Age of Discord: Book Review

Age of Discord: A Structural-Demographic Analysis of American History (2016), Peter Turchin; an interesting time series macro analysis of relative tranquility versus discord based on wages, elite dimensions and various economic and other factors. My initial interest was using these factors on primitive economics, consistent with my interest on food history (the original tests for this theory). However, most of the book is based on US history when statistics are available (from 1790). Much of this is useful to MVG analysis and therefore a useful blog. Question: Is America headed in the wrong direction? Note that this was written in 2016, completed before the Trump presidency. Consequently, the analysis is based only on economic and demographic statistics. The scary result: chaos was predicted based on 50-year cycles for around 2020.


The prediction is: labor oversupply leads to declining wages, then rising income and wealth inequality, elite overproduction and infighting, leading to rising polarization, then potential chaos. The standard way is to look at specific historic events (subject to hindsight bias) to describe and explain the "American story," a different approach that often seems to come to similar conclusions.


Turchin is an evolutionary anthropologist, focusing on cultural evolution and has written a number of books on similar topics. The history he focuses on is usually interpreted based on specific people and events, rather than macro time-series models using a variety of economic, demographic and sociological variables. The results are fairly strong (assuming the use of long lags is reasonable) and consistent. A variety of perspectives can use Turchin’s findings in diverse contexts, including MVG from an economic/political context.


Part One: A Theoretical Introduction


Chapter 1: Multi-Secular Cycles in Historical and Modern Societies. It opens with the Civil War and how myopic the political leaders were. A key point is the need to “gain a deep understanding of the fundamental mechanisms affecting functioning of complex marcosocial systems. … Progress is made by formulating general theories whose predictions can be tested with historical data, constructing large databases, capitalizing on natural experiments … [with] planned comparisons between the predictions of two or more rival theories and data. … Traditional history has generally failed to provide useful guidance for public policy … by cherry picking examples from the historical record” (p. 5-6).


The first questions is what are the general principles of social dynamics and do they apply to different societies? Turchin starts with pre-industrial states, claiming stable and unstable century-long periods, using Rome (500 BC to 400 AD) and post-Carolingian France (800-1700) based on his index of political instability. Historians named some “instability waves,” like the crisis of the Late Middle Ages and Seventeenth Century; ditto, China and its dynasties.


“Dynamical pattern of sociopolitical instability in agrarian societies is complex” (p. 11) involving multiple cycles. Based on structural-demographic theory: “population growth in excess of the productivity gains of the land has several effects on social institutions. First, it leads to persistent price inflation, falling real wages, rural misery, urban migration, and increased frequency of food riots and wage protests. Second, rapid expansion of population in “elite overproduction”—an increased number of aspirants for the limited supply of elite positions” (p. 11). The result is rivalry and factionalism, an expanding army and bureaucrats raising taxes, then fiscal crisis, often rebellion and breakup of central authority. Turchin predicts that pandemics should happen in disintegrative phases—which seems a bit of a stretch. (On the other hand, a pandemic can cause a downturn.) Strong states should protect peasants from threats. (Hobbes thought so, too.)


Turchin looks to medieval England and other examples to support the theory. He includes “Malthusian limits of population increase and feedbacks between supply of labor and real wages and popular misery and elite expansion. A major predictor of political instability is “elite overproduction.” This also is related to population pressure and inverse real wages (too many people with reduced wages).


Turchin then moves to industrial economies and a new structural-demographic theory, by using three concepts: 1) Neo-Malthusian principle: high population growth leads to falling living standards; oscillations in demand and supply of labor expected, 2) elite overproduction and 3) causes of political instability. Low labor costs lead to higher incomes and wealth for elites; this causes greater income inequality and an increase in elites—elite overproduction. This increases the probability of intraelite conflict. This should be the chief cause of political instability, aided by falling living standards (with counter-elites mobilizing the masses, and fiscal crisis. Instability (called disintegrative trend) caused by overpopulation, elite overproduction, and fiscal crisis. Note that popular misery (immiseration) not a problem as long as elites remain unified. The idea of the book is to test this theory from 1780 to now (slightly before 2016, but data dependent).


Chapter 2: Modeling Structural-Demographic Processes. The empirical data are population and labor wages, elites, the state and how they interact. Population variables include total population, age structure, urbanization, incomes, and social optimism. Elites include numbers, composition, incomes and wealth, conspicuous consumption, social cooperation norms, and intraelite competition or conflict. The state includes size, revenues, expenditures, and debt, plus legitimacy. Instability includes radical ideologies, terrorism and riots, revolutions and civil war. “Youth bulges tend to be politically highly destabilizing, because a sudden increase in new workers joining the labor force tends to depress their employment prospects and wages” (p. 22). Elites include established elites, new elites, and aspirant elites. Counter-elites are those frustrated. Income and wealth associated with power. Government issues associated with size (number of employees, proportion of GDP) and ideology.


Political stress index (PSI) should explain state collapse, rebellions, and civil wars; then broken down by structural causes and specific triggers. PSI looks at population-elites-state: mass mobilization potential (MMP), elite (EMP), and state fiscal destress (SFD) (p. 26). MMP includes relative wages, urbanization rate, and age structure (20-29 year-olds). EMP is measured by elite income compared to GDP per capita; number of elites (to total population).SFD: distress by national debt to GDP (or taxes), measures of dissatisfaction (like polls).


