Here are a few thematic features cross-posted to Tumbld Thoughts. You will discover the theme as you read -- it will "emerge", shall we say. But I'm not promising deep causality.
I. Cycles of Social Events With Little Causality?
Here are some random readings on cliodynamics and why my blogging endeavors exhibit little causality. The first set of articles [1, 2] focuses on blog mining, particularly when blog posts on a given topic yields subtle causality. In [2], the Rapport Corpus was used to compile thousands of (qualitative) accounts of the same event. These data were then statistically mined to find causal mechanisms among the convergent threads. This is (in theory) similar to the mining of lung cancer data for potential (and oftentimes false-positive) causal patterns.
Another way to establish historical causality among what are often highly-qualitative and contextually-contingent accounts of observed events is to use cliodynamics [3]. Cliodynamics uses a chartist approach, which is similar to Forex trading strategies [4]. This might be useful for finding cycles of violence in historical data. However, Jason Collins [5] offers a critical analysis of Turchin's approach. Notably, Turchin [4] boils most of history down to two uniform cycles: secular (in which societies cycle from egalitarian to elitist to egalitarian in 200-300 years) and father-and-son (where social injustices are found and addressed in 60-80 year cycles). However, this does not account for large-scales changes (so-called Black Swan events) nor other complex historical contingencies.
II. Or is there more causality than suspected?
Here are two perspectives on the Nobel Syndrome: does winning a Nobel cause brilliant minds to start investigating weird things, or does it happen all on its own? In the first feature by Bradley Voytek at Oscillatory Thoughts [6], we are introduced to the prodigy effect, where young investigators win Nobels (or similar such prizes), and then go on to investigate pseudo-scientific phenomenon later in their career [7].
But does the proverbial cart (Nobel) always come before the horse (oddball research topics)? That's where the second article (by Barry Ritholz at The Big Picture blog) comes into play: is Eugene Fama (this year's winner of the Economics Nobel) an example of someone who engaged in oddball behavior before winning the prize [8]? Ritholz thinks so, and explains why that may be a pre-emptive case of Nobel syndrome.
III. Or perhaps hierarchical network effects?
Here are a few blog posts/articles on human organization, cities, and economic payoff. The first is an intellectual excursion from Dizzynomics blog [9] on the phenomenon of buying housing as investment income in central London. This has lead to massive increases in housing prices which has displaced former residents to less desirable areas. The consequence of this strategy might be to create a ghost city (a city with no permanent residents) or, worse yet when the bubble bursts, a dead city. But what happens when a few cities (such as central London) serve as critical access points for the global economy? This trend, replicated across other cities in the global urban network, may provide a subtle causal mechanism for significant income inequality. This outsized effect (in terms of scope) of a real estate arms race on overall economic opportunity is discussed in a post [10] from Moneybox blog.
But why is it, in the age of easy global travel and internet connectivity, that opportunity found in the critical access point cities has not decentralized to a large number of urban centers? The answer to this is partly due to the inherent relationship between a given city's creative performance and its population size. This has been articulated by Geoffrey West and others [11]. According to this idea, the largest cities should be the most economically (and creatively) productive. This scaling relationship can occasionally be violated, but such exceptions are directly dependent on the evolution of the city in question.
But perhaps the nature of extreme concentration (or strict hierarchical organization) of places that are true engines of economic wealth creation have as much to do with the network topology that connects players in the global economy than the inherent properties of those players. In an network analysis of Twitter messages involving two grass-roots political organizations (the Tea Party and the Occupy), different network topologies might lead to different outcomes and sets of constraints on its function [12]. Perhaps the selective nature of a hyper-efficient, free-market global economy naturally leads to hyper-centralization and limited economic flexibility.
IV. Yes, Probably, with a Chance of Giant Components
It's hard to influence the "Giant Component". To make this point, here is an interesting book review: Robin Hanson of Overcoming Bias reviewing David Graeber's book "Debt: the first 5000 years" [13]. It is interesting not because of any particular insight or its length, but because here we have dyed-in-the-wool market capitalist reads book by a self-avowed left-wing Anarchist. And surprising because, overall, Hansen actually liked the book. I have also read "Debt", and understand Hansen's skepticism. However, there are two tacit assumptions to this dynamic that need to be understood:
1) Graeber is in an interesting position because while he is an Economic Anthropologist, he is also an activist. Therefore, the scholarship and plans for action don't always match up (as they should not). But then again, why does theory (or in this case, comparative historiography) need to be a catalyst for social change? Just because a particular theory fails to do so does not diminish what theories are actually for (e.g. explaining and predicting) [14]. And whether a particular set of theoretical assumptions actually does this is not a matter of a lack of activism.
