A prime focus for social scientists, and in particular political scientists, is on institutions. Institutions are stabilized sets of expectations that establish frameworks for social action that affect behavior because they affect calculations and inspire attachments. Institutions do change, but they change slower than life changes. This creates a paradoxical reality. On the one hand, the relative stability of institutions—the rules and procedures they establish for interaction and decision–compared to the fluctuations of circumstances and preferences is what makes it possible for human groups to take effective action. On the other hand, their very stability means that the decisions they enable are almost inevitably suboptimal.
Accordingly, although most political scientists are committed to a general view that the interests and beliefs of human beings and human groups are the primary drivers of political behavior and political change, a good deal of attention by “institutionalists” is directed to relishing the ironies or bemoaning the tragedies of rationality ignored and interests contradicted. Indeed you do not need a political scientist to point out numerous examples of institutional forms or collective beliefs or norms that are severely suboptimal for precisely those populations and communities that uphold and honor them. Political scientists, as well as pundits, are well aware of the obstacles sclerotic institutions pose to good policy, progress, and a general sense that our political communities work for us rather than against us. References abound to “institutional inertia,” “the stickiness of institutions,” or the institutionalization of answers to questions that current circumstances no longer pose.
However, if it is well understood that institutions cannot change fluidly with changing needs and changing insights, it is also known that institutions do change, and sometimes they adapt. What is not well understood are the limits to the effectiveness and pace of institutional change and, specifically, why some institutions are exceedingly resistant to change, even when the deficiencies of the practices, policies, and predicaments associated with them are fully appreciated by influentials as well as ordinary people. In this essay I want to suggest the contribution that evolutionary theory, as a tool in the hands of trained social scientists, can make to finding answers to these questions.
In part because political scientists are so aware of how bad things are compared to what they theoretically could be, they commonly reject evolutionary theory and thinking as inappropriate for application to the worlds they seek to explain. This rejection is usually based on a fundamentally incorrect understanding of evolution as “survival of the fittest”–short-hand for the (incorrect) idea that Darwinian evolution produces the “best” version of what could be out of a ferocious competition among versions that can be. Let us not linger over the reasons for this error. There are many candidates for explanations including past abuses of evolutionary theory by Social Darwinists and some sociobiologists, cultural or psychological fears, and religious commitments. One oft-neglected explanation is perhaps the easy identification of such a Panglossian understanding of evolution with the principle of neoclassical economics that free, unconstrained competition in a market place can yield a Pareto optimal set of prices for guiding the most efficient use of resources possible.
In any case, this notion—that “history is efficient”—is certainly not an implication of evolutionary theory, not in the natural world and not in the social world either. Indeed it is the very effectiveness of evolutionary theory in accounting for suboptimality that offers political scientists and social scientists more generally, an approach to explaining the prevalence, not only of institutional suboptimality, but to the combination of adaptation and extreme resistance to change that institutions display.
A common trope among historically minded political scientists is that severely suboptimal outcomes are the product of the inheritance from the past of an institutional form or policy implemented to serve a particular purpose under particular circumstances.[i] The simplest form of this explanation is that the outcome is fully explained by “path dependence” and “inertia;” that is to say by the contingency of what happened in the past and by the difficulty of changing the status quo, even if the status quo no longer reflects the purposes or circumstances that resulted in the contingent outcome that was deposited in the present by the past.
Of course many political scientists offer very nuanced and learned “process-tracings” of these outcomes and the ironies and tragedies associated with them. But for more systematic and generalizable explanation for the regularity of this type of outcome, we can profitably turn to evolutionary theory. To appreciate how, we must understand the standard concept in evolutionary theory of a “fitness landscape.”
From an evolutionary point of view, the particular history of a polity or society is one path through the “state-space” of possible ways that society could have changed over time, from one combination of characteristics to another. Since each of these distinctive combinations of characteristics is a separate point, or “state” in the space of possible states, the trajectory of the society over time is a path through the state space achieved by movement from one state to another “accessible” to it. The idea of “accessibility” here reflects whatever laws are governing the behavior of relevant entities so that once relevant elements are configured in a particular way, a subset of possible successor states exists that includes the state that actually materializes. As a social or political institution or practice changes over time, slowly or rapidly, it can be imagined to be exploring a particular, and, in all likelihood, relatively small, portion of the state space it inhabits. Evolutionary mechanisms are ways of explaining patterns of movement through the state space that do not rely on calculated strategic choices at the level of the entity moving through the space. Instead, these mechanisms rely on the outcome of competition for replicative success among large populations of variants at levels of analysis below the ontological level of the entity moving through the state space. Another way of saying this, indicating the links between evolutionary and complexity theories, is that the path through the state space is an emergent property of unguided evolutionary processes at a lower level of analysis.
Thus, in evolutionary biology, the unit of selection reflects a particular level of analysis. The effects of evolution are traced on levels higher than that of the unit of selection. Sometimes the unit of selection is the codon, sometimes the gene, sometimes the trait which might be expressed by a combination of genes; sometimes a species and sometimes varieties within a species. Whenever evolutionary, or at least natural selection, questions are asked at one of these levels, we identify variation, a selection criteria arising from circumstances/competition, and retention ability across generations of replication. The result of processes of variation, selection, and retention at any particular level is a pattern of outcomes at a higher (emergent) level. For example, giant tortoises in the Galapagos Islands vary, island by island, by the shape of their shells near their necks and by the length of their necks. This pattern of differences at the level of tortoise anatomy resulted from competition, at a lower level of analysis, among different rates of reproductive success for individuals on different islands with traits (slightly longer necks and slightly notched shells) that marginally advantaged the individuals possessing those traits wherever food was relatively higher. The result was that the state space for the configuration of neck and shell on giant tortoises was explored along different routes on different islands.
