State Space Models

All state space models are written and estimated in the R programming language. The models are available here with instructions and R procedures for manipulating the models here here.

Wednesday, July 9, 2025

World-System (1950-2000+) Iranian Globalization and Unemployment

 




The IR2 component (Unemployment and Globalization) is the second most important state variable in the IR_LM model and explains about 12% of the variation in the indicators. Unlike the IR3 (Environmental-Globalization feedback controller), IR2 is not a feedback controller but rather a historical componentThe White Revolution and the  Foreign Policy of the Kennedy Administration put a high premium on Modernization and Globalization which rapidly brought population into the cities without immediate job availability (see the video on The Iranian Revolution).




The result was very high levels of Unemployment (a peak of almost 15% in the mid-1980s--see the graphic above). Notice that Globalization peaked in the mid-1970s (first graphic above) while Unemployment peaked about a decade later. 

The two graphics above show the attractor path (dashed red line) for both  Globalization and Unemployment and show that each was above sustainable levels until the early 1990's. The IR2 historical component shows why Islamic Fundamentalism organized wide-spread discontent in Iranian cities before the Iranian Revolution.


Notes

The IR_LM Measurement Model is:


The first state variable (row of the Measurement Matrix, IR1) describes overall growth in the Iranian Economy. The second state variable, IR2, describes  Unemployment and Globalization, (IR2 = 0.9091 SL.UEM.TOTL.ZS + 0.335 KOF) and the third state variable, IR3, is an environmental-globalization feedback controller, (IR3 = 0.827 KOF - 0.3878 SL.UEM.TOTL.ZS - 0.258 EF - 0.217 EG.USE.COM.KT.OE)for unemployment, ecological footprint and energy use.

Further reading:

Blog Roll:

Monday, July 7, 2025

World-System (1950-2100+) Global Temperature Projections


A Recent Paper in Nature Communications: Earth and Environment was titled Warming of +1.5 °C is too high for polar ice sheets.** In terms of projections from my models (above), this would mean that if we reach the 15.5 temperature level, it will be too late for the Polar Ice Caps to maintain their integrity and sea level rise will inundate most countries around the World (to include Mar-a-Lago, President Trumps Empire). Does this mean catastrophe for the World-System

The first projection to consider is the Random Walk (RW). If Global Temperature were a  Random Walk it would mean that it cannot be forecast, a result that would please Climate Change Deniers. Given the high variability of the Global Temperature data, it should be no surprise that, in the short-run year-to-year, Global Temperature is a RW given the Akaike Information Criterion [-77.33, AIC=-65.15,-55.0]. However, in the long-run there is a clear trend. The issue is how far into the future can the trend be carried.

In the Stokes (2025) Global Sea Level article the conclusion is that, to prevent destabilizing the Arctic Ice Caps, Global Temperature should return to the levels somewhere around the 1980's. The graphic above plots a forecast of the Temperature BAU model that has no inputs (no radiative forcings). Without forcings from human systems (anthropogenic forcings) the model would say within the "safe" zones. Unfortunately, isolating Global Temperature from human systems seems impossible.



The next one of my models to consider simply uses CO2 emissions to predict global temperature, something similar to the endpoint of the Kaya Impact Model (graphed above) used by the IPCC. The model projects a linear increase in temperature that eventually reaches a steady state (the model is stable). Using the Akaike Information Criterion (smaller is better) [170.2,AIC=198.3 217.1], it is not a strong competitor.

Another important model involves the technical coefficients of the Kaya Impact Model (q, e, c, t in the graphic above). The model is important because improvements in Kaya Technology (efficiency) such as renewable energy, etc. should ultimately reduce Global Temperature. The model shows continuing increases in Global Temperature, but because the model is stable, a steady state will eventually be reached but at a higher Global Temperature than 15.5 degrees (dotted red line in first graphic). In other words, even though technological change can moderate temperatures, it will not be enough to prevent melting of the Polar Ice Caps. Still, the model is a good competitor using the Akaike Information Criterion [-47.94, AIC=14.95, 67.81].*** 

Finally, driving Global Temperature with the state of the World System using two different models (the dashed blue line and the pink line in the graphic at the beginning of this post), produce growth-and-collapse modes and keep global temperature below 15.5 degrees. The WL203 model (dashed blue line) is the best model (beyond the RWusing the Akaike Information Criterion  [-70.95, AC=-10.41, 34.52] compared to the WL20W model [-4.007, AIC=64.63, 107.9].

You can experiment with the stable WL203 model (here) or the unstable WL20W model (here).


Notes


The measurement model for the World System has three historical environmental controllers: W1=(Growth-LivingPlanet), W2=(LivingPlanet-TEMP) and W3=(P.Oil.-TotalFootprint).  Information about the indicators (LivingPlanet index, TEMP, TotalFootprint), etc. can be found in the Boiler Plate.

