I have studied about thirty cases of historical societies going into crisis, and emerging from it, ranging from Rome and China to France, Russia, and the United States. I scored the crisis severity in each by such parameters as the effect on the population (none, mild decline, catastrophic decrease), on the established elites (from mild downward social mobility to dispossession or even extermination), and on the state (territorial fragmentation, external conquest). Adding together these indicators, here’s the result:

As you see, the more positive outcomes (lower severity on the left side) are fairly rare (about 10% of historical cases), while the majority of outcomes cluster in the middle-high severity territory. In fact, this way of presenting outcomes is somewhat misleading, for the following reason. We know that the scale of collective violence in humans, measured by the number of people killed, is not “normally” distributed. Instead, it follows a power law. As an example from my own work, here’s the distribution of the severity of political violence events in the United States between 1780 and 2010:

The frequency distributions of war severity, including both external, inter-state wars and internal, civil wars have the same shape. What it means in non-mathematical terms is that there is no “typical” scale for outcomes of societal crisis. We cannot say that “on average 10,000 people are killed when a civil war breaks out.” The idea of “average” is misleading. A civil war can kill 100, 1000, 10 000, 100 000, 1 000 000 people and even more. The probability of a really severe conflict (e.g., more than 1 mln people killed) is fairly low, but it is much higher than what a naïve person would estimate. This point is admirably discussed by Nassim Nicholas Taleb in The Black Swan and other writings.

What it means for us here in the United States is that the severity of the troubles to come in the next few years to a decade is really impossible to predict. It could be as mild as the late 1960s–early 1970s, with the violent urban riots and a fairly ineffectual terrorist campaign by the Weather Underground. Or it could be as bad as the First American Civil War. Once again, a real catastrophic collapse of our society may not be highly probable, but it is much more probable than we think.

For me the greatest eye-opener in The Rainforest: The Secret to Building the Next Silicon Valley by Victor Hwang and Greg Horowitt was a realization that starting your own innovative company is a deeply irrational decision. Considering the amount of effort (and often your own money) that you are going to invest in it and the very low probability of success, it simply doesn’t make sense to start a company—as long as you measure success in dollars. As the authors write at one point, “the process of launching a startup company has many similarities to riding a roller-coaster. It is a highly irrational act.”

There is useful diagram on p. 224 of The Rainforest explaining the components that have to work in order for the venture to become success:

• Technology works
• Product works
• Customers buy
• Manufacturing works
• Business model works
• Enough capital
• Team can handle this whole thing

There are seven components of success in this scheme, and all of them must work in order for “this whole thing” to succeed. Even if we assign an unrealistically high chance of success—50%—to each step, the probability of all of the seven tosses coming heads up is less than 1%. A more realistic chance of success for each step, like 20%, yields an overall probability of around 0.001 percent. If you want to become a billionaire, there are much more certain routes to wealth—for example, working up the corporate ladder in a well-established large company or running a hedge fund. Or, as Thomas Piketty suggests, inheriting the wealth.

And I am not even talking about the “psychic costs” involved in starting a successful business. It can take many years before you find out whether you succeed, and all that time one must live with the very real possibility of failure, that all that hard work will be, in the end, wasted. At least, if you fail at becoming a CEO of a huge corporation, you will be still very handsomely remunerated along the way and can look forward to a very comfortable nest egg when you retire. Another huge psychic cost is having to ask people for money, and then almost always being turned down. As Hwang and Horowitt write, “successful innovation requires self-sacrifice.”

We have the examples of success in front of our eyes, the Bill Gateses and Steve Jobses, but for each of them there are thousands of entrepreneurs who ruined their lives by trying to launch a firm. Even the brilliant Nikola Tesla died in poverty, weighed down by debt (at the end of his life, he lived in a series of New York hotels, leaving behind unpaid bills).

Curiously enough, I have a great personal understanding of the costs of entrepreneurship, because launching Seshat: Global History Databank was in all respects, except one, an identical experience to an investor launching a successful business. The one difference, of course, is that the success of Seshat is measured not in dollars, but in the number of publications in top journals, the number of citations by other researchers, the theories that we either confirm or reject, and our increased understanding of how human societies evolve and function.