A “fitness landscape” is a heuristic device for analyzing the opportunities and challenges of changing for the “better” (by whatever metric is imagined as selecting or pressuring behavior). For a simple example, picture a three dimensional grid comprised of columns rising from a checkerboard along the Y (vertical) axis. (See Figure 1) The squares on the checkerboard represent all the different ways the entity could behave or be constituted so as to combine every available value on the X (horizontal) axis (Attribute 1, e.g. neck length) with everyone on the Z (depth) axis (Attribute 2 e.g. shell shape). In other words, the checkerboard is the space of possible states the entity can “be.” Since evolution seeks explanations for patterns of change in response to circumstances without endowing the units of selection with the capacity to look far ahead for their success, the only states that can be achieved by an institution, policy, or practice are those adjacent to it. The columns rising from the squares on the grid register their relative “fitness” by their different heights. If change is incremental and myopically driven toward whichever adjacent (i.e. only very marginally different) column is higher, then “hill-climbing” will be normal. The trajectory of an entity through the state space will be upward.
However, and this is the key to understanding why history is “inefficient” from an evolutionary point of view, ravines and chasms may exist in the landscape, or may come into existence as changes in circumstances degrade the effectiveness of strategies (ways of being; traits; call them what you will) that worked in the past while enhancing strategies that performed (or would have performed) poorly in the past. In such a rugged landscape, many better ways of being are not “accessible” without going “downhill,” or becoming less “fit” in the process of becoming more fit. But if the processes producing change are truly myopic, or only responsive to direct local stimuli and information, then they themselves—i.e. evolution, per se, will never produce downhill trajectories. Absent non-evolutionary, or at least non-natural selection, processes of change, a stabilized set of practices (i.e. an institution) can evolve to, or become stuck on, one of many suboptimal “local maxima” that may exist in the state space.
In our checkerboard illustration, this occurs if an entity is surrounded by columns that are lower than the column it occupies. Without look-ahead powers that evolution does not, per se, assume, that entity will not be able to improve its “fitness” by relocating to a higher, but distant and therefore non-accessible location. To do so would require some extraordinary event that exogenously relocates the entity to a position from which it could hill-climb to the highest peak available in the state space. This is of course not impossible, but it would run against the grain of normal interactions. In any event, the predicament outlined here proves the point that history’s inefficiencies can be modeled evolutionarily with no contradiction whatsoever to the fundamentals of evolutionary theory, i.e. to the claim that unguided change in the deployment or appearance of strategies drawn from a repertoire of those available arises from the immediate successes and failures that determine rates of replication of alternatives. In the world of institutions this means that memos identifying pathologies and offering plans for institutional change that entail high short-term costs in favor of long-term gains will tend to be out-replicated, and effectively suppressed, by memos warning of immediate pain, discounting the future heavily, and distorting the benefits of staying the course.
Let us look more closely at how these ideas could be used to do some work in the political world. We can do this by recognizing that political institutions, political practices, or the campaign strategies of politicians are entities that can be moved through state spaces. The trajectories can be understood as the product of evolution to the extent that they result from competition at lower levels among varieties of organizational forms, rhetorical appeals, political positioning, slogans, and formulas–these are the equivalent of the genes, codons, traits, or varieties that compete to drive the trajectory of evolution in biological contexts, depending on the level of analysis of the entity whose position in the state space is under examination.
Students of the political economy of advanced industrial societies are familiar with the “Japanese model.” This refers to the distinctive and enormously successful combination of institutional forms and attendant policies that produced the stupendous success of the Japanese economy in the 1980s. We need not go into detail here, but the Japanese economic miracle is widely understood to be due to effective planning and implementation by a developmentalist state intent on harnessing its resources for efficient production and prevailing against international competitors.[ii] Key ingredients in the “Japanese model,” imitated to one extent or another by a number of East Asian and Southeast Asian countries in the 1970s and 1980s, included a deferential political culture; well-trained, well-coordinated, and political protected bureaucracy; a dense patron-client network with massive corporations working closely with banks and the national bureaucratic apparatus toward goals of growth of the economy; job security for the middle class; a disciplined working class; massive subsidies for agriculture; and the effective exclusion of women from the upper levels of the work force. Trusting in the wisdom of the state, and willingly following its directives, Japanese firms fully cooperated in tax, trade, and monetary policies to override market forces and endow Japan with tremendous competitive advantages in the global market place. Between 1960 and 1980 annual per capita growth grew from 36% of the American rate to 72%.
But in the early 1990s the real estate market tanked. Banks incurred immense losses on their books in order to prevent large firms from failing. This was imperative based on the corporatist organization of the economy and the concomitant absence of a welfare state to cushion unemployment. The political system churned away, producing government after government, but no reforms that could help Japan confront the huge challenges from its own ageing work force and from increasingly severe global competition. In her closely argued and extensively researched study of what happened to the Japanese economic miracle, Jennifer Amyx identifies the gradual crystallization of the Japanese model, but asks why it seemed incapable of grappling with the challenges posed in the late 1980s and early 1990s. Amyx credits Japan with a gradualist, evolutionary path to the economic model to which it owed its immense success, but then notes the sudden reversal of Japanese economic fortunes and the seeming inability of Japan to adjust in a timely manner to patently new circumstances. The result was a national debt more than twice as large as the OECD average by 2008 and an economy, once the envy of the world, now entering its third decade of stagnation. In her 2004 study, Amyx focuses her questions on the crucial finance sector of the Japanese economy, noting the difficulty existing theoretical approaches to institutional adaptation had in explaining Japan’s predicament.
