Old world city systems and economic networks 950-1950 how the growth and decline of cities and the rise and fall of city-size hierarchies is related to the network structure of intercity connections Doug White UC Irvine November 20 2009 talk Social Dynamics and Complexity ASU economic networks and city systems: using physics models & measures with large samples, time series, & inferential statistics Physical measures 1 Entropy – minimum energy configuration given constraints/processes 2 LDC: Long-distance correlations 3 q-scale entropy is (1) with (2) and power law tails when q > 1 4 q-scale network – degree distribution parameter for a network where size is a LDC q-scale attractor to & between hubs 5 Time-lag cross-correlations for city size distribution q-scale parameters. P(X ≥ x) ~ (1-(1-q)x/κ)1/(q-1) (1 < q ≤ 2) As x max and P 0 the tail of this distribution converges to a log-log powerlaw slope -ß 1/(1-q), so P ~ (1- ßx/κ)-ß examples 1.Species, species populations, & energy in habitat areas (J Harte) 2.Cities depend on trading partners 3.City size distributions, power law tails 1 < q<2 4.q-scale social circles simulation model of complex networks (White,Tsallis, Kejzar,Farmer,White 2006) 5.q-scales of cities in city-system regions show temporal time lags from 0 (synchrony) to hundreds of years As q → 1, q-scale entropy converges to Boltzmann-Gibbs entropy Probability distribution q-fits for a person being in a city in the region with at least population x (fitted by MLE) Smooth lines are fitted curves in successive time periods, jagged lines bootstrap point distributions used to estimate error bounds Each distribution is for all the cities of a region, e.g., China, in one of the 8 time periods, at 50 year intervals from Chandler 1987 1.0 Cum prob P(X ≥ x) on a log scale Cum prob P(X ≥ x) on a log scale 1.0 .1 .01 .001 The mle Pareto Type II q-scale. Measures the shape of the body of the curve, while beta10 measure fits the log-log slope of the tails, which vary independently of q. x = City size log of 10 thousand .1 .01 .001 Goodness of fit for q and beta10 are found by bootstrap probability simulation, with iterations added around each of four of the 8 periods 0001 1 million Shalizi (2007) right graphs=variant q-fits City size log of 10 thousand city systems in the last millennium 1 million Whole period 900 – 1950 Credits: White, Tambayong, Kejzar 2008 Are there inter-region synchronies? Time-lag cross-correlations give lag 0 = perfect synchrony lag 1 = state of region A predicts that of B 50 years later lag 2 = state of region A predicts that of B 100 years later lag 3 = state of region A predicts that of B 150 years later, etc. The relation of q-scales in region “MiddleEast&Afghan&India” to Chinese cities is (1) synchronously inverse but with (2) 100-150 year lags affects them positively (2) (1) Moving to inter-Asian regions on the Silk Road, excluding India: Time-lagged cross-correlation effects of Mid-Asian q-scale on China q-scale mle_MidAsia with mle_China Coefficient Upper Confidence Limit 0.9 Lower Confidence Limit 0.6 CCF 0.3 These and the other crosscorrelations hold on average for the 1000 year time period. 0.0 -0.3 -0.6 MiddleEast&Afghan Robust Cities affect Robust Chinese Cities with 50 year lag -0.9 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 Lag Number (1=50 year lagged effect, 2=1 year lag, etc.) For endpoints further away on the Silk Roads: Time-lagged cross-correlation effects of China q-scale on Europe q-scale mle_China with mle_Europe (100 year lagged effect) Coefficient Upper Confidence Limit 0.9 Lower Confidence Limit 0.6 CCF 0.3 0.0 -0.3 -0.6 Chinese cities q-scale affect European cities q-scale with 100 and 300 year lags -0.9 -7 -6 -5 -4 -3 -2 -1 0 1 Lag Number 2 3 4 5 6 7 Time-lagged cross-correlation effects of the Silk Road trade on Europe’s beta (beta is the slope of the power-law tail of the urban distribution) (50 year lagged effect) logSilkRoad with EurBeta10 Coefficient Upper Confidence Limit 0.9 Lower Confidence Limit 0.6 mle_MidAsia with mle_Europe