2) The implicit goal of economics is to understand how resources are efficiently allocated. In fact, the definition of the suffix "-nomics" (or even "-omics") means "natural law", but is often used as a stand-in for quantification and optimization [15]. In fact, a goal of economics is to understand human exchange through the lens of optimal outcomes (whether or not they actually are optimal). By contrast, alternative approaches such as Economic Anthropology do not make this assumption. Such alternative approaches There is a "economics as natural law" vs. "economics as human agency" [16] dichotomy surfacing here that subtly influences much of the debate on post-crisis economics.
Now, one of Hanson's criticisms is that Graeber is inherently "anti-debt". And while Hanson does not explain why debt is a good thing (other than totally ignoring the phenomenon of predatory debt), Graeber does discuss how debt is part of a system of social and moral obligations. In this sense, debt enables a social order. However, these social orders can be unstable (due to natural disasters or wars), and it has been quite common throughout history to discharge debts. This is where Hanson has the most trouble with Graeber's position: what would happen to the world economy if debts were simply discharged? Would this not be cataclysmic? And who pays the price when debt holders are not just elites, but pension funds and endowments as well?
But this brings up a larger question: how does large-scale cultural change happen during the flow of life, and how does it happen without social collapse or (more immediately) a fundamental disruption to social life? We can view this in the context of social networks -- more importantly in terms of Renyi's Giant Component [17]. Social networks (in this case, the global economy) exhibit connectivity as a function of human exchange. In the case of modern economic social networks, a common system of finance permeates every part of the topology. This is why the financial crisis of 2008 had such an "giant" effect: the freezing of credit systems essentially had the effect of neutralizing connections throughout the network.
Setting everything back to zero can be quite destructive. Or creatively destructive.....
The giant component, or at least one interpretation, results from a phase transition in the network structure that results in a large, unified topological component. This giant component, once it emerges, is stable. But it may also be unevolvable (e.g. serves as a cultural constraint) and perhaps even makes the entire network brittle with respect to large-scale changes [18]. It is because of this giant component that large-scale social change provides as much a risk as an opportunity: simply suspending or changing a policy or arrangement that has lead to a giant component (or of similar scale) has the potential to completely dissolve the network.
NOTES:
[1] Blog mining. Economist, March 11 (2010).
[2] Tomai, E., Thapa, L., Gordon, A.S., and Kang, S-H. Causality in Hundreds of Narratives of the Same Events. Proceedings of the AAAI (2011).
[3] Turchin, P. Arise 'cliodynamics'. Nature, 454, 34-35 (2008).
[4] For some perspectives on Turchin's work, please see:
a) Pigliucci, M. Cliodynamics, a science of history? Rationally Speaking blog, August 4 (2008).
b) Finley, K. Mathematicians Predict the Future With Data From the Past. Wired Enterprise, April 10 (2013).
[5] Collins, J. Cliodynamics and complexity. Evolving Economics blog, August 6 (2012).
[6] Voytek, B. The Prodigy Effect. Oscillatory Thoughts, June 8 (2013).
[7] Orac Luc Montagnier: the Nobel disease strikes again. Respectful Insolance blog, November 23 (2010). Also, here is a Quora conversation in the topic.
[8] Ritholz, B. Fama has Shiller to thank for his Nobel Prize. Big Picture blog, October 20 (2013).
[9] Kaminska, I. Property bubbles and ghost cities. Dizzynomics, October 9 (2013) AND Goldfarb, M. London's Great Exodus. October 12 (2013).
[10] Yglesias, M. America's fast-growing cities aren't prospering. Moneybox blog, September 30 (2013).
[11] Robinson, R. Can cities break Geoffrey West’s laws of urban scaling? The Urban Technologist blog, July 23 (2013).
Bettencourt, L.M., Lobo, J., Strumsky, D., and West, G.B. Urban scaling and its deviations: revealing the structure of wealth, innovation and crime across cities. PLoS One, 5(11), e13541 (2010).
[12] Whitty, J. Tweet Forensics: occupy vs. tea party. Mother Jones, November 17 (2011).
[13] Hanson, R. Graeber's Debt book. Overcoming Bias blog, October 6 (2013).
[14] Johnson, T. How economics suffers from de-politicised mathematics. Magic, Maths, and Money blog, September 21 (2013).
[15] One man's quest to make "omics" all about his life (and biology): Dennis, C. The rise of the narciss-ome. Nature News, March 16 (2012).
[16] The performativity hypothesis, summarized in the aptly-named book: MacKensie, D. An engine, not a camera: how financial models shape markets. MIT Press (2008).
[17] Erdos-Renyi model: Erdos, P. and Renyi, A. On the evolution of random graphs. Publications of the Mathematical Institute of the Hungarian Academy of Sciences, 5, 17–61 (1960).
Hayes, B. The birth of the giant component. bit-player blog, November 20 (2009).
[18] Jones, J.H. Nearly Neutral Networks and Holey Adaptive Landscapes. Monkey's Uncle blog, December 29 (2008).
No comments:
Post a Comment