Existing theories cannot explain the length of delayed government response to banking problems, the magnitude of breakdown seen in Japanese finance, or why Japanese authorities are unable to restore the financial sector to health even thirteen years after the onset of the crisis.[iii]
As Amyx makes clear, the problem was not that Japanese experts did not, and do not, understand the problem or were/are not committed to change or reversing Japan’s economic performance. The problem is that the kind of transformation required threatened and has threatened even more pain for the Japanese people and for the government, in the short run, than would be experienced (in the short run) by continued stagnation. Evolution toward optimal policies arising from the kind of gradualism that has characterized Japanese institutional adaptation in the past has not occurred, and will not occur. In essence, Japan has been stuck on the “local maximum” it had achieved in the 1980s in a fitness landscape that changed to reduce the relative fitness of its historically evolved strategy, but whose ruggedness has prevented normal processes of institutional adaptation to replace the “less fit” strategy with a “more fit,” but not immediately accessible alternative. While it is likely, if not certain, that Japan will eventually change its policies, when and how that happens will likely have little to do with myopic, evolutionary processes of natural selection per se, but will depend heavily on shifts in circumstances that remove ruggedness from the fitness landscape, on far-sighted leadership prepared to whether high political costs, and/or the force majeure associated with devastating shocks.
As noted, the problem of institutional rigidity is well-known and pervasive, and it would be wrong to characterize all instances as the result of being stranded on a local maximum. Determined, albeit mistaken, policies of powerful elites; the effects of vested and well-positioned interests or veto-players; the existence of determinative but latent or obscured functions; or the suppression of ideas about how change might be achieved are examples of alternative explanation for the failure of institutions to adapt. Nonetheless, as suggested above, the effects of being stranded on a local maximum is a powerful explanation and valuable tool for thinking about the predicaments faced by political systems and how they might be escaped.
In the 1950s it was clear to most Frenchmen that the Fourth Republic was a severely dysfunctional institution, but it was governed by parties and leaders that had evolved to perform well within its institutions and no matter how many opportunities they were given to adapt the regime to the requirements of the political system, they could not. It was the genius of de Gaulle to realize this, to withdraw from “le systeme,” and then use the regime-threatening problem of Algeria to replace it wholesale with the “Fifth Republic.” There was no gradual, evolutionary path from the Fourth to the Fifth Republic, but there was a revolutionary path. It is widely understood, in Israel and outside the country, not only that the country desperately needs a peace agreement with the Palestinians. But it is just as well understood that Israeli political institutions, however effective they are at maintaining democracy and producing opportunities for political office and patronage to those in power, have insured that every Israeli government has been coalition-based and reliant on small religious and highly ideological right-wing groups that prevent any realistic peace plan from being put forward. Gaullist solutions in Israel have been attempted—by Rabin and, to an extent, by Sharon–but so far Israeli elites have failed to weather the storm of political opposition associated with efforts to “deinstitutionalize” deeply embedded arrangements.[iv] In the United States, the Madisonian system described in Federalist 10, that prevents tyranny by dividing and balancing power among states, Houses of Congress, and branches of the Federal government, has also institutionalized a kind of gridlock in so many domains that the confidence of the American people in its government is falling to record lows. Just as Washington may well be understood as stuck on a local maximum—fit enough to allow incumbents to be re-elected, but not fit enough to solve the problems posed to it in the twenty-first century while enabling re-election–so may we understand the predicament of the Republican Party in this election cycle. As has been widely observed, any candidate wishing to win the Republican nomination may be forced to position himself or herself in such a way as to attract Tea Party support; thereby greatly complicating if not rendering impossible the rapid adaptation that will be necessary to achieve a position on the rugged “electability” fitness landscape near the position that wins by attracting independents and conservative Democrats.
Identifying a syndrome in politics with an evolutionary dynamic does not itself solve any problems. On the other hand, our understanding of evolution in fields as far removed from one another as psychology, botany, agronomy, pest control, pharmaceutical research, and cancer treatment, has helped enormously to explore state spaces for solutions and improvements that were not imagined beforehand. By understanding key predicaments in political life with the same conceptual, theoretical, and analytic equipment used to solve problems in evolutionary theory, we can begin to see the quandaries we face more clearly and imagine more systematically possible opportunities to escape or overcome them.
[i] For a detailed assessment of “historical institutionalism” from the standpoint of evolutionary theory, see Ian S. Lustick, “Taking Evolution Seriously: Historical Institutionalism and Evolutionary Theory,” Polity (2011) pp. 1-31. http://www.polisci.upenn.edu/faculty/faculty-articles&papers/Lustick_Polity_2011.pdf
[ii] Chalmers Johnson (1982). MITI and the Japanese Miracle. Stanford, CA: Stanford University Press; Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization, Oxford University Press, 1989.
[iii] Jennifer Amyx, Japan’s Financial Crisis: Institutional Rigidity and Reluctant Change (Princeton: Princeton University Press, 2004) pp. 17-18.
[iv] Ian S. Lustick, Unsettled States, Disputed Lands: Britain and Ireland, France and Algeria, Israel and the West Bank/Gaza (Ithaca: Cornell University Press, 1993)
Daniel Nettle, University of Newcastle, email@example.com
The individual and the greater good: A comment on Lustick
Many of us biologically-minded folk have been appealing for years for social scientists to take evolution seriously (Nettle 2009). Thus, it is very gratifying to read Lustick’s thoughtful analysis of how institutional political scientists could employ evolutionary concepts. There are good and bad ways of bring evolution into social science (and you can find plenty of examples of both in recent literature). Bad ways have two characteristics: the evolution they appeal to is a highly simplified and sometimes wrongly characterized version of the nuanced edifice of evolutionary biology; and in their enthusiasm to embrace their Darwinian idea, they are less rigorous than they could be in the deployment of methods and knowledge base of the social science discipline which they purport to be expanding. The good ways, happily, are represented by Lustick’s work: it grows from a deep grounding in political science itself, and employs a sophisticated evolutionary metaphor in which both adaptation and history are important.
