1. To participate please register first

2. Accept Terms & Conditions on first use

3. Staff must select Single Sign On to login

This Proposal is hidden. Change Visibility
Cross-matching Data to Uncover Corruption in Colombia
Dec 19 2019

Corruption in Colombia, present at all levels of government, is abetted by financial and interest disclosure systems that are not effectively detecting and preventing graft and conflicts of interest. We propose designing an innovative data cross-referencing digital solution to improve substantially the use of Colombia’s existing public sector disclosure systems and information managed by Colombia’s Public Servants Department (Departamento Administrativo de la Función Pública - DAFP). Focused on national legislative and executive power senior officials (hereafter 'senior public officers'), this digital solution would cross-match two existing databases containing senior public officers’ asset and special interest declarations, one of which became available only in November 2019, to automatically detect potential conflicts of interest and improve the oversight functions of Colombia’s Public Servants Department. Additionally, the information generated by this digital solution would then be tallied with other separate government systems that store data on political campaign finance and public contracting to enhance its coverage and reliability. 

The objectives of this digital solution are: 

- Significantly strengthen corruption surveillance in Colombia. The solution, in a form of a prototype, would use inexpensive software and cloud hardware components, as well explore the utilization of Machine Learning, to automate, through a process of cross-referencing, the detection of noteworthy changes in wealth and other suspicious patterns of the senior public officers, including inconsistencies as they appear across asset and interest disclosures, campaign finances and public tendering databases.

- Serve as a valuable corruption indicator for the IMF. The solution would leverage different government data storage systems that at present don't exchange and validate information among them. It would produce robust empirical information providing the Fund with a metrics-based way to quantify and track corruption tendency in Colombia. This approach, including the prototype that would be developed under this proposal, is potentially scalable to other countries with similar untapped scope for applying data mining and crossing asset and conflict of interest declarations data with other disclosure sources. 

There are reasons to focus specifically on senior public officers’ financial and interest disclosure systems. One is a recently modified law under which, specifically, national congressman and executive power senior officials’ special interest declarations must be made public. The advent since November last year year of this fresh untapped dataset, complementing the existing SIGEP managed system (please see our "Problem Statement" for an explanation of both systems), makes timely the task of gathering and tallying legislators and public officers' wealth and interest disclosures. Focusing our analysis to senior public officers also makes the task of obtaining declarations from SIGEP, available on request only, manageable, and allows the development of the prototype proposed in a period of between 8-10 months. Furthermore, this approach makes sense from the standpoint of promoting media and citizens attention with interest in this specific segment of public servants and congressman.

In sum, the digital solution proposed l would have two main interconnected functions:

1. It would automatically cross-reference and verify Colombia's national legislators' and public officers' interest and financial disclosures - currently obtainable via two sources, SIGEP and Congress' Official Gazette - to detect suspicious patterns. This data itself would be repackaged and displayed in far more user friendly ways than is currently the case to promote social accountability and citizen participation.

2. It would then cross-check this information with two other public sector information systems. The first is the Project OCEANO. This is an acclaimed anti-corruption initiative adopted by the Office of the Comptroller General that maps public contracting data in Colombia (and which further links to an older related platform called Colombia Compra Eficiente). The second information system, Cuentas Claras (Clear Accounts), is the database on political campaigns finance hosted by the National Civil Registry. Subsequently the solution would be able to widen the process further to include data stored by the National Tax and Customs Office, the Financial Intelligence Unit, Superintendencies, and financial and social security entities.

Currently Colombia presents growing cross-sectoral interest in advancing - innovatively - in anti-corruption initiatives. The country has already shown itself receptive to technology, and - in this context - Colombia is among a minority of countries that insists on 'electronic only' financial and interest disclosure submissions from declarants. Furthermore, a bill has just been approved in the Congress and is likely to be enacted in the coming weeks that would make all SIGEP declarations publicly available. For its part, Colombia's Chamber of Representatives recently approved an 'Open Parliament Plan' setting out a series of core commitments, including one to improve public online access to congressman financial and interest disclosures. These developments and the climate overall suggest there would  be strong backing for this proposal including from across government.

The specifications of the solution would be customized to Colombia's normative framework. The solution proposed could feature an easy-to-maintain updatable register of declarants as well as a searchable database of disclosures. Up-front and maintenance costs could be kept reasonably low by using open source technology and cloud infrastructure. 

The  solution proposed is highly replicable. Like Colombia, there are many countries in and beyond Latin America that, despite having weak or poorly implemented normative frameworks, possess multiple financial and interest disclosure systems which aren't being validated and referenced with each other to detect corruption. While since variations in those legal frameworks mean there is no generic digital solution that can be used to create a one-size-fits-all model for different countries, the necessary tailoring from one to the other need not be onerous. If we were aided in piloting this solution and, of course, subject to a positive evaluation, we would wish to support the IMF and country authorities in exploring its rollout.

Expertise Required



No evaluations ready
Recorded Outcomes
Record an Outcome