So, what does actually motivate entrepreneurs? Some of them are in it just for the money, and most of those fail. Successful ones are typically motivated by extrarational reasons: “thrill of competition, human altruism, a thirst for adventure, a joy of discovery and creativity, a concern for future generations, and a desire for meaning in one’s life.”

Even more eye-opening was the new understanding of venture capitalists that I gained after reading The Rainforest. I used to refer to them as “vulture capitalists”, but I now understand that it was unfair, especially when talking about successful VCs. Because successful VCs, as The Rainforest explains, also cannot be motivated by greed alone. The problem is, again, the high probability of failure. It is simply not rational to invest in a startup at the very early stage. Usually VCs will know that the first hurdle (“technology works”) has been successfully passed. But the remaining six hurdles still need to be surmounted. One way of increasing the payoff in case the venture succeeds is to force the entrepreneur to yield a large percentage of equity in the company. But that’s self-defeating, because that destroys the motivation of the entrepreneur to work hard.

At a later stage, once most of the hurdles have been cleared, it makes perfect sense to invest in a startup, but how do they get to that stage? It takes either investors behaving extrarationally, being motivated by the same values cited above, or collective action channeled by governments—in fact, it takes all of that.

Most venture funds that help entrepreneurs during the early stages are funded by governments who hire VCs to run them, or by VCs who venture their own money in parallel with a government fund. I had some experience of this recently, when one of the Seshat directors, Kevin Feeney, launched a new company, Data Chemist (read about it in The Irish Times).

I end my review of Hwang and Horowitt’s book with this quote “Rainforests depend on people not behaving like rational actors.” It requires extrarational motivations of all the key players, including the inventors and investors, and also governments. I am not saying that all business is like this. From what I read in the press, the world of financial organizations, large corporations, and corporate law seems to be driven largely, or entirely by greed. Branko Milanovic is right in that. But not all business is like that. Innovation is really the key to why we live better than people did one hundred, or one thousand years ago. And that business requires extrarational motivations, self-sacrifice, and cooperation.

One of the chief reasons I became an advocate of the Cultural Multilevel Selection (CMLS) theory is that it wonderfully clarifies the relationship between competition and cooperation.

• Competition between groups (up to whole societies) fosters within-group cooperation.
• Competition within groups (between their members) destroys cooperation.

If you wanted a two-sentence summary of Ultrasociety this would be it.

I’ve been applying these principles in my research on the evolution of complex societies (and a lot more is to come, now that the Seshat Databank have come of age and is generating a tremendous volume of empirical results—which you will see in a year or two, as academic publication mill-stones grind sooo slowly).

But I have also felt that the CMLS theory tells us a lot not only about political organizations like the states, but also about economic organizations—firms competing in markets. This goes very much against the grain, especially as far as economists are concerned. Milton Friedman, of course, always argued that economic agents should strictly follow their material interests; there is no place for “extra-rational motives” in business. Most economists today feel the same way, although few are willing to state it as boldly as Friedman did.

Thus, it was very refreshing to receive two years ago an email from Branko Milanovic, an economist whom I greatly admire, in which he was willing to go on record and state this position very forcefully (I published Branko’s letter as a guest blog on Cliodynamica). I also invited two economist friends to comment: Bob Frank and Herb Gintis, as well as writing a response myself. The whole exchange was recently re-published by Evonomics and generated a lot of discussion.

More recently Branko wrote a review of Kate Raworth’s new book Doughnut Economics. A central question in the ensuing debate between him and Kate (see it here) is the same we debated two years ago (see also a good summary on Vulgar Economics). Here’s what Branko wrote about the “human nature under hyper-commercialized global capitalism”:

Here I respectfully decline to be moved by the results of any of the “games” that Kate cites and that are supposed to reveal human nature. These games are indeed games; they are not the way people behave in real life. Games are good in generating publishable papers but they tell us nothing about how the same people would (or do) behave in real life.