Lustick’s essay is based on the idea of institutional forms as replicating themselves through time, giving rise to a process of institutional descent with modification which happens more slowly than the change in the environment of people’s lives. Although institutional adaptation does occur, the fitness landscape has a complex shape, such that institutions can become trapped on local maxima. This analysis potentially has a lot to recommend it. However, I was surprised be did not invoke another important principle in evolutionary thinking, namely conflicts of interest between the individual and the collective as a source of suboptimality in design. The idea of the social dilemma – the tragedy of the commons – is already much discussed in political science, and indeed evolutionary biology got much of its thinking in this arena from social scientists (e.g. Ostrom 1990).
In evolutionary genetics, for example, there are many aspects of genomes which perpetuate themselves despite having no functionality for the organism as a whole (Burt and Trivers 2006). Transposable genetic elements make many copies of themselves despite not serving any function at the organismal level. Segregation distorters bias meiosis in their favour, and can spread despite being costly to the health of the animal carrying them. The replicatory interests of the segregation-distorting allele and the rest of the genome of the organism carrying it are partly, but not perfectly, aligned. The consequence is that things which are inefficient at the organismal level often persist.
How might we apply these ideas at the institutional level? The long-term efficiency of government might, hypothethically, be served by reducing the size of the political class. However, any individual leader introducing this reform reduces the size of his alliance base, and consequently risks personal loss of power. All politicians might agree that this is the reform needed, but it is not good for the career of any of them. Thus, it is a very hard reform to introduce. The same is true by definition of any policy which makes governments unpopular over the timescale of the electoral cycle, even if it would be good for society in the very long term. These social traps essential arise because the interests of society and of individuals, whether politicians or not, are partly but not perfectly aligned.
Something I found interesting in Lustick’s analysis is the potential role of rare massive upheaval in overcoming these traps. If society occasionally becomes completely destabilized, then no individual has any possibility of doing well personally by continuing with the status quo, then an adaptive radiation of new and better institutional forms is possible. I can only hope that, in this era of financial crisis and looming environmental problems, our political masters understand this.
Burt, A. and R. Trivers (2006). Genes in Conflict: The Biology of Selfish Genetic Elements. Cambridge, MA: Belknap Press. Nettle, D. (2009). Beyond nature versus culture: Cultural variation as an evolved characteristic. Journal of the Royal Anthropological Institute 15: 223-40.
Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press.
Comment on Ian Lustick’s “Institutional Rigidity and Evolutionary Theory: Trapped on a Local Maximum”
David Sloan Wilson
President of Evolution Institute and SUNY Distinguished Professor of Biology and Anthropology, Binghamton University
Thanks to Ian Lustick for his stimulating essay. He correctly notes that stasis and change are two sides of the same evolutionary coin, for biological evolution no less than the cultural evolution of human political institutions. Sewall Wright, one of the fathers of population genetics theory, was the first to appreciate that when phenotypic traits have a complex genetic basis, natural selection can result in multiple stable local equilibria (the “peaks” of a multi-peak landscape), which are internally stable by definition but can differ in their absolute fitness (the “altitude” of each peak). His shifting balance theory was a complex scenario involving selection among populations occupying different adaptive peaks (Provine 1986). He originally developed the theory for individual traits with a complex genetic basis (such as coat color in guinea pigs), but it applies equally to social adaptations with a complex basis, where it is called “equilibrium selection” (Binmore and Samuleson 1997, Boyd and Richerson 1992, Samuelson 1997)
Applying these and other evolutionary ideas to stasis and change in political institutions is even more complex than Lustick suggests. I would like to introduce three additional factors and conclude with a reflection on how to manage the study of highly complex systems in both biology and the human-related sciences.
1) Multilevel selection theory needs to be distinguished from evolution on multi-peaked landscapes. The classic group selection model posits two traits, selfish and altruistic, in a multi-group population. The altruistic trait is selectively disadvantageous in all groups containing both types; there is no local equilibrium favoring altruism. Nevertheless, altruism can evolve in the total population if the differential fitness of groups containing the most altruists outweighs the selective disadvantage of altruism within groups. A political institution can be dysfunctional, not because it is trapped on a small peak, but because individuals or subgroups are maximizing their relative advantage within the institution, at the expense of the institution as a whole and even their own long-term welfare (Wilson 2004).
2) The term “evolutionary mismatch” refers to adaptations to one environment that become dysfunctional in a changed environment. Our adaptations for evaluating and copying behaviors evolved in the context of small-scale social interactions and can easily malfunction in the context of large-scale political institutions, resulting in the paradox of practices that work but don’t spread and spread but don’t work. This class of institutional dysfunction needs to be distinguished from both multiple adaptive peaks and multilevel selection (Wilson et al. 2011)
3) Rational thought and intentional planning might seem to be furthest removed from evolution, especially with respect to escaping local maxima. However, these are better regarded as evolutionary processes in which both variation (efforts to imagine alternatives) and selection (explicitly stated goals) are highly organized. Moreover, our genetically evolved reasoning abilities might be better adapted to winning arguments than solving collective action problems (Mercier and Hugo 2011), accounting for some of our reasoning inabilities in addition to our abilities.
This degree of complexity might seem daunting, but it is not evolutionary theory that makes the topic complex. It is inherent in the subject matter and must be faced by anyone who studies political institutions from any perspective. The question is whether an explicitly evolutionary perspective adds value to other perspectives. Along with Lustick, I think that the answer is emphatically “yes”–for the cultural evolution of political institutions on rugged adapted landscapes and the additional factors that I have briefly identified in this commentary.