Coe 0.3 Upp Lim Low Lim 0.6 0.0 0.3 -0.3 CCF CCF 0.9 -0.6 0.0 -0.3 Chinese Silk road trade affects Elite tails of European Cities with a 50 year lag -0.9 -7 -6 -5 -4 -3 -2 -1 0 1 Lag Number 2 3 4 5 6 -0.6 -0.9 7 Mideast cities q have a small effect on European cities q with 150 year lag -7 -6 -5 -4 -3 -2 -1 0 1 Lag Number 2 3 4 5 6 7 mle_Europe with ParisPercent Coefficient Upper Confidence Limit 0.9 Lower Confidence Limit 0.6 CCF 0.3 0.0 -0.3 -0.6 -0.9 European cities q-scale synchrony with % of France population living in Paris, with 100 year decay -7 -6 -5 -4 -3 -2 -1 0 1 Lag Number 2 3 4 5 6 7 Variations in q and the power-law slope β for 900-1970 in 50 year intervals Credits: White, Tambayong, Kejzar, Tsallis, 2006, 2008 3.0 China Mid-Asia Europe MLEqExtrap 2.5 Beta10 MinQ_Beta 2.0 1.5 1.0 0.5 0.0 911111111111111111111111191111111111111111111111119111111111111111111111111 001122334455566778888999900112233445556677888899990011223344555667788889999 0 0 0 5 0 5 0 590 1 50 05 02 01 0 5101510 1 50 05 25 70 50 5 7101510 1 50 25 1 5171015 1 10121517 1 151710 1 1 5101517 9 101510 1 1 7101215 1 1 0101510 1 15101510 1 1 2951710 1 171 000000000005000005050500 000000000005000005050500 000000000005000005050500 1111111111111111111 001122334455566778888999900112233445556677888899990011223344555667788889999 000505050505705050257025700050505050570505025702570005050505057050502570257 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0 5 0 5 0 5date 00 000000000005000005050500 000000000005000005050500 date city systems in the last millennium Are these random walks or historical Periods? Runs Test Results Runs Tests at medians across all three regions Test Value(a) Cases < Test Value Cases >= Test Value Total Cases Number of Runs Z Asymp. Sig. (2-tailed) MLE-q 1.51 35 36 71 20 -3.944 .0001 Beta10 1.79 36 37 73 22 -3.653 .0003 Min(q/1.5, Beta/2) .88 35 38 73 22 -3.645 .0003 Runs Test for temporal variations of q in the three regions mle_Europe mle_MidAsia mle_China Test Value(a) 1.43 1.45 1.59 Cases < Test Value 9 11 10 Cases >= Test Value 9 11 12 Total Cases 18 22 22 Number of Runs 4 7 7 Z -2.673 -1.966 -1.943 Asymp. Sig. (2-tailed) .008 .049 .052 a Median city systems in the last millennium Is there Eurasian synchrony or are there time-lagged effects? Some synchrony when dependent variable closely related in same region but conservative (Euro beta, Paris %) Mostly time-lagged effects of leading regions, i.e., directional not symmetric, with lagging regions, consistent with Modelski & Thompson 1996 w Devezas 2008, i.e., globalizing econ./pol. leaders Next: we test whether, if economic competition is increased by multiconnectivity (structural cohesion): The dynamics of trade is influenced by the trading network: whether monopolized by chokepoints or competitive. Leaders-to-lagger economic effects follow the dynamics of trade. 0900 AD From first stirrings of globalization to the 21st Century Credits: White, Tambayong, SFI Europe Central Asia Medit. Silk routes China Near East Q (scaling sizes) Changan India Bagdad & Changan (Xi’an) In these slides I will connect the city network & city size distributions and power-law tails connected to q-exponential scaling of city sizes low q with thin power law tails of global hubs CORRELATES with global network links 1000 AD N~3 960: Song capital at Kaifeng, invention of national markets, credit mechanisms diffuse Silk routes Global network links characterize low q (power law tail for city sizes) 1100 AD Silk Routes Global network links characterize low q (more exponential body with power law tail for city sizes) 1150 AD 1127: No. Song capital of Kaifeng conquered, Song move to south, capital at Hangchow Silk Routes diminish Global network links characterize low q (more exponential body with power law tail for city sizes) 1200 AD Song capital at Hangchow Golden Horde silk routes Silk Routes diminish Global network links characterize low q 1250 AD cutnodes edgecut Broken network links lead change to high q – led by China, 50 years 1300 AD 1279: Mongols conquer Song