Some ideas on a Post-Keynesian model of dynamic economic simulation

[Brutus]

Hi all,

I have been playing Democracy 4 for quite a long time, but I had always been hesitant to try some modding myself. Luckily it seems like with the new modding tool it would be easier than I had expected, but before trying to implement my ideas I would like to share some of these and get some feedbacks.

I always think that the economic simulation in Democracy 4 is not that dynamic. Sure, the feedback loops and interconnectedness of all simulation values make it incredibly vibrant than most other games, but it is too easy to reach a stable equilibrium in, say GDP. I have studied a little bit of Post-Keynesian economics and I find their model for explaining the fluctuations and business cycles of capitalist economies to be very convincing. I am certainly no expert in economics and I certainly do not pretend to incorporate all elements of PK models into this game, but I would still like to present them in a grossly simplified and distorted way:

Kaleckian economics believe that the most important determinant of any capitalist economy is the investment decisions of private capitalists. Capitalists do not invest in a hyper-rational way: they look at historical and current profitability to decide whether expand production or not. The profitability, in turn, is related to labor-capital struggle: if labor manages to advance their position, it will eat into the profitability of capitalists; they will either choose to pass that onto consumers through monopolistic pricing, leading to inflation; or they will choose decrease or even stop investing, leading to economic contraction. Thus the first new simulation value: Labor Bargaining Position, a value constantly floating between 10% to 90 % (all numbers are undecided) that depends on factors like labor law, unemployment, socioeconomic segregation, workers’ productivity and education, etc, and roughly corresponds to how large the proportion of economic surplus would be attained by the workers.

While investment is the most volatile part of economic activity, consumption is the more stable part. Different social classes have different marginal propensities to consume, meaning that as wealth becomes more concentrated into the hands of the wealthy, the consumer demand will contract, meaning that the vitality of the economy depends even more on investment, either domestic or foreign. As I discussed above, domestic investment decisions depend primarily on how the capitalists expect the profitability of their investment will be, with the information of recent profitability in mind. However, it also involves an irrational factor that Keynes had dubbed “the animal spirit”: spontaneous increasing or decreasing investment that cannot be explained by profitability alone is happening every time.

Foreign investment is also an important component of total investment, but it is even more volatile and tricky than domestic investment. Foreign investment can be done in portfolios, which would hardly lead to any significant real economy activities; it can be in extractive or menial manufacturing, in which domestic capitalists and workers will only receive a tiny proportion of profits, and it will not generate the multiplier effect if the export sectors do not have forward and backward linkages with other domestic industries. To simplify this situation, I propose another value similar to Labor Bargaining Position called Domestic Capital Bargaining Position, again between 10% to 90%, that depends on the country’s current GDP (thus how much you need to rely on foreign investment), technology, domestic or international policies and regulations concerning foreign investment, etc, that corresponds to how large a proportion of economic surplus generated by foreign investment would be retained as the profits of domestic capitalists. This value, combined with other stuff (such as concession policies to international investment, foreign relations, stability, global economy), will determine how much foreign investment you will receive. Again, like the Labor Bargaining Position, this is supposed to be a constantly floating value that will hardly reach an equilibrium.

There are many other aspects that I want to include (for example, I have not yet thought about how to incorporate the monetary side of the model), but this simplified model already provides some very intricate dilemmas. For example, sustaining welfare capitalism might lead to decrease of profitability, thus diminished investment (“capital strike”) or outsourcing. This might lead to economic contraction, unemployment and poverty, which might harm the bargaining position of labor and lead again to higher profitability. On the other hand, you can preside over a neoliberal wet dream with stark social inequality and thus diminished consumer demand, but your economy can still prosper if the profitability is high, leading to domestic or international capitalists’ investment. However, it might be very difficult to maintain such a situation, as business confidence and global economic can vary drastically than you had expected.

Here are some of the tentative formulas about how it would be implemented in-game wise. Again, all the numbers and exact formulas are tentative. I tried to let all values have varying degrees of latency as to prevent any stable equilibrium.

Labor Bargaining Power (0.1-0.9):
+Trade Unionist%, 2
+Education, 4
+GDP
-Unemployment
-Industrial Automation, 6
-Poverty, 2

Domestic Capital Bargaining Power (0.1-0.9):
+BusinessConfidence, 2
+GDP, 4
+Technology, 4
+Productivity, 8

Wages = GDP * LBP, 4
Profits = GDP * (1-LBP) - ForeignInvestment * (1-DCBP) , 4

//I want to make GDP an absolute or quasi-absolute value (that is, it has a great space for increasing or decreasing), since it will determine the absolute level of wages and profits

LowerIncome = Wages * LBP * 50% + Profits * (1% + LBP^3)
MiddleIncome = Wages * (1 - LBP *50%) + Profits * (2% + LBP^2)
HighIncome = Profits * (97%-LBP^3-LBP^2)

//I actually want to include another value called Labor Market Segmentation to simulate the uneven distribution of wages among different classes of workers. Here we will use Labor Bargaining Position as a proxy: it is reasonable to assume that as labor as a whole gains a more equitable position, the intra-working class differentiation might decrease somewhat.

//It is also reasonable to assume that, a better bargaining position for the workers will allow them to assume parts of the investment function of the capitalists. I am not sure how to implement this idea exactly. Here this is done in an indirect way: a proportion (ranging from negligible to significant) of profits will be added to their earnings, and parts of that extra earnings will be used for additional consumption. Another way of implementation is formally incorporating the part of investing done by non-capitalists.