Two years ago I wouldn’t know how to respond to Branko on this point, but fortunately recently I finished reading a remarkable book, which provides me all the ammunition I need.

Before discussing it, let’s define the question more explicitly. I agree that most people much of the time, and some people all of the time, are motivated by very “rational” calculations. When I decide which supermarket I am going to go for food, I try to minimize the amount of money I will pay and the amount of time I will spend, while maximizing produce quality and selection. It’s a very straightforward optimization problem.

But capitalism is not just about buying and selling things—people have been doing commerce for millennia before capitalism. Surely the amazing capacity of capitalism to transform knowledge into innovation, and innovation into economic growth is one of the central of its attributes? So let’s talk about such successful innovation hotspots, as the Silicon Valley. What are the motivations driving successful entrepreneurs within such hotspots?

If you want to find out the answers to these questions, read The Rainforest: The Secret to Building the Next Silicon Valley by Victor Hwang and Greg Horowitt. Now, Hwang and Horowitt are not scientists, and their book is not a rigorous scientific study. But they spent decades bringing together venture capitalists and entrepreneurs, both in their native California and all over the globe—“Japan, Taiwan, Scandinavia, and New Zealand, … Mexico, Egypt, Kazakhstan, Colombia, Saudi Arabia, and the Palestinian Territories.” Given their enormous experience, the empirical base, with which they operate in the book, transcends the dismissive academic characterization as “anecdotal.”

A central theme that recurs throughout the book is that successful entrepreneurs, and the successful innovation systems in which they operate, such as the Silicon Valley or Route 128 in Massachusetts, are the antithesis of the rational businessperson postulated by Branko, one who is solely motivated by money. In fact, “Rainforests [their term for successful innovation systems] depend on people not behaving like rational actors.” “For Rainforests to be sustainable, greed must be restrained.” “Predatory venture capitalists might win a few in the short run, but they do not last long in the business and are unable to build lasting firms.”

The evolutionary logic of entrepreneurship, according to Hwang and Horowitt, is precisely the opposite of that posited by Branko. Predatory, super-competitive individuals and firms are eliminated by natural selection, and only cooperative ones survive. They write:

Extra-rational motivations—those that transcend the classical divide between rational and irrational—are not normally considered critical drivers of economic value-creation. …  These motivations include the thrill of competition, human altruism, a thirst for adventure, a joy of discovery and creativity, a concern for future generations, and a desire for meaning in one’s life, among many others. Our work over the years has led us to conclude that these types of motivations are not just “nice to have.” They are, in fact, “must have” building blocks of the Rainforest.

Many successful entrepreneurs (think Steve Jobs or Elon Musk) clearly were not motivated solely by money. Naturally, they did not give their billions away, and for some other very successful innovators, perhaps, all they wanted was to become filthy rich. But the point that Branko makes, that capitalism “is a system really built on the best use of our vices, including greed” is clearly wrong.

This is why when governments and corporations try to incentivize innovation by focusing on financial mechanisms, the overall result is failure. By the end of the book, Hwang and Horowitt boil down their own recommendations as to what makes successful “Rainforests” thrive. They are four.

First, diversity, which brings people with very different knowledge and skills together, such as a scientist, a venture capitalist, an engineer, a sales specialist, and an administrator (a CEO).

Second, extra-rational motivations, because self-regarding rational actors are simply unable to cooperate to launch a successful innovation enterprise.

Third, social trust, because successful cooperation is the only way to beat the terrible odds against a successful innovation startup, and cooperation requires trust.

Fourth, a set of social norms that regulate the behavior of various cooperating agents, and willingness both to follow them and to enforce these rules by various sanctions.

In other words, Hwang and Horowitt describe a system that uses precisely the same components to bring about cooperation that have been studied in other settings (a foraging group, a military troop, a religious sect, and a state), and in the abstract, by cultural evolutionists.

The Rainforest, then, provides ample empirical material to reject the theory that economic growth, which is based on innovation, is moved by self-interested rational agents. But—and it was one of the real eye-openers for me—it also explains why this is so. I discuss this in Part II.