** Recent data from NASA's CRES (Clouds and Earth’s Radiant Energy System) satellite data also indicates that Earth's Albedo (ability to reflect sunlight and reduce warming) is decreasing!

*** There is another type of Technological change that involves productivity (the dashed black line in the graphic at the beginning of this article). It shows a similar, if somewhat worse, time path than Kaya Technology and reaches 15.5 degrees earlier in time.

Sunday, June 22, 2025

World-System (2000-2150) Seven Futures for Iran


The Trump II Administration in the US has just attacked Iran and claims to have seriously damaged Iran's Nuclear Sites (here). Iran claims it will retaliate, but the future after the unprecedented attack using B2 Stealth Bombers and Bunker Buster Bombs remains unclear both for Iran, the US and Israel (who started the attacks on Iran). The future is unknowable and there is a lot of current speculation. I can use the IR_LM (Iran, Late Modern) model to project seven possible future scenarios for Iran (graphic above).

The Business-as-Usual (BAU) model is clearly the best and it forecasts unending exponential growth for the Iranian Economic System. The worst scenario is Geopolitical Alignment with China (CN) which leads to collapse. The scenario for alignment with Russia (RU), (light-blue dashed line) is quite cyclical and unstable. Alignment with the US produces stable growth out to 2100. The World System (W) and the EU Alignment models are not much better than a Random Walk (RW).

These seven scenarios suggest that Iran would prefer to be left alone to pursue its own future, but that seems unlikely to happen. Geopolitical Alignment with the US  would also produce a desirable future, but that also seems unlikely to happen. Russia (RU) might be attractive in the very short run but that alignment will prove to be unstable. This leaves Iran with no desirable futures within the IR_LM BAU model.

You can experiment yourself with the IR_LM BAU model here. The period prior to the Iranian Revolution of 1979 sowed possibilities of alignment with the West, but after the Revolution the economy was converted into strong central planning aimed at Autarky (see the Economic History of Iran). In these circumstances, the IR_LM BAU model could be stabilized at any time by reducing growth rates (a counterfactual you can run here). Notice in the IR_LM BAU Model that the feedback controllers are dominated by Globalization, Unemployment, Ecological Footprint and Energy Use.

ChatGPT reports the following:


For more information about how the statistical models were constructed and estimated see the Boiler Plate.

The Axis of Evil


Further reading:

Blog Roll:

Friday, June 13, 2025

World-System (2000-2150) China and the Global South

 


In a prior post (here), I looked at China's possible Geopolitical Alignments, first with the World-System and then with Russia. In this post, I look at China's possible Geopolitical Alignments with the Global South. In the long-run, alignment with Latin America (LA) would clearly be in LA's interest; Africa (AF), not so much.

Thursday, June 12, 2025

World-System (2000-2100) Six Futures for China, One Scary

 


Most Geopolitical Alignments are mildly beneficial, except for alignment with the World System which is not. The order of benefits (in terms of comparison to a Random Walk, RW) are Asia Region (AXJ), Business as Usual (BAU, no alignment) and United States (US). 



One forecast missing from my standard forecasts (and the most concerning) is the one involving Geopolitical Alignment between Russia and China. Although the alignment has been publicly proclaimed by each country, there is increasing concern about CN intelligence activities in Russia (here). For more information about China-Russia relations, see the Wikipedia page (here).

I will investigate Global South alignments for China in a another post here. For more information about how the statistical models were constructed and estimated see the Boiler Plate.

Notes

Excerpts from ChatGPT:

Wednesday, June 11, 2025

World-System (1950-2150) India as a Small Regional Country

 


A recent article in the New York Times (here) about the evident decline of US Hegemonic Power suggests that India (IN) might be one of the countries competing for world leadership once US dominance has passed. India might well try to resurrect the "... Indosphere, a zone of Indian cultural and linguistic influence that stretched far beyond the subcontinent in the period roughly between 250 B.C. and A.D. 1200..." Another article (here) suggests that India is in a difficult position, closer to the US than to Russia but unwilling to alienate other BRIC countries (see below in the Notes).

The question not being asked here is what would be the best Geopolitical Alignment for a country such as India and does that alignment involve Hegemonic domination of the World-System. This post investigates the question by estimating a state space model for India in the period (1950-2000) and then testing various Geopolitical Inputs to see the effect on the Indian Economy.

The graphic above shows the two state variables, IN1=(Growth) and IN2=(Q-N), with forecasts out to 2150. The best model (in terms of the AIC statistic, AIC=-6272.714 when compared to the BAU model AIC=-385.2805, smaller is better) to produce those forecasts (see the Boiler Plate for a description of the state space model and the data sources used) is the IN_LM_SC model (a Small Open Country) model. It shows that India would be approaching a steady state around 2150 with part of the Asia Region.