Consumption = 80% * LowerIncome + 60% * MiddleIncome + 40% * HighIncome

//Values are tentative to show the different marginal propensity to consume. Specific values about different classes’ MPC can be created, though that might be too complicated

BusinessConfidence = 50%* (GlobalEconomy + Profits) + 10% * Stability, 2

//BusinessConfidence measures the willingness of capitalists to invest. I use the GlobalEconomy value as a proxy for the part of the inherent randomness of investment decisions.

Investment = 50% * (1+BusinessConfidence) * Profits + ForeignInvestment, 2

//Correct me if I’m wrong, I don’t think the current game has any mechanisms for different classes’ savings, so sadly we can only assume that proportion of profits not spent on investment simply disappears. Anyway, a percentage of profits will be used for expanding production; if the outlook is more optimistic, the percentage is larger. Foreign Investment should have a more sophisticated formula that sadly I don’t have time to think through now

GDP = Consumption + Investment * (1+Productivity), 2

//The 1+Productivity simulates the economic surplus in the classical Marxist term, and it allows for a simulation of sustained economic growth. However, with all the intervening mechanisms I think it will be very difficult to maintain either equilibrium or a perpetual growth of GDP.

[Brutus]

I have not come up with a way to dynamically simulate unemployment. I’m wondering if there’s a way to use GDP growth as a distinct value: GDP growth leads to lower unemployment, vice verse.

[PARAdoxiBLE]

Though there are some points I don’t agree or equations can’t see its reasoning, your ideas generally look interesting and thankfully feasible.

• ‘Animal Spirit’ or speculative behaviors should be possible to implement in your mod. Let’s see an example.

Just a few lines of simulation.csv would be enough to implement fluctuations shown above. It might take some fine tuning to get properties you need though. If you want more cyclic flux, you consider using global economy as a base or create simulations for that.

• For income simulations, though there already are 3 income simulations at the economy section, they might be unable to represent what you intend for several reasons. A) there are 2 kinds of income effects and ones with _fixed don’t affect these simulations. They do have impact on disposable income curve but no impact on Low/Middle/High Income simulations. B) Middle Income won’t represent actual income of middle-income voters for another reason - they also have other voter identities and thus have even more income modifiers, leading to different income level by different memberships. C) as people get or lose their income the ratio of poor/middle/wealthy will also change so just adding them up might give different results or even a chance to cheat the system (like boosting low income when nobody is deemed poor anymore). Though, this won’t be a big deal if you are going to see those who are poor with the original income to be poor. Just keep in mind that people’s income changes rather drastically in this game, at least in the disposable income screen.
• I also agree that it would be nice if labor market segregation or income gap among workers. I’ve recently tried adding unemployment as an additional factor of the wages-to-income link in my income adjustment mod (Lists of proposed changes to income simulations +misc) and made the poor more vulnerable to unemployment so that high wages & high unemployment would mean wider gap between the poor & middle income. Your Labor Bargaining Power approach sounds interesting too but unsure whether higher LBP will mean less pay for the middle income.
• It looks quite likely that income-boosting measures outside wages will effectively become a GDP stimulus. I guess it may need more factors to curtail consumption such as Inflation or dedicated Consumer Price.

[Brutus]

Thank you for your reply! I am really new to this game’s modding, and even though I have several dozens of hours playing this game there are still many game mechanics that I am not aware of.

• ‘Animal Spirit’ or speculative behaviors should be possible to implement in your mod. Let’s see an example.

That would be awesome! I used global economy as I thought that was the only genuinely random variable in this game, but if you can generate randomness otherwise the global economy can be used to determine specifically foreign investment.

For income simulations, though there already are 3 income simulations at the economy section, they might be unable to represent what you intend for several reasons

I have always found it difficult to understand how earnings and classes work in this game. Is there any detailed documentation that provides this information?

I also agree that it would be nice if labor market segregation or income gap among workers.

I actually just thought about labor market segregation after I wrote those formulas (somehow it didn’t occur to me when I wrote them). I agree a higher LBP might not necessarily lead to less pay for the middle income, but I think I really need to learn more about how income and classes work in this game to implement what I want.

It looks quite likely that income-boosting measures outside wages will effectively become a GDP stimulus

I agree. I was planning to first create a general fiscal framework, and then adding systems of monetary simulation. I’m not sure how much monetary stuff should and can be implemented in this game though, but inflation, exchange rate and foreign investment should certainly be there in this model.

[Brutus]

Also, profitability and labor-capital struggle should also lead to interesting consequences on industrial automation, which I forgot at first

[PARAdoxiBLE]

There are lines somewhere near I’ve drawn as shown above. That’s roughly how income groups are divided. If you click the ‘Original Income’ below the income screen, you will see original distribution of people’s income. These incomes will be modified by various factors such as Income Tax (-), Private Heathcare (-), Bus Subsidies (+), State Pensions (+), or Wages (+/-). Modified income is called disposable income here and it ultimately decide who belongs to which income group. Let’s see an example here.

Person A is Capitalist, Liberal, Environmentalist, Commuter, and State Employees. He has original income of $ XXXXX and it will be in the range of middle income. Now he pays Income tax at middle-income rate, gets some tax credit (which doesn’t benefit the poor), spends money on essential parts like healthcare & education (unless the gov covers them all or private ones are banned), happens to realize he have more money than before thanks to high wage level, and many more. He also have other income modifiers such as pro-environment policies benefiting environmentalists, income boosts for State Employees from public spending programs, and possibly Bus Subsidies making it cheaper to commute to work. Now his income will be very different from the original one. He may become slightly poorer if taxes & expenses were too high and even be re-classified as poor.

This happens to every single dot in the disposable income screen.