A manuscript titled “Evolution as a General Theoretical Framework for Economics and Public Policy” (Wilson and Gowdy 2011) makes some of these points at greater length and is equally relevant to the field of political science.
Binmore, K., & Samuelson, L. (1997). Muddling throught: Noisy equilibrium selection. Journal of Economic Theory, 74, 235-265.
Boyd, R., & Richerson, P. J. (1992). Punishment allows the evolution of cooperation (or anything else) in sizable groups. Ethology and Sociobiology, 13, 171-195.
Mercier, H., & Sperber, D. (2011). Why do humans reason? Arguments for an argumentative theory. Behavioral and Brain Sciences, 34, 57-111.
Provine, W. B. (1986). Sewall Wright and Evolutionary Biology. Chicago: University of Chicago Press.
Samuelson, L. (1997). Evolutionary games and equilibrium selection. Cambridge, MA: MIT Press.
Wilson, D. S. (2004). The New Fable of the Bees. In K. R. (Ed.), Advances in Austrian Economics (Vol. 9, pp. 201-220). Greenwich CN: JAI Press.
Wilson, D. S., & Gowdy, J. (2011). Evolution as a General Theoretical Framework for Economics and Public Policy. First article of special issue planned for the Journal of Economic Behavior and Organization.
Wilson, D. S., Hayes, S. C., Biglan, A., & Embry, D. (2011). Evolving the Future: Toward a Science of Intentional Change. Behavioral and Brain Sciences, submitted.
Evolution predicts suboptimality, but not only because of getting stuck on local peaks
Hanna Kokko – Evolution, Ecology & Genetics, Research School of Biology, ANU College of Medicine, Biology & Environment. Hanna.Kokko@anu.edu.au
That institutions can become trapped on local maxima (with respect to some measure of performance) does not come as a surprise to any evolutionary biologist. I am probably not alone in my field when welcoming a move towards interdisciplinary understanding of complex systems, political and social science included. In addition to ‘food for thought’ provided by essays such as that by I.S. Lustick, researchers working in this area would also greatly benefit from moving towards quantitative rigour in this area (e.g. Turchin 2006), a step that biologists took long ago (for a historical review see Kingsland 1995).
There is one aspect in Lustick’s otherwise thought-provoking essay that warrants comment: it remains silent on the fact that even if an evolutionary process found itself on a slope towards highest peak, subsequently reaching it, the solution thus found can be maladaptive compared with a hypothetically reachable peak that could be found if entities forming the group (e.g. individuals in a society) forgot about their selfish short-term interests and pulled together to reach a common goal. This problem is one of the levels of selection, and it provides a clear counterexample for those who believe in invisible hands that create the best possible society simply by letting competition run free.
Lustick mentions that the unit of selection can be a codon, or a gene, or perhaps an entire species. Biologists agree, but only when the statement comes with a reminder that processes acting on the lower end of this spectrum often override selection on the higher units. This is why reaching outcomes that are good for the group is difficult, except — arguably — in the special case of multicellular individuals, where specialized adaptations exist to make sure that individual cell lineages do not take over and start proliferating at the expense of how well the entire organism functions and survives.
After all, it’d be terrible if some cells in your liver, or brain, started dividing haphazardly, without the law and order that states they have to serve the greater good of the multicellular society (i.e. you). Unimaginable? In fact, I have just described cancer, which quite commonly causes death in senescing animals (though not in plants, which do not have animal-like circulatory systems that would allow rogue cells to spread around the entire body). Despite cancer-killing cells and other similar adaptations shown by tightly regulated multicellular creatures like you and me, the system may, to its detriment, sometimes fail to extinguish the selfish, short-term interests of individual cells (Lewis et al. 2008). Here, it is irrelevant that cells do not really have aspirations in any cognitive sense. Short-term proliferation of cells can be selected for in the simple sense that more cells of the cancerous type is by definition a short-term reproductive improvement in this lineage. Given a short enough timeframe, this remains true despite extinction looming in the near future: cancer often kills, and the death of the organism kills the cancer.
We tend to think of cancer as a medical rather than an evolutionary problem. The beauty of evolutionary theory, however, is that it provides a framework for thinking about the commonalities of problems occurring at every level of selection. Fishermen unable to resist the temptation to overfish this year, despite fish stocks depleting to levels that threatens the entire industry? A virulent pathogen spreading in the local population of schoolchildren, even though this means that soon everyone will be immune, and the virus has nowhere to go? These are all examples of the tragedy of the commons, which means that evolution often favours short-term success over more prudent behaviour, even if the latter meant better performance as a whole in the fitness landscape (Rankin et al. 2007).
In fact, human institutions could be viewed as adaptations that try to keep some level of order intact at a higher level of selection that is vulnerable to the invasion of short-term interests. Societies fund police forces that have been given the power to punish thieves, for essentially the same reason as our bodies produce cancer-killing cells: to prevent detrimental selfishness from spreading. It is also the same reason why we spend money on negotiations over quotas for dwindling stocks of cod in the sea — or over the right to dump CO2 in the atmosphere. The struggle between interests of different entities, each of which takes a shorter term view than would be ideal, is the root of much what is problematic in the world. Add to that the types of myopia that Lustick mentions — the inability of an evolutionary process to be farsighted enough to jump to distant peaks — and the power of evolutionary theory to predict suboptimal design should be clear to anyone.
Kingsland, S.E. 1995. Modeling nature. The University of Chicago Press, Chicago.
