Job Market Paper

  • "Quantitative Analysis of Climate Heterogeneity via an Unconditional Quantile Vector Error Correction Model"
  • Abstract: This paper introduces a time-series methodology to quantify heterogeneity in the dynamics of the unconditional temperature distribution and its association with climate drivers. By recognizing that temperature unconditional quantiles represent temperatures at different locations—or seasons—, I establish an equivalence between a structural One-Dimensional Energy Balance Model (1D-EBM) and a statistical reduced-form Vector Error Correction Model (VECM) for a range of distributional characteristics of temperature—mean and quantiles—and total radiative forcing, including radiative forcing from anthropogenic greenhouse gases (GHGs). The VECM is estimated employing time-series methods agnostic about the type of trends (stochastic or deterministic) in climate data, and is utilized to produce the following outcomes of practical interest for economic analyses: i) estimation of long-run responses of temperature distribution to changes in GHGs, ii) long-term temperature density forecasts, iii) conditional projections of temperature distribution under different scenarios of future GHGs emissions, iv) extraction of the common-trending component behind the existing warming trend, and v) identification of distributional climate shocks and impulse-response analysis. An empirical study using station-level temperature records (1880–2021) reveals strong climate heterogeneity at global, hemispheric, and continental (Europe) scales, with important potential implications for damage analysis and integrated assessment modeling. A deeper understanding of global warming dynamics is crucial for informing more efficient adaptation and mitigation policies.

    Publications

  • "Trends in Temperature Data: Micro-foundations of Their Nature", (2024) with Lola Gadea and Jesús Gonzalo
    Economics Letters, 244: 111992. doi: 10.1016/j.econlet.2024.111992
    Replication files
  • "Which Crisis Support Fiscal Measures Worked During the Covid-19 Shock in Europe?", (2024) with Evi Pappa and Eugenia Vella
    SERIEs Journal of the Spanish Economics Association, 15: 327-348. doi: 10.1007/s13209-023-00288-w
    Replication files
  • "Heterogeneous Predictive Association of CO2 with Global Warming", (2023) with Liang Chen, Juan J. Dolado, and Jesús Gonzalo
    Economica, 90(360): 1397-1421. doi: 10.1111/ecca.12491
  • Working Papers

  • "On the Effects of Wildfires on Poverty in Bolivia", with Gustavo Canavire and Alejandro Puerta. R&R Journal of Development Economics
  • Abstract This paper examines the impact of severe wildfire events on Bolivia’s poverty and labor market outcomes. We construct a panel dataset from 2005 to 2020 at the municipality level utilizing NASA’s MODIS Collection-6 MCD64A1 burned area product, and merge it with household surveys. To attain survey representativeness at a given geographical level, we aggregate neighboring municipalities through the max-pregion algorithm. Using the Interactive Fixed Effects Counterfactual Estimator, we estimate the causal effects of severe wildfire events on poverty, household per-capita income, and the agricultural sector. We find a significant short-term increase in poverty explained by a temporary decline in household per capita and, specifically, agricultural labor income.

  • "The Heterogeneous Effects of Changes in Precipitation on Poverty and Labor Outcomes in Ecuador", with Gustavo Canavire and Alejandro Puerta
    Abstract This document examines the effect of precipitation shocks on poverty status in Ecuador. Using gridded monthly precipitation data from 2007 to 2021, we define measures for the excess and deficit in precipitation levels at the parish geographical level. This data is merged with household socioeconomic information obtained from the National Survey of Employment, Unemployment, and Underemployment (ENEMDU). Our empirical findings reveal that both excess and deficit in precipitation significantly affect poverty status, with these effects displaying strong heterogeneity across economic sectors. Variations in the Standardized Precipitation Index (SPI), whether positive or negative, lead to an increased probability of poverty among workers in the primary sector (specifically, those engaged in fishing and agriculture). In contrast, we observe poverty-reducing effects for the secondary and tertiary sectors. Factors such as formality status, urban/rural location, and the nature of employment play crucial roles in moderating the estimated effects. Per-capita household income and labor income are key channels to explain our findings.