In another post (here), I look at alternative forecasts for World-System Hegemonic Dominance. India is one of the also-ran countries (in fact, the worst performer), not providing large growth stimulus to the World-System. In the upcoming period of Environmental Crisis and Hegemonic Chaos, maximum growth might not be the best choice for the World-System but it will be a Dynamic Attractor.

Notes

In term of Systems Theory, a Small Open Country is one that has no state variables. The state of the system is driven by the dominant Input Economy (the Asia Region in this case).  Here are the System Matrix (F) and the Input Matrix (G) for the IN_LM_SC model:


The small values for the coefficients in the System Matrix (e.g., -2.605049e-23) are effectively zero to within machine precision.

You can run the IN_LM_SC model yourself at the following Google Sites (here). You can also run the other BRIC country models:


Monday, June 9, 2025

World-System (1950-2150) Hegemonic Dominance Forecasts


A few new books have been written and reviewed in the New York Times (here) about the evident decline of US Hegemonic Power in the Trump II Administration. From what I can tell, all of the discussions are qualitative and speculative. Since the Future is unknown, there is plenty of space for conjecture.

What has not seemed to have been done is to make multi-model forecasts for the World System given inputs for different countries that are competing for Hegemonic Dominance. Since I have state space models for the World System and for all of the Hegemonic contenders (you can inspect and run the models here), I might as well see what the forecasts say.

The graphic at the beginning of this post shows alternative forecasts for overall growth in the World System. There are three exponential growth contenders: Japan (JP), the US (US) and Russia (RU, see below). The "also rans" are the United Kingdom (UK), Business-As-Usual (BAU), China (CN), India (IN) and a Random Walk (RW) to find a stable, sustainable attractor path.




The surprise forecasts is that Russia dominance of the World System (RU) would create unending, unstable exponential growth. Dominance by other countries have uniformly more modest impacts.

Aside from the fact that living in a Russian-dominated World-System would be an unpleasant authoritarian nightmare, the reason for unending exponential growth is that Russia would supply a rich natural resource base to be exploited by the World-System. Maybe from some perspectives, unending exponential growth is a positive thing, but from a Environmental perspective, it would be a disaster. We have to find a way to limit growth rates and, evidently, linking the World System to Russia is not the way to do it.

You can experiment with the World System model here. You can run the individual countries on the following sites:

  • Russia (RU_LM Model) An unstable model with strong Export-Employment historical controllers. Becomes cyclical when stabilized.
  • India (IN_LM Model) A Malthusian Economy currently in collapse mode. Stabilizing the economy is beneficial.
  • China (CN_LM Model) An unstable economy with strong historical Export-Employment controllers. Stabilizing the economy produces a steady state after 2100.
  • Japan (JP_LM Model) The Japanese economy has an unstable, sensitive relationship to the Export-Employment (X-HOURS) historical controller. Stabilizing the economy produces an immediate steady state without further growth. Prior to the Great Recession and the 2008 financial crisis, Japan practiced Shūshin koyō (employment for life). Adjustment to market conditions were based on modifying hours. In 1979, Ezra Vogel published Japan as Number One: Lessons for America. The Great Recession and the 2008 financial crisis put an end to fears of Japanese domination of the World-System.
  • US (US_LM Model) The US is the current Hegemonic Leader of the World-System, but it's leadership is collapsing in the Trump II Administration. The US Economy is an unstable, cyclical model with unstable Export-Employment historical controllers. Stabilizing the model produces a cyclical steady state with wide swings.
  • UK (UK_LM Model) The UK model is also unstable and cyclical but the cycles are mild until well into the 22nd Century. There is no historical controller for HOURS and both HOURS and the (X-N) population controller follow cycles in the overall economy. The UK withdrawal from the EU (Brexit) can be viewed as an attempt to shield the economy from European shocks.
  • Some other countries you might investigate are:

    • Brazil (BR_20 Model) An unstable model that leads to collapse (stabilizing the model also leads to collapse). There are strong historical unemployment and environmental Controllers. See if you can find a steady state for this economy! 
    • South Africa (ZA_20 Model) An unstable economy with strong Globalization-Environmental historical controllers. Stabilizing the economy would lead to a steady state in the distant future (well after 2100).
    One thing to keep in mind when exploring potential hegemonic leaders is that the countries being dominated will to some extent (depending on the strength of their historical feedback controllers) take on the historical patterns of the hegemonic leader (India being an extreme Small Country example). Cyclical patterns that may be good for a decade, might not be so desirable during the first downturn.

    You can learn more about how the models were constructed in the Boiler Plate. For more discussion of the Indian Economy, see India as a Small Regional Economy.