Lewis, Z., Price, T.A.R. & Wedell, N. 2008. Sperm competition, immunity, selfish genes and cancer. Cellular and Molecular Sciences 65:3241-3254.
Lim, M., Metzler, R. & Bar-Yam, Y. 2007. Global pattern formation and ethnic/cultural violence. Scence 317:1540-1544.
Rankin, D.J., Bargum, K. & Kokko, H. 2007. The tragedy of the commons in evolutionary biology. Trends Ecol. Evol. 22:643-651.
Turchin P. 2006. War and Peace and War: The Life Cycles of Imperial Nations. Pi Press
Commentary on Ian Lustick’s “Institutional Rigidity and Evolutionary Theory: Trapped on a Local Maximum”
Bradley A. Thayer
Professor Department of Political Science
Ian Lustick has produced an important argument by thoughtfully applying evolutionary ideas to a major problem in the study of institutions. Lustick demonstrates how an evolutionary approach may explain why institutions resist change even when their faults and limitations are well understood, while at the same time remain adaptive. To advance his argument, he draws on the central concept of evolutionary theory: natural selection. Evolution through natural selection operates through variation within a population, a selection criterion or criteria arising from competition, and replication.
From this foundation, Lustick artfully creates a “fitness landscape” for illuminating why institutional change is so difficult—in sum, entities are trapped on a local maximum, plans for change entail short term cost for long term gain, and will lose out to those discounting the future. When these ideas are applied to political problems, they can explain major phenomenon like the rigidity of political systems in important and novel ways.
In this brief commentary, I evaluate Lustick’s use of evolutionary ideas and conclude that Lustick’s approach is a model of how evolutionary ideas and theory may be applied to social science.
Lustick’s use of evolutionary ideas is well done. Here I offer two points to place the discussion in greater context. A critical point in the operation of evolution through natural selection is the selection pressure on variation within a population. Selection pressure is broader than Lustick’s treatment, and should be thought of as competition for resources among conspecifics, for example, institutions and states, but also with other species, predators, and changing immediate, seasonal, and long-term ecological conditions (the impact of the international system, if you will). The last point underscores the consideration of time. Benign ecological conditions, for example, abundance of resources or few predators reduce selection pressure and may do so over a considerable period of time. The reverse is likely to accelerate change over a relatively short period. Recognizing the broader scope of selection pressures will only assist the study of institutions.
Second, evolution lacks teleology. With that in mind, it is important to stress a major point: “better” and “best” are relative, and are so for given ecological conditions. As conditions change, what was the “best” for a specific condition may be fatal for the species in new conditions. Better to be “good enough” and adaptable to changing ecological conditions than the “best” and inflexible. Here evolution agrees with Voltaire’s quip that the perfect is the enemy of the good enough. As I read Lustick’s consideration of “fitness” in his presentation of his “fitness landscape,” he is implicitly sensitive to this point, which I would encourage to be drawn out in the future.
Finally, Lustick deserves great credit for advancing the goals of consilient social science, the use of insights from the life sciences to inform, improve, and augment our understanding of major social problems. Sven Steinmo’s recent examination of political economies of Sweden, Japan, and the United States joins Lustick as another excellent example of the use of a consilient approach to aid our comprehension of institutional change. As other scholars join Lustick and Steinmo in a consilient approach, the knowledge of institutions, and politics and, more broadly, social science may advance.
 Sven Steinmo, The Evolution of Modern States: Sweden, Japan, and the United States (New York: Cambridge University Press, 2010).
Commentary on Ian Lustick’s “Institutional Rigidity and Evolutionary Theory: Trapped on a Local Maximum”
Department of Biological Science
Binghamton University University
I would like to thank both Ian Lustick and David Sloan Wilson for their essays on a subject so closely related to my own work and for the invitation to participate in this conversation. Lustick does an excellent job presenting the case for why institutional inertia is such an important problem and how it can be approached from an evolutionary perspective. Wilson nicely brings in some valuable distinctions that may at first glance seem to complicate the problem but that ultimately make it more tractable. I would like first to offer a few cautionary notes and then briefly to situate the conversation within the context of more recent theoretical work that is part of a new “extended synthesis” of evolutionary thought(1).
First, there is something terribly seductive about adaptive landscapes. When plotted out on tidy, three dimensional graphs, the differences between local and absolute maxima become salient, plausible and compelling as an explanation for why systems can become so recalcitrant to change. However, it is rare for fitness to depend so neatly on combinations of two discreet traits. More often, fitness derives from combinations of large numbers of traits, leading to high-dimensional fitness landscapes that are impossible to imagine and difficult to model. What do peaks and valleys look like in, say, 10 dimensions? Are populations that find themselves on local maxima in such bogglingly complex fitness landscapes trapped in the same ways they would be in a three dimensional landscape? Those who have undertaken such mind bending mathematical challenges have found that our intuitions about movement in three dimensions tend to fall apart in more complex spaces(2). This is not to say there is no utility in framing the problem of institutional inertia in terms of fitness landscapes. Indeed, it can be quite useful in certain contexts. However, it is an oversimplification and, as I hope to show later, there are important theoretical reasons for keeping the more complicated picture in mind.
Second, I would caution that the notion that institutions are “inevitably suboptimal” is a bit problematic. From an evolutionary perspective, it is not always “optimal” to respond to immediate conditions. Suppose, for instance, that you have a population of individuals that have different tolerances for cold and heat. If temperatures steadily increase over many generations, then individuals with higher tolerance for heat should have a selective advantage and the population should increase in its proportion of heat tolerant individuals. Similarly, if temperatures steadily decrease, the proportion of cold tolerant individuals should increase. This is the familiar case of directional selection. However, what if temperatures are highly variable? What is advantageous for one generation in such a scenario becomes maladaptive for the next. The most adaptive response under such conditions might be to become relatively insensitive to changes in temperature(3). In one sense this kind of response is, indeed, “inevitably suboptimal” to the immediate environmental conditions, but systems under natural selection often must take a longer view of things. This may hold particularly true for institutions which, as Lustick points out, often outlast single generations.
