We develop indexes of economic policy uncertainty (EPU) for Latin American countries.
First of all, we estimate the EPU for the six largest Latin American economies: Argentina, Brazil, Chile, Colombia, Mexico and Peru, as we have enough newspapers for all of them. More details about our methodology can be found in these two recent papers: Ghirelli, Pérez y Urtasun (2021)
and Andres-Escayola, Ghirelli, Molina, Pérez y Vidal (2022) 
(2 MB).
For each of the countries we construct three types of EPU index, using the largest set of available newspapers with different levels of proximity to the country of interest
- an index based on local newspapers (in Spanish or Portuguese for the case of Brazil);
- an index based on foreign newspapers (both Anglophone and Spanish sources);
- an index based on all of the aforementioned press, weighting equally each source irrespective of the level of proximity to the country of interest.
To construct the EPU index, we follow Baker, Bloom y Davis (2016) and similar papers complied in their webpage Policy Uncertainty website. Below we post the keywords used for the Anglo-Saxon press:
- Uncertainty: uncertain, uncertainty/uncertainties, unstable, instability/instabilities, risk(s).
- Economy: economic(s), economy.
- Policy: the name of the central bank of the country, the name of the government's workplace in the country, parliament, government, federal reserve, treasury, tariff(s), deficit(s), budget(s), public spending, debt ceiling, exchange rate(s), currency crash(es), sovereign debt, public debt, fiscal policy/policies, monetary policy/policies, legislation, reform(s), tax(es), overhaul, rule(s), norm(s), normative, regulation(s), law(s).
(for the words in Spanish and Portuguese please see Andres-Escayola, Ghirelli, Molina, Pérez y Vidal (2022) (2 MB).)
We rely on Factiva's indexation to exclude all articles related to sports and leisure and to select articles about the country of interest.
The construction of the indexes is as follows. First, for each newspaper we count the number of articles published in a given month that contain words in the aforementioned blocks of keywords. Second, we express this count as a proportion of articles published by the newspaper in the month. When dealing with the local press, we divide this count by the total number of published articles that month, whereas in the foreign case we use the number of articles published that month that are about the country of interest. Third, we standardize each monthly series of scaled counts by dividing it by its standard deviation. Fourth, for each country we average the newspaper-based standardized series across newspapers to compute an aggregated index. Fifth, we rescale the resulting index to mean 100 to obtain a set of homogeneous country-based EPU indicators.
These indexes will be regularly updated in this webpage, and can be downloaded from this link 
(101 KB).
For each country, we consider the most-read local generalist newspapers available in Factiva:
- Argentina: Clarín, La Nación, Infobae.
- Brasil: O Globo, Folha de Sao Paulo, O Estado de Sao Paulo, Correio Braziliense, Estado de Minas, Agencia Brasil.
- Chile: El Mercurio, La Tercera, El Diario Financiero, Pulso.
- Colombia: El Espectador, El Nuevo Siglo, Portafolio.
- México: Reforma, El Universal, La Jornada, El Financiero, El Economista, Agencia Mexicana de noticias.
- Perú: El Comercio, La Republica, Gestión, Agencia Peruana de noticias Andina
For Brazil, we process articles written in Portuguese, while for all other countries we process articles in Spanish.
The EPU based on local press are available as from 2002m11 to present.
We select the following types of foreign press:
- Anglo Saxon press: Los Angeles Times, The Boston Globe, The Globe and Mail, The New York Times, The Telegraph U.K., The Times U.K., Chicago Tribune, The Guardian U.K., The Wall Street Journal, The Washington Post, The Economist.
- Spanish press: El País, El Mundo, Expansión, ABC, Cinco Días, El Economista, La Vanguardia.
All EPUs based on foreign press are available as from 1997m1 to the present.
The EPU index based on all press are constructed by pooling together all available sources (i.e. local and foreign newspapers), that is we pool together the local, Spanish and Anglophone press. All EPUs based on all press are available as from 1997m1 to the present.
EPU indices have also been estimated for the Central American region, which includes Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama and the Dominican Republic. In this case, country-specific EPU indices are constructed using a combination of local and regional sources and validated using the narrative approach to ensure that changes accurately reflect relevant policy developments. Moreover, we show using a BVAR model that a shock to our indices translates into a decline in economic activity. Details of the methodology can be found in a recent Bank of Spain Occasional Paper (Diakonova, Ghirelli and Quiñónez (2024) (681 KB)). These indices will be updated regularly on this website, and can be downloaded at this link (34 KB)
