Colombia Atlas of Economic Complexity

distant cityscape of Medellin, ColombiaIn Colombia, income gaps between regions are huge and have been growing: new job opportunities are increasingly concentrated in the metropolitan areas of Bogotá, Medellín and Cali, as well as a few places where oil and other minerals are extracted. The average income of residents of Bogotá is four times that of Colombians living in the 12 poorest departments.

In a country as heterogeneous and regionally disconnected as Colombia, it is unrealistic to expect that regional gaps can be resolved by implementing the same policies in all departments and municipalities. In the absence of effective regional strategies, large cities will continue to develop new and increasingly sophisticated production activities, while most lower-income regions will continue to produce a few unsophisticated products, with their labor force working mostly in construction and very low productivity services.

Graphic button


July 2014 - April 2018


Bancóldex and Fundación Mario Santo Domingo

More about the Research

The main contribution of Datlas Colombia to the body of economic statistics will be a new set of economic variables of productivity and employment by department and municipality, using annual data since 2008. These economic variables will be derived from official information from the Integrated Account of Labor Contributions (Planilla Integrada de Aportes Laborales - PILA) of the Ministry of Health, detailed international trade data (from DIAN) and data on industrial activity (from DANE). The consistent and integrated combination of these data sources will allow for visualizations by department, municipality and metropolitan area (combining several municipalities) of economic variables such as:

Employment and wages by sector of economic activity (CIIU at 4-digits)

  • Number and average size of productive establishments by sector of economic activity
  • Exports and imports by product (International Harmonized System at 4 digits)
  • Exports by product and by country of destination for each location studied.

Datlas Colomiba will also use information on vacancies obtained through major Internet job sites, both official and private, which will allow us to show labor and employment variables by economic sector both nationwide and for some of the largest cities for 2014, such as:

  • Occupational structure, according to the international classification of occupations ONET
  • Wages by occupation
  • Key skills required by occupational groups.

Apart from these new economic variables developed exclusively for the Atlas, we will also include key official statistics produced by DANE and the National Planning Department by department (d) or municipality (m) on population (d, m), gross domestic product (d) and fiscal outcomes (d, m), as well as competitiveness indicators (d) produced by the Private Council on Competitiveness, for the years they become available.

In addition to these descriptive statistics, Datlas Colombia will present a set of economic complexity indicators that will allow analysts informed on the concepts of complexity to perform specialized diagnostics of productive alternatives, export potential and alternatives for employment generation, which will serve as a basis for proposing strategies for local productive development. These indicators include, by department and "city" (metropolitan area or urban municipality with more than 50,000 inhabitants):

  • Product diversity
  • Export diversity
  • Product complexity
  • Export complexity
  • Potential gains of product complexity
  • Potential gains of export complexity

Datlas Colombia will also include indicators of export complexity by product, as well as productive complexity and occupational complexity by sector.

To enable users to easily manage the abundant and detailed information of Datlas Colombia, it will offer intuitive and flexible forms of visualization that will allow for navigation from one variable to another and for the creation of comparisons between departments, municipalities and cities, both for any given year and over time (from 2008).

Aside from the graphics for the economic variables and indicators mentioned, Datlas Colombia will also include interactive graphics that concisely demonstrate the similarities between products, sectors, occupations and cities, which are derived from economic complexity indicators. These graphics would be very useful for growth diagnostics and policy design. Our team is exploring the following options (which will appear in the Atlas to the extent permitted by the quality and consistency of the data and the visualization methods):

  • The product space for exports (built upon world trade data) and the location of departments or cities within that space (similar to the "Product Space" of  the international Atlas of Economic Complexity)
  • Graphics of similarities between productive sectors (developed using co-location, and/or labor transitions and/or occupational sector similarity), which will allow for comparisons of a city or department with national and international patterns
  • Graphics of similarities between cities (built from the similarity of productive structures of cities and/or the similarity of occupational structures of cities), which will allow for comparisons of one city with other similar Colombian cities
  • Graphics of occupations (built from the joint sectoral use of occupations or other criteria), which will allow for comparing characteristics of different occupations and occupational structures of sectors and/or cities.

Datlas Colombia aspires to be useful not only for policymakers but also for specialized economic analysts. While some users might wish to have more economic variables than those contained in Datlas, it was necessary to leave out many data sources, both public and private, that are characterized as sample data or for which coverage does not match the criteria of regional and/or sectoral aggregation used in Datlas.

In the process of developing Datlas we have maintained the highest standards of statistical consistency and privacy of information, in accordance with international statistical practices and legal standards in Colombia. We are grateful to the many public and private organizations that have generously collaborated with us on the same objectives.


Juan Tellez

Juan Tellez

Former Research Fellow
Vice President, Algorithmic & Automation Risk, Bank of Montreal
  • «
  • 2 of 2