Finally, there seems to be an inherent hubris in discussions of institutional change. Namely, there is an assumption that proposed solutions, rationally derived, are inherently better than what has come before, even when what has come before is acknowledged to have been vetted through a process of natural selection. This may often be the case, but history is likely as rife with examples of the unintended, disastrous consequences of rationally implemented institutional change as it is with the ironies and tragedies of “rationality ignored.” A favorite example comes from the green revolution that overtook Bali in the 1970s(4). This was an attempt to bring Balinese agriculture under the rationality of scientific agronomy. Farmers were encouraged to abandon their superstitious planting and watering schedules, which for centuries had been under control of an elaborate system of water temples known as subaks. Incentives were instituted to encourage farmers to plant new hybrid rice varieties, to use industrial fertilizers and pesticides, and to plant as many crops per year as they could manage. Unfortunately, none of the advocates for rational farming practices appreciated the exquisitely sophisticated ways the Bali subaks had evolved to sustainably use water, to limit nutrient leaching and to control pests. After a few years of impressive increases in crop yields, the new system began to unravel: water shortages became common, pests overtook fields, soils became less fertile and productivity rapidly declined. Institutions are often highly complex systems embedded within larger environmental and social contexts. Any particular institutional trait may interact with multiple parts of the system and even the most sophisticated rational designs for institutional change can only take a handful of the operational variables into account. For instance, a new policy that leads to increased efficiency in one part of the organization may lead unintentionally to decreased efficiencies elsewhere within the same group. As long as selection is operating at the level of the organization and not its independent parts, evolutionary theory allows us to predict that the new policy will not be favored.
All that said, institutions frequently do fall into maladaptive patterns. Sometimes it really is necessary to get “there” from “here” and some proposed solutions to institutional problems are objectively better than others. Lustick’s call for approaching institutional change from an evolutionary perspective is well placed and well timed. The study of evolutionary forces that resist change is hardly new. Sewall Wright, for instance, first introduced the idea of fitness landscapes in 1932. However, the extended evolutionary synthesis places these ideas on firmer theoretical footing under the headings of robustness and evolvability.
Robustness refers to the ability of a system to persist or function in the face of perturbation. It is a familiar term across a wide variety of disciplines. As such, I prefer it to Wilson’s “cultural stasis,” which has some rather unfortunate connotations in disciplines such as anthropology and the humanities, disciplines that must be engaged in a positive dialog if we are to reach a better understanding of how robustness plays out in cultural systems. Evolvability, by contrast, refers to the ability of a system to respond to natural selection. At first glance, these seem to be antagonistic concepts. Robustness is generally seen to be the antithesis of change while evolvability is seen as a desirable quality for a system to be able to adapt. As it turns out, however, the two are intimately interconnected concepts(5). Understanding how they are related may yield new insights into the problems of institutional change and how it can be better managed.
It is relatively easy to understand how evolved, adaptive systems might become robust. To begin, natural selection requires some degree of stability in a system in order to have an influence. Otherwise, favorable traits would not be reliably passed on to future generations. Similarly, highly fragile systems tend not to survive variable environments or the inevitable variations that creep into systems as they are propagated across generations. Becoming resilient to such perturbations can be highly adaptive. But can increased robustness ever lead to increased evolvability? The answer seems to be yes. To understand why, it is important to realize that is very rare for single changes in evolved systems to be adaptive. In fragile systems, single changes most often lead to decreased fitness. In robust systems, by contrast, their effects on fitness are most often neutral. The implication is that robust systems allow for the accumulation of neutral mutations and this brings us back to the question of how systems move through high-dimensional landscapes. Robustness creates areas of neutral space, thus allowing systems a degree of creativity they might not otherwise enjoy. A single change to a robust system might have very little consequence to fitness, but some combination of changes may allow the system to hit upon novel solutions that are either more adaptive in the current environment or that allow the system to occupy novel niches in which it was previously excluded. In other words, the neutral spaces afforded by robustness can allow systems to move from one adaptive peak to another even in otherwise rugged adaptive landscapes. Again, this is why our intuitions about movement through three dimensional space can lead us astray.
Work on robustness and evolvability in biology is still in its infancy and I have found very few attempts to apply these dual concepts to cultural systems. However, I believe this approach holds great promise. Instead of treating robustness as something inherently problematic that needs to be overcome, even this preliminary sketch suggests that robustness might be employed to search for novel solutions to institutional problems and that, wisely managed, could actually facilitate institutional change.
1 Pigliucci, M. An Extended Synthesis for Evolutionary Biology. Annals of the New York Academy of Sciences 1168, 218-228 (2009).
2 Gavrilets, S. in Evolution: The Extended Synthesis eds Massimo Pigliucci & Gerd B. Müller) Ch. 3, 45-79 (The MIT Press, 2010).
3 Kawecki, T. J. The evolution of genetic canalization under fluctuating selection. Evolution 54, 1-12 (2000).
4 Lansing, J. S. Priests and Programmers: Technologies of Power in the Engineered Landscape of Bali. 2 edn, (Princeton University Press, 2007).
