Work in progress

Minimum Wages and Informality: Who Bears the Welfare Cost in a Small Open Economy? with A. Grajales, J. Ospina and M. Ramos. 

Abstract: Colombia's 2026 minimum wage decree raised the nominal floor by 23%---the largest single adjustment in recent history---implying an 18\% real increase. We analyze the macroeconomic consequences of this policy using a two-agent New Keynesian small open economy (TANK-SOE) model with sticky prices, active monetary policy, and a formal--informal labor margin, calibrated to Colombia. Formal low-skilled employment falls 19% on impact and 22% in the long run, raising the informality share by 5 percentage points. The resulting cost-push inflation (0.8,pp annualized) triggers a transitory monetary tightening (0.7,pp) that dissipates within eight quarters. Low-skilled consumption rises 1.5% on impact but subsequently falls 1.5% below the pre-shock level; aggregate GDP contracts permanently by 0.82%. Welfare losses, measured by consumption-equivalent variation, are larger for the low-income household, revealing that the policy fails its redistributive goal in general equilibrium. Results are robust to a representative-agent and a closed-economy specification.

Oil shocks and Informality with O. Jaulín, and M. Salazar

Abstract: We explore how oil-price shocks affect informality in a small open economy. The analysis proceeds in two steps. We first provide empirical evidence of the causal effect of oil-price shocks on informality, showing that a positive shock increases the informality rate. We then rationalize this evidence with a small open economy model featuring tradable and non-tradable sectors. Consistent with the data, informality concentrates in non-tradable services; accordingly, the model assumes that the tradable sector employs only formal labor while the non-tradable sector also employs informal workers. The tradable sector further comprises export and import sub-sectors. A positive terms of trade shock expands the former and contracts the latter; the resulting income effect raises demand for non-tradable goods, boosting demand for informal workers and raising the informality rate. Consistent with the empirical evidence.

Filtering Through the Storm.  Comparing Volatility-Treatment Methods for DSGE Filtration and Estimation with F. Castañeda, O. Jaulín, and J. Rincón. 

Abstract: COVID-19 exposed a gap in standard DSGE practice: the constant-variance, mean-reverting shock processes these models rely on have no principled way to absorb a handful of quarters an order of magnitude more volatile than anything in the estimation sample, and practitioners need a general treatment strategy for extreme-volatility episodes, not an ad-hoc pandemic patch. We compare four such treatments, no treatment, missing real-side data, heteroskedastic AR(1) shocks, and heteroskedastic ARMA(1,1) shocks, on CUMBIA, Banco de la República's estimated small-open-economy DSGE model for Colombia, across Bayesian parameter estimates, impulse responses, historical shock decompositions, and out-of-sample forecast accuracy. CUMBIA's transmission mechanism proves robust to the choice of treatment, but the treatments disagree sharply on what a crisis quarter's variance represents and how long its effects should persist. The heteroskedastic AR(1) treatment is the most accurate post-COVID forecaster on every variable and horizon we evaluate, while the ARMA(1,1) treatment is easier to implement without error and easier to communicate to a policy committee. We recommend the ARMA(1,1) specification as a production default, with the AR(1) treatment reported alongside it as a supplementary, scarring-robust check in the years following any future extreme-volatility episode.