Part Two: Overview of Structural Demographic Variables: 1780-2010.


Chapter 3: Demography and Wellbeing. The US started with lots of land and few people. Then immigration increased. Immigration generally increased from the 1830s to about 1910, then decreased until the 1930s, then increased again; many factors involved. Until the 20th century most migration came from Europe. Real wages increased by 300% from 1790-1850 (1.8% a year), then decreased from 1850-80. Wages increased substantially from 1930-70, then stagnated after that. Relative wages (deflated by GDP) increased from 1900 to 1960 and decreased after that. Other measures of wellbeing were similar—which fits the idea of “structural-demographic.” Turchin divides the US into four phases: 1780-1830 for improvements; 1830-1910 for decline associated with massive immigration, wage stagnation and declining wage labor to GDP; 3rd was 1910-1960 for improvement based on rising real wages and other measures; 4th from 1960 for a decline based on massive immigration and wage stagnation.


Chapter 4: Elite Dynamics. Elites are the ruling class, with power defined as the ability to influence behavior of others: military, economic, political (including administrative) and ideological. Most power in modern states is administrative; likely economic in the US: “Its structural economic power allows it to dominate the political class through lobbying, campaign finance, and appointments to key government positions. … The corporate community also controls the ideological basis of power through ownership of mass media corporations and a policy-planning network made up of foundations, think tanks, and policy-discussion groups” (p. 76). Power should shift over time. Turchin looks at the wealth of the 1% as a measure of elites (also the extreme value index).

Inequality declined between 1920-80, with the period from the 1940s-70s considered the “Great Compression,” where the top 1% wealth fell from 44% to 20%. Once again, these fortunes went through cycles, up slowly until 1830, then a big rise in fortunes until 1912: a rise from the 1820s-1900, down until the 1960s, then up from the 1970s. Different measures of elite status showed similar results. One test was number of lawyers and lawyers’ salaries. Number of lawyers dramatically increased from the 1970s, with a small percentage of high paid lawyers. Similar results for MBAs.

Congressional positions from liberal to conservative also is tested. One test is scoring by the Americans for Democratic Action (liberal) versus American Conservative Union. Another test is political polarization based on average scores of Democrats and Republicans in Congress each session. The “Era of Good feeling” with virtually no polarization from the 1820s, then rising polarization until about 1908, then a decline and low polarizations from the 1930s until about the 1980s (especially the New Deal and World War II), the era of Reagan and the start of rapidly rising polarization; inequality and immigration helped fuel this rise.


Elite fragmentation (political polarization) peaks around 1910, low point about 1940, rise begins in late 1970s and peaking about 2010 (and it will go up from there. Similar pattern for labor oversupply (% foreign-born), and elite overproduction (wealth inequality index). Immigration increases poor and those that cannot vote; voter age rage moves to the right. “Over the course of American history elite overproduction and popular wellbeing have moved in opposite directions” (p.96).


Chapter 5: Th State. The relationship of the state to elites differs over time. The Roman Republic was dominated by elites. The Tang Dynasty in China (618-907) was governed by the state bureaucracy determined by civil service exam. The role of the federal government has changed over time. Initially, its main cost was war (and paying for prior debt) financed by customs duties. The big change was the shift to income tax and, with the New Deal of the 1930s, focused more on public welfare—especially until the 1970s. Preparation for war continued to be a major cost. Local governments depended on property taxes, mainly for local spending including education.

“High levels of inequality are destructive of cooperation. … Faith in government peaked in 1965. … The wealthy segment of the population is primarily interested in reducing their taxes” (p. 106). “A social mood of national consolidation and broad-based cooperation, which typically takes hold during the integrative phase of secular cycles, is usually necessary to generate concerted collective action … including a willingness of elites to tax themselves” (p. 109). The top tax rate was 91% until 1963.


Chapter 6: Dynamics of Sociopolitical Instability. “Political instability is violent, group-level conflict within a state” (p. 113). This includes attacks on officials and groups for political and other social reasons. One database from 1780-2009 had almost 1,600 events, such as riots (56%), lynchings (28%), and terrorist attacks. If mass shootings are included, these events have increased since then. Categories include race, labor-management, and politics. Big instability happened around 1870, 1920 (e.g., Red Summer of 1919), and 1970 (e.g., Vietnam protests); big decline started with the New Deal.


Part Three: A Complete Secular Cycle: From the Revolution to the Great Depression, c. 1780-1930.


Chapter 7: Long-term Trends, 1780-1930: A Synthesis. The theoretical variables should be interconnected through feed back loops. Turchin is interested in “trend-reversals;” a change in one variable affects the others, suggesting the potential to predict outcomes. Political instability should correlate with labor oversupply, elite overproduction, and interelite polarization (measures of wellbeing), subject to lags. Looking at multiple variables, he sees trend reversals between 1820-40 and 1900-20. Graphing these and drawing an overall trendline suggests those trend-reversals up about 1820 and 1930; down about 1860 and 1960s.