  • "High-frequency Density Nowcasts of State-Level Carbon Dioxide (CO2) Emissions in the U.S.", with Ignacio Garrón
    Abstract Accurate tracking of anthropogenic Carbon Dioxide (CO2) emissions is essential for the formulation of effective climate policies and meeting long-term international decarbonization commitments. However, data on energy consumption and CO2 emissions are typically released annually with significant publication delays, posing challenges to a timely and informed decision-making process. This paper introduces a panel nowcasting methodology designed to provide timely predictions of the statelevel growth rate of per-capita energy consumption and CO2 emissions in the United States (U.S.). Initially, we estimate a panel mixed-data sampling (MIDAS) model for per-capita energy consumption growth, employing a variety of predictors including quarterly personal income, monthly electricity consumption, and the weekly economic conditions index of Baumeister et al. (2024). Those predictors feature a shorter publication lag with respect to the energy consumption data. In a second stage, a bridge equation that links per-capita CO2 emissions growth with the timely predictions of energy consumption is estimated using panel quantile regression methods. The obtained density nowcasts provide important information about both the expected path of CO2 emissions growth and the uncertainty surrounding the central trajectory. Predictive accuracy is evaluated through a pseudo out-of-sample study spanning 2009 to 2018, simulating the real-time data release schedule. Compared to a simple historical mean benchmark, our findings show that incorporating predictors such as electricity consumption and the weekly economic conditions index significantly enhances the predictive accuracy of per-capita energy consumption growth. These improvements also translate into more accurate predictions for the density of per-capita CO2 emissions growth, compared to a historical quantile benchmark. In a comparative evaluation, the most effective nowcasting model is the one that integrates information from all predictors sampled at mixed frequencies.

    Work in Progress

  • Becoming Green: Aggregate and Firm-level Effects of Green Technology News Shock (with Oscar Jaulin)
  • Abstract We empirically explore the macroeconomic and firm-level implications of news about future technological advancements in the green sector. Utilizing the economic value of green patents granted to publicly listed companies in the U.S., we identify green technology news shocks via a convenient and meaningful rotation of the innovations from a Bayesian Vector Autorregresion Model (BVAR). These shocks are decomposed into two orthogonal components: i) a common technological component shared by both green and non-green innovation, that reproduces response patterns akin to those expected from a technology news shock with long-run impacts on productivity; and ii) an idiosyncratic component to green innovation inducing inflationary pressures and stock price reductions. The responses to the orthogonal component suggest the existence of a green transition news content that can be related to expectations of more rigorous carbon policies or stricter environmental standards in the future. Our focus on green innovation deepens our understanding about the effect of technology-specific news shocks and provides information of practical importance for macroeconomic and environmental policies.

  • From Paleo-Cooling to Paleo-Warming (with Jesús Gonzalo)
  • Abstract A time-series quantitative methodology to describe trends in the temperature distribution under the assumption of local-stationarity is introduced. The methodology consists of estimating certain distributional characteristics of interest (mean, quantiles, and dispersion measures), testing for the existence of trends, and characterizing heterogeneity in the trend evolution along the temperature distribution. Applying the proposed methodology to the temperature paleoclimate record over the last 800 thousand years, we are able to determine to what extent the currently observed Global Warming dynamics are unusual compared to past changes in Earth’s climate. Our findings indicate that the Global Warming during the last 200 years is completely different to what is observed over the paleoclimate record. Over long-horizon samples covering several glacial-interglacial cycles (800 and 420 thousands of years), our approach detects a clear long-term Paleo-Cooling. If the analysis is restricted to shorter-horizon samples characterized by abrupt increases in temperature, as the last deglaciation (last 27 thousands of years), evidence of Paleo-Warming is obtained; however the intensity and heterogeneity of the current warming are significantly different. We highlight the importance of studying the whole distribution of the temperature to obtain a wide angle picture of climate evolution.