Kublai Khan Mongol trade Broken network links characterize high q (here: tenuous interregional connectors) 1350 AD Mongols refocus on Yuan administration of China Silk routes unimportant Broken network links characterize high q (here: tenuous interregional connectors) 1400 AD 1368 Ming retake China Silk routes unimportant Renewed network links characterize low q (power law tail) 1450 AD World population growth turns super-exponential 1421 Ming move capital to Peking Silk routes unimportant Renewed network links characterize low q (power law) – high q led by China, 100 years 1500 AD Renewed network links characterize low q (power law tail) – but China high q leads change 1550 AD Broken network links characterize high q 1600 AD Renewed network links will lead change to low q (here: tenuous interregional connectors) Britain India British/East India: circumferences of the trading circles are small, sufficient by 1720 and 1760 to induce fully competitive market pricing Network cohesion plus close regional distances COMPANY ROUTES in 1620 evolve thru malfeasance by ship captains to independt market price capitalism from 1720 Erikson, Emily and Peter S. Bearman. 2006 Malfeasance and the Foundations for Global Trade: The Structure of English Trade in the East Indies, 1601–1833 American Journal of Sociology (2006) 112(1):195-230. Fig. 3 1650 AD Renewed network links characterize low q (power law tail) – China crash synchronized 1650 AD Renewed network links characterize low q (power law tail) – China crash synchronized 1700 AD Broken network links return to high q – esp. for China leading 1750 AD Broken network links typify high q – China leading – bifurcated world 1800 AD Circum-European cities start to overtake China in number Broken network links typify high q – bifurcated world 1825 AD European cities overtake China in number and size Industrial revolution British opium trade from India Broken network links typify high q – trifurcated world – best example of high local navigability 1850 AD British benefit from peace trea (here: tenuous interregional connectors) Broken network links typify high q – trifurcated world – but China developing power-law tail 1875 AD British benefit as opium legalized (here: tenuous interregional connectors) Broken network links typify high q – bifurcated China power-law tail thinning toward low-q 1900 AD British benefit opium legal Broken network links typify high q – trifurcated Eurodominant - China leads shift to low-q 50 yrs 1925 AD British trade but opium banned Britain lease on Hong Kong from 1898 Broken network links typify high q – trifurcated rise of Japan - China returns to high q 1950 AD Start of a low q Zipfian tail for world city distribution – trifurcated – but linked by airlines N-cohesion (2=competitive 1=monopoly land trade) leads q-scale, dichotomized (city rise/fall), in moving averages for 150 year periods: World land C routes integrating ↑F ↑CF ↑F ↑C ↑F N leads q to 1500, competitive trade cities Nq Inverse of N leads q to 1750, Portuguese & British Indian markets create choke-point trade city q-scale Nq Nq qN World land routes N.Sung S.Sung Genoa Portugal disintegrating __________________ /Mongols /Venice Dutch Engl.British USA--- Decol. q leads inverse of N (more choke-points) 1750-1900 (industrial revolution; maritime displaces land trade) Transaction costs, hegemony and inflation as q-correlated temporal variables Europe and Mediterranean C CCC CCCC CC Commercial I ??? ?I I I I I I I I I I I ? ?I I I I I I q Hi ? ? ? ? L ? ? ? h h h h L L L L L L L L h h h h h h L L L L h h L L L L L h h h L Inflation Lo/hi P P ? ? pP ? ? ? EEEEEEE? ? EEEEEEE? EEE q Lo FFFFFFF FFFFF FFF Financial capital 1 1 1 1 1 1 1 1 1 1 2 0 1 2 3 4 5 6 7 8 9 0 02570257025702570257025702570257025702570 05050505050505050505050505050505050505050 L/h lo/hi inflation figures (L=depression) are for that year forward Dominant Routes Conflict on Land Sea trade routes safer than land, 1318-1453/4+ Maritime (low cost) versus Land routes trade (pop. growth) (Spufford:407) Landed Trade