04-23-2017, 05:41 AM
I have tried to interest some of the other folks in posting here since this site is better for serious posts, threads don't get buried for instance. Also it is better from presenting data. People keep saying they want more theory. Well here's a theory post I just made at the other site. Let's see if it gets any engagement over here.
I have a new modeling result. Look at the figure below. Shown is the model output (dashed line) and a measure of sociopolitical instability obtained from a weighted average of Turchin's database of political violence, and Chas's and my database of "cultural instability"--new conceptions of religious meaning and associated enthusiasm (revivals, alternative lifestyles, founding of new religions and denominations). To understand the significance of what I am presenting, note that these results have no additional adjustable constants (fudge factors) beyond what Turchin's model uses, but fit the data pretty well.
![[Image: Instability-cycle-model.gif]](https://mikebert.neocities.org/Instability-cycle-model.gif)
Background:
Peter Turchin proposed a model for generational cycles of instability (related to but not exactly the same as turnings) that is based on models developed for the spread of infectious disease. The population is divided in three categories: naïve people, radicals and moderates, who correspond to uninfected, infected and recovered populations in an epidemic. Basically the idea is there are radical ideas that can spread to naïve people by social engagement with radicals, which convert them into radicals. Every year a certain fraction of radicals lose their radicalism and become moderates. Radical formation is proportional to the number of radicals (more radicals mean more exposure of naïve people to radical ideas). The proportionality constant is modified by the number of moderates--more moderates suppresses radicalization. The result of this process is periodic bursts of radicalism that burn out in time.
This process operates on a range of ages from coming of age (assuming to be 21) and retirement from political life, AR. In addition the process of radical conversion to moderates occurs after lag period (it takes time to become disgusted with radicalism). Turchin shows that will some sets of parameters, regular cycles of 50 years length can be obtained. He also reports that cycle length is largely dependent on AR and the lag period.
My version:
I implemented Turchin's model with two modifications. First instead of a fixed AR as Turchin used, I set AR = AL+10 where AL is the average leader age obtained from data provided by Neil Howe at his website and which I have previous discussed. Thus AR rises from the late fifties around 1790 to 72 today. This has the effect of increasing cycle length as time progresses. The second modification was I increased the radicalization parameter (the propensity of naïve people to become radicalized upon contact with radicals) for the first year when a new liberal generation of leaders came to power (as forecast by the generational model). These dates were 1802, 1831, 1862, 1896, 1933, 1968, 2008. They are simply 1775 (assumed to the start of a "revolutionary generation coming of age) with AL-21 added to it successively.
The result of this boost was nil when it happened in a "down period". When it happened in an "high" period it adjusted the date of peak radicalization slightly closer to the dates when the new generation comes to power. For the recent period I added a boost in 2001 for 911 in addition to the boost for the 2008 coming to power of the Boomers--as defined by the generational model. Thus, the next peak in instability is forecast for the middle of next decade (without these additions it would be later, around 2030).
Now this is not a turning model. It is a model designed to fit the empirical instability data which shows that the last 4T was not a period with a lot of unrest. Thus, the fact that the model forecasts no period of instability for the last 4T. The period around 1920 had lots of unrest and the model predicts peak unrest in 1922. The model explains low levels of violence during the Depression by suppression of radicalization by moderates, who peaked in the mid 1930's.). These moderates were Lost radicals of the teens and early 1920's who had become disenchanted with radicalism. This might be why we avoided fascism last time.
I have a new modeling result. Look at the figure below. Shown is the model output (dashed line) and a measure of sociopolitical instability obtained from a weighted average of Turchin's database of political violence, and Chas's and my database of "cultural instability"--new conceptions of religious meaning and associated enthusiasm (revivals, alternative lifestyles, founding of new religions and denominations). To understand the significance of what I am presenting, note that these results have no additional adjustable constants (fudge factors) beyond what Turchin's model uses, but fit the data pretty well.
Background:
Peter Turchin proposed a model for generational cycles of instability (related to but not exactly the same as turnings) that is based on models developed for the spread of infectious disease. The population is divided in three categories: naïve people, radicals and moderates, who correspond to uninfected, infected and recovered populations in an epidemic. Basically the idea is there are radical ideas that can spread to naïve people by social engagement with radicals, which convert them into radicals. Every year a certain fraction of radicals lose their radicalism and become moderates. Radical formation is proportional to the number of radicals (more radicals mean more exposure of naïve people to radical ideas). The proportionality constant is modified by the number of moderates--more moderates suppresses radicalization. The result of this process is periodic bursts of radicalism that burn out in time.
This process operates on a range of ages from coming of age (assuming to be 21) and retirement from political life, AR. In addition the process of radical conversion to moderates occurs after lag period (it takes time to become disgusted with radicalism). Turchin shows that will some sets of parameters, regular cycles of 50 years length can be obtained. He also reports that cycle length is largely dependent on AR and the lag period.
My version:
I implemented Turchin's model with two modifications. First instead of a fixed AR as Turchin used, I set AR = AL+10 where AL is the average leader age obtained from data provided by Neil Howe at his website and which I have previous discussed. Thus AR rises from the late fifties around 1790 to 72 today. This has the effect of increasing cycle length as time progresses. The second modification was I increased the radicalization parameter (the propensity of naïve people to become radicalized upon contact with radicals) for the first year when a new liberal generation of leaders came to power (as forecast by the generational model). These dates were 1802, 1831, 1862, 1896, 1933, 1968, 2008. They are simply 1775 (assumed to the start of a "revolutionary generation coming of age) with AL-21 added to it successively.
The result of this boost was nil when it happened in a "down period". When it happened in an "high" period it adjusted the date of peak radicalization slightly closer to the dates when the new generation comes to power. For the recent period I added a boost in 2001 for 911 in addition to the boost for the 2008 coming to power of the Boomers--as defined by the generational model. Thus, the next peak in instability is forecast for the middle of next decade (without these additions it would be later, around 2030).
Now this is not a turning model. It is a model designed to fit the empirical instability data which shows that the last 4T was not a period with a lot of unrest. Thus, the fact that the model forecasts no period of instability for the last 4T. The period around 1920 had lots of unrest and the model predicts peak unrest in 1922. The model explains low levels of violence during the Depression by suppression of radicalization by moderates, who peaked in the mid 1930's.). These moderates were Lost radicals of the teens and early 1920's who had become disenchanted with radicalism. This might be why we avoided fascism last time.