What the simulation is worth today

What is realistic

• The mechanics. A hundred years, 213 countries, shocks, elections and defaults included, with no aberrant value, accounts correct to the cent every day, and two runs with the same seed identical to the bit (§7.1).
• The responses to modest decisions, of one or two points, fiscal as well as monetary. Their direction, their size and now their timing are those of the studies (§7.2).
• The world moves. Each country varies as much as in the data in ordinary times, with as many years of recession, and countries move together. Pandemics, El Niño and disasters strike at random, and cost what their sources measure. Governments fall at elections about one time in two, on the dates of their own calendar. Some countries default, through the chance of a confidence crisis or at their limit.
• Excesses have a price. Held by a player who would never lose an election, a 10% policy rate takes France to default in August 2028 (debt at 148% of GDP), then again in 2033 and in 2038, when the government falls. Ten extra points of spending lead to default in November 2028, then in 2036. The absurd budget (revenue 10%, spending 80%) does so as early as August 2026, then in 2031 and in 2036. These last two governments survive their defaults, which bring down those in power only one time in two (E16).
• Politics can be played. Purchasing power links taxes and benefits to popularity, the electoral calendar rewards a well-timed stimulus, and debt punishes a stimulus that is never withdrawn (§7.4).

What is not

1. Advanced countries are still a little too calm. Even with world events and the common share of the shocks, France has a reason to act in only 58% of years, and 49% outside elections; the United States in 65% and 54% (without the events or the common share of the shocks, 51% and 55%). The major disasters mostly strike small, exposed countries, such as Taiwan, the Philippines or the islands, not France or the United States. Above all, they lack trade, which spreads the crises of other countries.
2. Inflation remains easy to tolerate in advanced countries. A 0% policy rate held for twenty years raises French inflation by only 0.6 points, and credibility only falls to 0.67. The Phillips curve is flat there (0.09), and the rate lowers prices by only 0.065% after five years per point, half as much as the corrected studies. The exchange-rate channel is missing.
3. High inflations end too slowly. Without shocks, Argentina, starting from 118%, is still at 56% after ten years. Turkey, Iran and Nigeria, starting from 35 to 44%, are still around 18% after twenty years, for a target of 10%. Yet episodes above 100% last 40 months on average (Fischer, Sahay and Végh 2002), and even “moderate” inflations, of 15 to 30%, most often last three years. Only 6 episodes out of 55 go beyond five years (Dornbusch and Fischer 1993). They end with a change of regime (fiscal contraction, an exchange-rate commitment, re-anchoring of expectations) that the model does not have. Credibility lost at the start is only regained in small steps. Quarter after quarter, inflation falls much more slowly than the credible regime predicts, and the signal of E17 keeps siding with the regime that is not credible.
4. Too few defaults. 0.17% per country per year, several times less than history. In the model, an emerging country defaults at 139% of GDP at the median (from 128 to 245%). At the median, its risk starts 67 points above its starting debt, because its limit is derived from that of advanced countries. Yet more than half of the defaults of middle-income countries from 1970 to 2001 happened at an external debt below 60% of GNP (Reinhart, Rogoff and Savastano 2003). On top of this come foreign-currency debt, currency crises and commodity shocks, which the model does not have.
5. The big stimulus before the vote wins the coming election (§7.4). It loses the 2027 election only three times out of forty. It no longer dominates since cohabitation. The default almost always follows, cohabitation two times in three, and it survives no more often than the others. But in France the default no longer brings down the president (E22). What it costs the player is a lost assembly, not the game.
6. Advanced countries led by a populist contract too often. Without a player, France contracts in 36% of years, Germany in 41% and Italy in 47%, while the data count 5, 15 and 25% (2000-2019). This is mostly the growth cost of a populist leader (E24, populist_growth_loss). It takes one point of potential growth away every year from countries that grow by 0.7 to 1.1%. With this cost removed for the measurement, France contracts in only 17% of years, Germany in 29% and Italy in 27%. Funke, Schularick and Trebesch measure one point on average over their episodes. Whether it holds for a country that grows by 0.8% requires rereading their subsamples. This cost never affects the player's country, which the model does not govern. Settled. The direct loss is removed, and populism costs through its policies.
7. Elections are mostly down to the draw. A well-timed stimulus gains about one point of the vote, and the standard deviation of the draw is worth three. This is realistic, since the economy explains only part of the vote, but a player may well experience it as chance.
8. Too many small quarterly recessions. Counted in years, they have the right frequency. Dated quarter by quarter, they are three times too frequent in advanced countries (§7.2, E13).

The current limits

• Data. IMF data cannot be redistributed without permission (§6), and starting inflation depends on them too. 155 countries start from the policy rate given by the rule, for lack of a source. Wikidata gives no election for Denmark, and the terms that follow an early election are short (Germany, Korea). The starting output gap is zero everywhere. Social benefits are a share of spending by group of countries, not by country. 38 territories are outside the economy, and 2 countries are out of domain.
• Theory. All the unknowns in the list of unknowns: the exchange rate, foreign-currency debt, who holds the debt, the number of terms of the incumbent camp… The debt limit of emerging countries and the 20-point risk zone are choices; the monetary transmission delay is a calculation, not an estimate. The voting equation is American. The neutral rate of emerging countries is extrapolated from that of advanced countries, so India, starting from its observed rate of 6.5%, climbs towards 9.6%. The step from the frequency of price changes to the slope of the Phillips curve goes through the Calvo model, which is a choice. Tax multipliers fall into three irreconcilable families of results: 0.25 in VARs and models, and 2 to 3 at the peak with the narrative method (Romer and Romer 2010), whose inference is disputed (Mertens and Ravn 2013; Jentsch and Lunsford 2019). The model takes the low value. The studies contradict each other on the strength of monetary policy in emerging countries (weaker in Enzinger et al. 2025, stronger in Brandão-Marques et al. 2020), so the model keeps it the same as in advanced countries.
• Events (E18). The loss from a cyclone fits into one year and its rebound (IMF), whereas Hsiang and Jina see it growing for fifteen years. Exposure to earthquakes, cyclones and droughts is estimated by a method of the studio (the footprints of past events, the threshold of 5% of the population), for lack of an index that can be redistributed, and every country is equally exposed to floods and epidemics. The eruption and the solar storm rest on choices.
• Missing mechanisms. Trade, the exchange rate, credit ratings, hysteresis, the detail of taxes and spending, public investment. The player still has no lever on their country's long-run growth, and no way to re-anchor expectations at a stroke.

What it takes to go further

Two levers are missing for the player more than anything else: a stabilisation plan, which would restore credibility at a stroke on condition that the budget is put in order, and public investment, which would give the player a hold on the long run. And the world of advanced countries needs more shaking: trade and the exchange rate, which will make countries depend on one another and bring the currency crises that defaults are missing.