Population Change Doesn't Affect Maximum GDP, Youth, Retired

Democracy 4 General Discussion
1

Raising or lowering the population does not change the maximum GDP, or the costs of policies that should have a per-person cost.

This seems like it wouldn’t matter immediately, since the percentages would be the same, but in theory one should be able to combat debt by increasing the population in a surplus-generating government, thereby increasing the surplus, or vice versa.

Furthermore, population changes do not change the youth demographic, when they come from birth policies and not from immigration. They should have an effect on “youth” and “retired” which should last for about 25 years.

1 Like
2

Changes to GDP - V1.44

You can refer to the post above for a dev answer on ‘population and GDP link’ and related ideas. In short - population affecting GDP, tax revenues, and many others will mean a massive increase in complexity with few player choices. Nonetheless, I agree that it would be nice if there’s another way of fighting debts instead of just taxing more & spending less.

Regarding youth & retired demographics, you may be right but imo the youth isn’t getting much love from devs. At least the retired population will increase as lifespan & health go up though. But, to be fair, I won’t give attention to statistics that will have an impact 100 turn (=100 quarters = 25 years) later. There will be a massive amount of time and its effects should be creeping at unnoticeable speed.

1 Like
3

Yes its true that in theory we should model changing demographics. In a technical sense, its quite easy to write code that shows the age of every citizen, and we could age them each turn, add new ones in line with the birth rate, kill off older ones in line with lifespan, and so on…
The reason I choose not to do this, is because it breaks the consistency of the simulation. In the way the game is coded, nothing is ever created or destroyed, its a single mesh of connected items, whose connections may be on or off at any one time. This allows for a very ‘clean’ implementation of the game, in code terms, and means processing stuff is easy to debug, and modding things is quite easy.

But we could probably achieve all the same results using the existing system. We can certainly adjust membership of the retired group more strongly in line with lifespan changes, and the young voter group can change more sharply in line with pro-parenting policies. AFAIK those effects are already in, but they may be a bit too weak for the player to notice?

(Lifespan has a potential 12% boost to membership of the retired group, for example, and FreeParenting classes boosts the parents group by 3%, but we do not model an 18-years-later boost to young voters. Sadly IIRC we cap out at 32 turns for inertia (8 years)).

4

Just to add, there are another duplicate links regarding Health/Lifespan.

#,Health,PUBLICSERVICES,0.5,0,1,HIGHGOOD,icons_health,#,#,"Retired_freq,-0.2+(0.4*x),8","WorkerProductivity,-0.15+(0.15*x)","ImmigrationDemand,0.07*(x^6)",,,,,,,,,,,,,,,
(...)
#,Lifespan,PUBLICSERVICES,0,0,1,HIGHGOOD,icons_lifespan,#,"_year,-0.1+(0.02*x)","Health,0+(0.65*x),4","Environment,-0.2+(0.4*x),6","Technology,0.0+(0.15*x),8","ViolentCrimeRate,0.0-(0.2*x)",#,"Retired_freq,0.0+(0.12*x)","Population,-0.1+(0.2*x)","HealthcareDemand,0.13*(x^3)",,,,,,,,,,

Health-to-Retired_freq: -0.2+(0.4*x),8
Lifespan-to-Retired_freq: 0.0+(0.12*x)

So, Health has a bigger impact on the membership of the Retired compared to Lifespan, which has been mentioned already.

The retired living long & healthy won’t make any difference to pension costs compared to them living long & unhealthy.

5

Do people get destroyed by abortion in game, because it’s the only policy in game which directly reduces population in game, or is that an insignificant impact?

6

:eyes:

another correction) there ARE 3 policies reducing population

7

What are the other two?

8

One Child Policy & Right to Die

9

Oh, right, forgot about those two.

10

We neither create or destroy voters, as simulated. But the voter group membership of voters changes in response to demographics.
It might seem easier to simulate individual voters, but it doesn’t actually get you anything that changing the group-membership variables does not. It just sounds like it would be easier to code (but actually would be more complex, and far less optimal for processing).

11

So if you drop population to 0%, it makes no difference?

12

Yes, no difference, only displayed numbers change

1 Like
13

Should probably fix this, this combined with there being no positives to population growth (like more tech, for ex) and population not tied to taxation means that there’s no incentive to having more population.

Same problem with private housing, no positive outcomes (generational wealth gap, high earnings from rent, etc.).

14

Hopefully Democracy 5 completely abandons neural network as its too limited. :stuck_out_tongue_winking_eye: :clown_face:

1 Like
15

:slight_smile:
Indeed.

16

I solved this by adding extra links between population and voter group densities, also voter groups and expenses/ incomes.

Also in my overhaul voter group percentages affect expenses, like:
Retired_perc affects:
increase: state pension, state health, benefits (except unemployment), and cc. all expenses which are primary for retired (free bus passes, winter heating, etc…)
decrease: payroll tax for example

Problem was and still is: taxes where have the multypleincome flag - still had not find a workaround when can affect the final numbers over 1.0