Thanks for the thoughtful response! I’ll try to reply to all the points you raised.
1. The argument “in a free market the overall wealth in the system increases, so even though inequality ramps up everyone ends up with more than what they initially held” is kind of flawed. It is flawed because it’s based on three flawed hypotheses. The first is that the idea that the overall wealth in the Western world has been increasing since the end of WWII because of free markets doesn’t have enough evidence in its support because there were many other factors at play. Among those, technological advancements that didn’t necessarily happen because of neoliberal policies (China surely can’t be labelled as a neoliberal country, and yet it has produced more technological advancements than the Western world over the last 10 years), and Keynesian policies based on government and central banks intervention to support aggregate demand (look no further than the New Deal and the Marshall plans) have arguably contributed to the sustained growth of the past 70 years more than Friedman’s cynical monetarism. What some neoliberal economists often forget is that correlation doesn’t mean causation. The second flawed hypothesis is that growth will happen naturally in a free market, and the average wealth increases as the economy grows. What really increases in a laissez-faire market is the “temperature” of the system, defined as ratio between traded/perceived wealth vs. real wealth: growth increases linearly with the temperature of the system until some point, but after some threshold nothing can prevent the crash. The past 90 years have been struck again and again by financial crises caused by lack of regulation within certain areas, and the world has recovered from all of them thanks to better regulations and redistributive policies, not thanks to the market itself. The third argument is that in a free market everyone can contribute to the system according to their real skills and potential: the opposite is true in many cases. In a mechanism that drifts towards growing inequality by design those left behind will be less and less able to contribute to the system because the barriers for them to participate will increase. It’ll be harder for them to receive proper education, harder for them to land in good jobs, even if they have plenty of potential, and therefore harder for them to participate to the market as much as they could, and that hurts aggregate demand on the long run. Sure, they can still buy things that were unthinkable for poor generations 60 years ago, but in such an active market they could buy more if they had more — also because concentrating more wealth in the hands of fewer doesn’t guarantee that they will spend more, indeed the opposite is often true. Those are wasted opportunities for further growth. And when we look at the increasing numbers of people living in absolute or relative poverty, it’s inevitable to ask ourselves whether the benefits of growth due to lack of regulation are really worth the relative loss of purchasing power of a growing slice of the population.
2. A degree of wealth inequality is of course acceptable. Economics should be a trade-off between merit and fairness, and history has proved that systems that value too much one of them over the other aren’t sustainable on the long term.
3. What you showed in your numerical argument is a clear example of “high-gain gamble”. If on a coin toss you either double of half your wealth then you’re in a scenario where you’re likely to win more and more, because what you get when you win is much higher than what you lose otherwise (+100% >> -50%). In the real world the margins between wins and losses in a trade are actually much lower. As an exercise, try to tweak your win/loss percentages (e.g. reduce the win fraction or increase the loss fraction) and see how the “profitability” changes when you increase the number of games N. There are many interesting insights that can be found from this type of analysis. In my simulation I do indeed trade only a fraction of the agents’ wealth every time, not the whole wealth, but the outcome won’t be very different even if I took the whole wealth into account at every turn — the system would simply converge earlier towards inequality.
4. True, a free market works (at least according to the theory) only if every agent acts rationally and tries to maximize the owned wealth, but reality has shown us already that this is a very rough and almost utopian approximation — comparable to the utopian approximation of “the State will provide everyone with what they need” of the communist ideology. Crises have occurred again and again because of irrational behaviour of the agents in the market - simply because a system where all agents act rationally is a system where every agent is fully informed, and that’s a very rough approximation. A real free market is more like poker and less like chess. And the fact that the ratio of wealth redistributed into the market decreases when the wealth of a small set of agents increases has already been widely proved. Simply because as the wealth of a small set of agents increases their decisional and lobbying power will also increase, and they’re unlikely to take decisions that harm their own wealth. It’s also true that in a real market agents don’t trade randomly, but a proper real-world simulation should be half way between these two approaches. The case of agents trading rationally to optimize a specific metric would converge towards a point of Nash equilibrium, and it would experience an overall decrease in the number of trades made over time as we approach that point (something that we aren’t actually experiencing in the real world). The case of agents trading randomly inevitably ends up with increasing inequality instead. A new model can indeed take into account aspects of both the scenarios, but still I don’t expect it to converge on the long run towards any outcome other than increasing inequality. Simply because the tendencies of the system to increase its entropy will eventually take over the forces that would push it towards a point of Nash equilibrium.