How Brazil’s indexed minimum wage allows inflation to grease the wheels of the labour market

Many economies index their minimum wage to inflation to preserve its real value during inflationary periods. This column shows, however, that the timing and predictability of the policy matter for how inflationary shocks propagate. In Brazil, minimum wage workers are less likely to receive a nominal wage increase between indexation events, which leads to lower real wages for these workers. However, both firms and workers anticipate the future increase in the nominal minimum wage, which lowers output in the future and in the present, dampening inflation and raising real wages.

Wage indexation is pervasive. This is true explicitly, such as in Belgium, and implicitly, such as in Spain through collective bargaining agreements. The purpose of indexation is to preserve the real wage (Gray 1976, Fischer 1977). Thus, it acts against the ability of inflation to grease the wheels of the labour market, i.e. to lower the real wage in the presence of nominal rigidities (Tobin 1972, Card and Hyslop 1997). However, in practice, indexation does not occur in real time with inflationary shocks. Instead, it happens on a pre-determined schedule and is pre-announced. These two features of how indexation is implemented affect the scop for inflation to lower the real wage.

The indexed minimum wage provides an opportunity to study how indexation shapes the propagation of inflation shocks. In a recent paper (Colombi and Di Tella 2025), my co-author and I examine this question in the context of Brazil between 2015-2017. During this period, the minimum wage was updated each January based on the inflation rate over the previous 12 months. In January of 2016 it rose by 11%, and in January of 2017 it rose by 6.5%.

Our analysis reveals two key findings about wage setting in this context. First, we show that minimum wage workers are less likely to experience a nominal wage increase between February and November, suggesting that their real wages are less downwardly rigid. Second, we find evidence of anticipation, since workers who are not bound by the current minimum wage, but earn less than the upcoming one, also experienced fewer month-to-month nominal wage increases.

In a New Keynesian framework, periodic and predictable indexation of the minimum wage has countervailing effects on the ability of inflation to lower the real wage. On one hand, the lower frequency of nominal wage increases in between indexation events directly amplifies its ability by lowering the rigidity of real wages. On the other hand, the indexation event itself raises marginal costs and lowers output. The expectation that this will happen lowers output and inflation in the present through intertemporal substitution. This indirectly dampens inflation’s ability to grease the wheels of the labour market by dampening the inflationary shock itself.

In between indexation events, the real wages of minimum wage workers are less downwardly rigid

We start by plotting the likelihood that different worker types experience a nominal wage increase each month (Figure 1). Those earning more than the upcoming minimum wage are classified as unexposed workers (green and blue lines), while those earning less are classified as exposed workers (red and orange lines). Firms are categorised as exposed (solid lines) if they employ at least one exposed worker. We proxy nominal wage increases with sustained earnings increases, defined as any monthly earnings increase that is at least maintained in the following month.

Figure 1 Likelihood of experiencing a nominal wage increase in each month for workers differently exposed to the January 2017 minimum wage increase

Note: Exposed workers are those that earn less than the following year’s minimum wage, we call them “Binding MW” (orange line) and “Non-binding MW” (red line) workers. Unexposed workers are those that earn more than the upcoming minimum wage, we call them “Mid-wage” (green line) and “High-wage” (blue line) workers. Firms are exposed if they employ any exposed workers.
Source: Author’s calculations from RAIS.

Exposed workers are less likely to experience nominal wage increases than unexposed workers (13% compared to 19%, on average). The less nominal wages rise, the less they keep up with inflation and the more downwardly flexible the real wage is. The descriptive evidence suggests that minimum wage exposure, rather than other worker or firm characteristics, drives this difference. Mid-wage and high-wage workers exhibit similar wage-setting patterns despite vast differences in their earnings, and despite being employed by different firm types.

The primary concern is that a worker’s relative position in the earnings distribution, rather than minimum wage exposure, explains the differential wage-setting behaviour. To address this, we estimate a state-level regression that compares workers in the same state-specific earnings decile across states where the minimum wage is more versus less binding. Figure 2a, which pools the data between February and November, presents two key findings. First, workers in lower earnings deciles are 9 percentage points less likely to experience a nominal wage increase when the minimum wage is one standard deviation more binding. Second, higher earners, who serve as a placebo group since they are unaffected by the policy, show no significant cross-state differences in wage setting behaviour.

Figure 2 Workers at lower earnings deciles in states that are more exposed to the current minimum wage are less likely to experience a nominal wage increase between February and November, but more likely to experience one between December and January

Note: The figures plot the effect that state-level minimum wage exposure has on the likelihood that workers in each earnings decile experience a nominal wage increase. State-level exposure is captured by the Kaitz index, defined as the difference between the state-specific median wage and the national minimum wage. The sample covers private sector workers from 2015-2016.
Source: Author’s calculations from RAIS.

Wage setting anticipates the indexation event

To establish that wage setting anticipates the indexation event, we highlight that non-binding minimum wage workers (red line in Figure 1) and resemble binding minimum wage workers (orange line) more closely than unexposed workers. Non-binding minimum wage workers are informative because they earn less than the upcoming minimum wage but more than the current one. Thus, the only reason their behaviour should differ from that of unexposed workers is their anticipated exposure to the future minimum wage.

In a firm-level event study framework, we find supportive evidence that wage setting anticipates the indexation event. Leading up to the indexation event, the firm-level average of workers’ earnings is falling in firms that will be more exposed to the upcoming minimum wage increase relative to firms that will be less exposed (Figure 3). The trend persists even after controlling for natural alternative explanations, such as time-varying region and industry-specific fixed effects (Panel 3a) or past exposure (Panel 3b).

Figure 3 Leading up to the indexation event, average nominal earnings grow slower in firms that will be more exposed relative to firms that will be less exposed

Note: The figure plots the event-study coefficients describing how the difference in the firm-level average earnings across firms that are more versus less exposed to the January 2017 minimum wage evolves over time. The sample covers private sector workers from January 2016-December 2017.
Source: Author’s calculations from RAIS.

Conclusion

We show that the time-dependent and foreseeable nature of Brazil’s indexed minimum wage makes the real wages of exposed workers more flexible downwards. Specifically, they are less likely to receive nominal wage increases between indexation events. For now, we remain agnostic about why this happens. However, recent evidence on the cost workers’ pay to preserve their real wage is consistent with our findings (Guerreiro et al. 2025). Minimum-wage workers who anticipate a costless nominal increase in January may choose to wait for that adjustment rather than undertake costly efforts to raise their wages beforehand.

Overall, the effect of indexing the minimum wage on the ability of inflation to lower the real wage, and therefore ‘grease the wheels of the labour market’, is ambiguous. The policy amplifies inflation’s grease by introducing greater real wage flexibility for minimum wage workers throughout the year, but dampens it by raising future marginal costs in a foreseeable manner. In a New Keynesian framework, anticipated future marginal cost shocks lower demand and therefore inflation in the present via intertemporal substitution, ultimately raising the real wage. Our research illustrates how institutions mediate the propagation of inflation shocks, sometimes in unexpected ways.

Source : VOXeu