Chapter 8: The Antebellum Era. Start with the era of good feeling around 1820, when there was only a single party: Monroe was president, a Democratic Republican; the Federalist Party faded away and was only gradually replaced—eventually by the Republican Party in 1856. Tocqueville visited in 1831. The country was growing. Slavery became a big issue later. White Americans were relatively homogeneous. Jackson became president in 1828 as a populist Democrat. Overpopulation on the eastern seaboard became a negative factor. The US “sectionalized” into North and South. Elites in the two groups became less collaborative. The South was the richest (for the elites); this changed with Northern manufacturing (favoring high tariffs and domestic manufacturing). Slavery became a moral issue. Textile manufacturers still allied themselves with Southern elites, also true for Northern bankers. The canning of Charles Sumner by Preston Brooks indicates the degree of political instability.


Chapter 9: The Road to Civil War. Problem of oversupply of labor because of immigration and movement of rural to urban areas. Elites grew substantially after 1840. Mass Mobilization Potential (MMP): real wages, urbanization rate and age structure. Elite Mobilization Potential (EMP): relative elite income, intraelite competition, both up after 1840. Civil War destroyed much of Southern wealth; also fueled Northern wealth, creating a national banking system and push for high tariffs.


Chapter 10: From the Gilded Age to the New Deal. Persistent instability after the Civil War, with violence against minorities, strikes and assassinations (e.g., James Garfield, William McKinley), a class war in 1919 (21% of labor disruptive, Red Summer riots; labeled as labor conflicts, interracial tensions, and political extremism. Progressive era reforms started 1900-20, then the New Deal. Northern business and politicians became the new “national upper class;” aka: “the Establishment.” This included interlocking foundations, think tanks and policy organizations: Carnegie Corp., US Chamber of Commerce, Rockefeller Foundation, National Industrial Conference Board, National Bureau of Economic Research, Brookings Graduate School of Economics (merged with others to form the Brookings Institution), and Social Science Research Council, all between 1911-23. After great merger movement around 1900, Sherman Act, Clayton Act and FTC passed. Immigration (e.g., replace unions with immigrants) led to protests and riots. Wellbeing trends up from 1900 to 1960. Impact of Great Depression on employment, wages “sticky.” New Deal was an economic growth period. New Deal bad for the rich. Most of the programs had been contemplated for some time. Skepticism of government intervention started with 1970s stagflation (plus bear market).


Part Four: The Current Secular Cycle: From the Great Depression to the Present.


Chapter 11: Long-term Trends, 1930-2010. The post-WWII period proved good on most dimensions, including “convergence.” Most reversals took place from 1960-80. Possibly the Reagan Era Trend Reversal, although reversals started before then. They moved bigtime with Reagan: suppression of unions, income and wealth inequality.

The US went from a net exporter to a net importer in 1975, reducing the demand for US labor. Other key reversal trends: labor oversupply, economic wellbeing (economic, health and social), inequality, elite over production, polarization, trust in government, sociopolitical instability. Real wages peaked in 1970s. College costs increased starting in 1980. Overall, 29 of 30 structural-demographic variables reversed. The exception was life expectancy (which has gone down recently, in part because of Covid).


Chapter 12: From the New Deal to the Reagan Revolution: A Dynamic Model. A basic question is why real wage growth stopped in the 1970s, followed by stagnation and decline. Labor productivity growth was 1.8% a year from 1927-48. Supply of labor increased faster than demand beginning about 1970. Political compromise ended with Young Republicans (Newt Gingrich, Vin Weber, Dick Armey, and Tom DeLay). This included a fight against labor unions. Also the rise of multinationals, focusing on self-interest. [Private equity also rising, not considered in the book.] A regression model of wages = GDP/capita, labor demand/supply and non-market forces tracks wages well for 1930-2010, R2 = 93%. Since 2011 labor supply increased faster than demand mainly because of migration, plus more women working. Median wage for men declined as median household income rose (but more slowly than GDP per capita).


Chapter 13: Social Pressures Toward Instability. Relative wage and elite wellbeing. “The elite aspirants who end up among the winners tend to be rewarded disproportionately, but at the same time there is a growing proportion of losers. As a result, intraelite inequity explodes” (p. 235). As an example, lawyers per thousand went from 1.6 to 3.9 over the last 40 years. Only some became super-rich elites (especially since 2000). Political stress indicator (PSI) when wages drop and elite overproduction exists. Measures include debt per capita and distrust in government (Pew poll). PSI stable from 1945 to the 1970s, then up gradually to the 1990s, then increased dramatically to 2012. Dynamics of these indicators increasingly complex.


Chapter 14: Conclusion: Three Ages of Discord. Turchin expect long secular cycles of political instability; positive in Era of Good Feelings around the 1820s and post-war prosperity of the 1950s. Instability spiked in 1870 [presumably the Civil War represents a lag], 1920 and 1970—Ages of Discord. Instability principle: population growth a big factor (labor oversupply), the elite overproduction and political fragmentation. The timing of 50-year cycles suggests 2020 the next one (the book was published in 2016, before Trump’s election).