Mexico: The Paradox of Public Investment and Fiscal Policy

Mexico today combines a weak economic cycle (which could amplify the multiplier) with rising public debt, despite a lower deficit, according to its own Pre-criteria. It varies with the cycle, the level of regional development, and the state of public finances. But there is a question that this column did not answer, and which is worth asking in all its discomfort: what if public investment also does not yield what is expected of it? The answer, at least for Mexico, is not reassuring. A multiplier below one, which implies some displacement of private activity, and it is consistent with previous evidence for Mexico and emerging economies. And the public investment multiplier is statistically zero. This does not mean that fiscal policy should abandon investment, since the long-term effects on productive capacity are real, although difficult to capture in short horizons, but rather that the evaluation of both instruments should be done with honesty about what they produce and in what timeframe. But when public debt is high, the multiplier collapses: in states with the highest debt burden, the effects become statistically null or negative for government consumption. In less developed states, the effect is greater, consistent with higher marginal propensities to consume and shallower financial markets. It is consistent with the international literature on emerging economies, where the public investment multiplier tends to be lower than in advanced economies for well-documented reasons: lower execution capacity, higher leakage into imports, less complementarity with the private sector, and greater uncertainty about the horizon of projects. It is not a devastating result, but it is not a multiplier greater than one that is usually implicitly assumed when it is stated that public action can be a 'lever' of growth. What is discouraging is the breakdown by type of spending. The channel through which Mexican fiscal policy transmits its effect is not capital formation, but demand. And the third variable (the composition of the adjustment) operates in the least favorable direction: the cut falls on investment, which is the component with the lowest multiplier, in an environment where the mechanisms that would make that investment efficient (agile execution, private complementarity, projects with clear social profitability) face known constraints. The paradox then becomes more acute than the one I described last week. The Infrastructure Investment Plan may have strategic rationality and real long-term effects. But its benefits are of a different nature than what the public discussion usually assumes when it links investment with immediate growth. So what does work, then? Fiscal stimulus is less effective precisely where fiscal pressure is most acute. This pattern has a direct implication for the current moment. Those two forces push in opposite directions. During recessions, the aggregate multiplier rises, which gives support to the counter-cyclical logic. Government consumption (personal services, goods and services) produces an effect of 0.2. That gives us an exogenous source of variation to identify how much growth each additional peso of public spending generates. The aggregate result is 0.8: for each additional peso that the public sector spends, GDP grows 80 cents in the year of impact. It is not just that the spending that matters most for the long term is being compressed. Public spending acts as a short-term stabilizer, not as a builder of productive capacity. This finding is not an anomaly. The evidence points to transfers. They are the instrument with the most short-term traction, they operate mainly through demand, and they have amplified effects in regions and households with higher propensity to consume. In an ongoing research with my co-author, we estimate the fiscal multipliers for the 31 states of the country between 2003 and 2022, exploiting the fact that about 90% of state revenues come from federal transfers whose allocation formula is independent of local economic conditions. They are seen in the pace of budget execution, in the gaps between approved and exercised spending, and in the history of projects that usually start in one administration and are abandoned in the next. The multiplier, moreover, is not a constant. Last week I argued that the fiscal adjustment of the 2027 Pre-criteria has an invisible cost: it suggests a contraction of public investment for 2027 at the same time as it is presented as the visible engine of the cyclical recovery of the economy. It is that the narrative of recovery is built on the component of spending whose short-term effect is the weakest in the evidence. In Mexico, these constraints are not hypothetical. Transfers to households have a multiplier of 0.3.