5 Wagner, A. Robustness and Evolvability in Living Systems. (Princeton University Press, 2005).
Evolution and Social Science: Toward a Real Conversation
Ian S. Lustick
University of Pennsylvania
The replies and comments posted in response to my essay on institutional rigidity and evolutionary theory are heartening. This is a promising start to the Social Evolution Forum’s effort to encourage fruitful dialogue and mutual learning across a divide that has too long separated the life sciences from the social sciences, in particular in matters related to evolution. It is no coincidence that all the respondents to my essay emphasize what social scientists have to learn from natural scientists. That is, in large measure, the message of my posting. Animated by the wonders, subtleties, and power of evolutionary theory, and aware of the challenges, predicaments, and inadequacies of social science, we each hail and encourage efforts to mobilize substantive evolutionary theory to do work in the worlds of social scientists.
My response here is designed to encourage the conversations that will make this possible by emphasizing that while social scientists must make a serious and sustained effort to become literate in evolutionary theory, so too will evolutionary biologists need to become more sophisticated about social science and its accomplishments. Otherwise social scientists, hearing well-intentioned but naïve suggestions from evolutionists, will take the easy path, turning away from a conversation based on a quick impression that “I already knew that. Nothing new here.”
I was delighted that Bradley Thayer cited Sven Steinmo’s important book on the political economies of Sweden, the United States, and Japan as a promising example of a prominent social scientist turning toward evolutionary theory. Indeed I was recently a participant in a panel at the annual meeting of the American Political Science Association entirely devoted to that book. The panel featured vigorous discussion of the value of bringing evolutionary thinking into contact with political science and of different strategies for doing so. One of the main points I made in that discussion is that Sven’s strategy—to use evolutionary vocabulary, but not theory, to depict the distinctive trajectories of three political economies as idiosyncratic phenotypes—was insufficiently ambitious. To advance this project we need to mobilize evolutionary propositions and explanations, not just evolutionary vocabularies. Evolution must do work, and be seen to do work, in the social sciences—work that could not be done without it.
As I read the four responses to my posted essay, I am convinced that colleagues in evolutionary biology and related fields feel similarly and are anxious to contribute to that effort in ways that social scientists will be able to appreciate. Thayer emphasizes the absence of teleology in evolution, successful replicators are “better” or “best” only as replicators, not according to some metric of progress toward a destined endpoint or on a scale of values of any sort. Indeed my decision to use “stranded on a local maximum” to analyze a particular problem in political economy was precisely intended to help disabuse social scientists of the misconception—only too prevalent—that evolution cannot explain sub-optimality. The falsity of his idea is obvious, of course, to any evolutionary theorist, but not to many social scientists. In fact, based on the misconception that evolution explains why the “fittest” survive, and knowing full well that optimal outcomes are almost never observed in the social world, social scientists are naturally expect the disutility of evolutionary thinking for their kinds of problems.
David Sloan Wilson’s comment also focuses on the question of explaining sub-optimality, while also noting evolutionary mechanisms to escape from it. Thus his multi-level selection theory can help explain why selection against a trait of value to a population at the individual level could still lead to its successful replication (because of disproportionately successful replication of groups that feature individuals with that trait). David’s comment about rationality as “better adapted to winning arguments than solving collective action problems” also highlights one of the most influential political science theory of sub-optimality, namely the “prisoner’s dilemma.” The combination of rationality and uncertainty under common kinds of incentive structures leads not only to maximum jail terms in the famous game theoretic fable, and various tragedies of the commons. It is also the source for powerful explanations of wars and arms races that are much more destructive or expensive than the interests at stake would warrant, and of the widely observed failure of governments to produce nearly the amount of public goods that would be “rational” for their societies.
Interestingly, the key focus of Daniel Nettle’s comment is also on the variety of ways that evolutionary theory can be mobilized to explain sub-optimal outcomes. The notion, familiar to Nettle from biology, that traits replicating successfully at lower levels can be dysfunctional to the organism, is rightly identified as isomorphic to the tragedy of the commons. It is also, not so incidentally, the basis for the crystallization of the state and the “social contract theory” of the state in the philosophy of Hobbes, Locke, and Rousseau. Few political scientists have thought that the problems they find familiar in their domain correspond quite closely to patterns observed at various analytic levels in evolutionary biology. But I also suspect that few evolutionary biologists are aware, or are presently equipped to appreciate, the variety of solutions political scientists have come up with to what is familiar to political scientists as the ‘collective action problem”—solutions that help explain why, despite the disincentives for rational actors associated with contributions to group goals, so much of it occurs. My point is that what we need is a conversation across disciplinary boundaries. Not only can social scientists benefit from learning evolutionary theory, but it is highly likely that evolutionary biologists can learn from the social science literatures on sub-optimality—literatures that explain why, under some circumstances, outcomes are not as sub-optimal as one might have expected. Among these solutions, for example, are exploitation of the strong by the weak, cultivation of small groups, centralized enforcement mechanisms, selective benefits, ideology, and indoctrination or programming.
I would make the same point in response to Hanna Kokko’s comment that evolutionary processes often feature sub-optimal outcomes at higher levels because selection processes at lower levels “override” processes of selection “on higher units.” Kokko rightly identifies this pattern as corresponding to a critique of the invisible hand as always likely to produce, as a result of bottom-up processes of competition, Pareto optimal outcomes at the macro level. Of course in both political science and economics there is a vast literature on “market failure” and the perverse incentives that arise depending on how competition is structured by institutions and the rules that comprise them.
In sum, I am grateful and appreciative of these four generous comments and of the prospect for a real and professional dialogue between social scientists and evolutionary biologists. But I would also emphasize that the curiosity that social scientists must cultivate about evolution will need to be matched by the curiosity of natural scientists about the world, perhaps equally mysterious to them, of sophisticated social science. It would be a pity to make Thomas Kuhn’s error—to reject social science as so much unscientific gobbledygook, even while using, or misusing, an underspecified social scientific theory of “revolutions” to analyze the progress and process of science.