Secure Peace of Westphalia Struggle for Sea Empire: Global Maritime routes Sea Battles Economy safe to 1815 Industrial Rev. French Sov. Landed Armies: safe land routes 1500-1650 Maritime Conflicts (Jan Glete) from 1760 Political Revolutions to 1814 Baltic conflicts: connection to Novgorod and Russia (lost) Swedish hegemony European access Trade net (low cost) versus (high cost) Euro-Hegemon examples (Arrighi 1994) Commercial Financial Constantinople Venice Genoa Amsterdam London New York The Medieval pause and Conclusions • • WHY DOES THE MEDIEVAL EUROPEAN RENAISSANCE ECONOMY FALTER CIRCA 1300? Major problem in Population Growth/Resource base Partly conflicts on land, internecine struggles Credit crisis between North and Southern Europe China invaded, change in Silk routes Trade dominance in the long terms begins to shift from betweenness centrality (Genoa; commercial capital) to global Flow Centrality (Bruges; financial capital), with later oscillations. Major collapse, long recovery "Long 13th century” reaches to today • Conclusions: city systems in the last millennium • City systems unstable; have historical periods of rise and fall over hundreds of years; exhibit collapse. City system growth periods in one region, which are periods of innovation, have time-lagged effects on less developed regions if there are active trade routes between them. NETWORKS AFFECT DEVELOPMENT. • • • • • • • Parts of the story in a nutshell Pax Mongolica: The routes are subject to policies of polities and empires, part of periods of pol./econ. dominance (Modelski et al. p.78,217)___Regional N. Sung 930S. Sung 1060Genoa/Venice 1190-___to Global____ Mongols 1250-90-1360 trans Eurasian Portugal 1430- Global system mapping Holland 1540- Global capital England 1640- Global industrial exports Britain 1740- Global organization United States 1850-Global informationGlobal market United States 1950- Decolonization -1990 Depolarization - Global hyperspace Globalization • • The Mongol administrative improvements of postal routes and support for merchants on the Silk roads were key to the rise of the Mongol Empire (on the scale of the later British Empire) and a first planned policy attempt at creating new routes for global trade and political globalization, i.e., going beyond earlier Roman and Greek (Alexandrian) attempts, for example. Modelski refers to the Mongols as a failed empire because they retreated to the east to dominate China until 1912. Their success and the benefits of East-West trade, however, were the spur to Portuguese and subsequent attempts at policy engineering towards globalizing trade and the periods of attempted West-East domination. A next study of planned globalization will start in 1290 and review globalization policies and pitfalls. – Globalization as a learning process – Globalization policies at attempts as dominance – The cycles of leading polities • And the two shorter economic cycles within them – The costs of losing dominance – The effects of wars over dominance – Paths to mutual regional support and peaceful resolution of competition end Cohesive extension of trade routes leads to a host of other developments… (figures courtesy of Andrew Sherratt, ArchAtlas) Multiconnected regions => structural cohesion variables (but the circumferences of these trading circles are large, not sufficient to induce fully competitive market pricing) Multiconnected regions => structural cohesion variables Multiconnected regions => structural cohesion variables Multiconnected regions => structural cohesion variables Some changes in the medieval network from 1000 CE Multiconnected regions => structural cohesion variables to 1500 CE (note changes in biconnected zones of structural cohesion) Project mapping is proceeding for cities and trade networks for all of AfroEurasia and urban industries for Europe in 25year intervals, 1150-1500 (our technology for cities / zones / trade networks / distributions of multiple industries across cities for each time period includes dynamic GIS overlays, flyover and zoomable web images)