ANOVA Policy Research · Independent research since 2016

Rigorous research, from field data to policy evidence.

ANOVA is a quantitative policy research firm working across Latin America. We produce rigorous evidence for anyone who has a decision to make: national and local governments, multilateral institutions, companies and civil society organizations. We design the data collection tools each question calls for, build our own estimates where official figures are missing or unreliable, and evaluate public programmes.

Read our research

Capabilities brochure (PDF)

Evidence-based thinking

Numbers from our recent research

249,268

Venezuelan adults had a concrete plan to emigrate in mid-2026. They carry more human capital than the average adult: 12 years of schooling on average, against 10.5 for the adult population as a whole.

Estimated from household survey microdata. Policy Brief, October 2026

3,466

buildings in coastal La Guaira, 22.6% of the 15,339 in the affected area, were left with severe damage by the June 2026 earthquakes, and another 1,720 (11.2%) were destroyed.

Field verification survey of 1,000 buildings, EMS-98 scale. Policy Brief, September 2026

494,763

people are out of work because their households receive water or electricity precariously, the case for 70.2% of Venezuelan households. The output they no longer produce comes to about US$1.6 billion, 1.4% of GDP.

Estimated from household survey microdata. Policy Brief, September 2026

Practice areas

What we do

Applied econometrics and impact evaluation

Impact evaluation, quasi-experimental designs and causal inference applied to public programmes. We estimate what a programme changed and for whom, using administrative records and our own surveys.

Big data, machine learning and computational modelling

Satellite imagery, small-area estimation and predictive models for places surveys do not reach. We use them to map poverty and disaster damage where official data are missing or too coarse to use.

Poverty, labour and human capital

Poverty, labour markets, education and income distribution, measured from household microdata. We build our own estimates when official figures are missing or out of date, and follow them over time.

Design of economic experiments

We design economic experiments, run in the field or online, to see how people respond to policies and incentives. One example is our study of hiring discrimination in Labour Economics (2024).

Macroeconomic and sector diagnostics

Macroeconomic, fiscal and sector analysis, from the effects of sanctions to remuneration indices. We also build forecasting models, among them one that covers fifteen Latin American economies.

Design and advice on public policies and programmes

Applied policy design, from a new national food basket to school infrastructure surveys. We work alongside the teams that will run a policy, from the first draft to the indicators they will later use to monitor it.

How we work

The full cycle of empirical research

Instrument design and sampling strategy
Questionnaires, sampling designs and experimental protocols built around the policy question, not adapted from a template.
Data collection and field supervision
Household and establishment surveys, online platforms and field verification, run with our own supervision and quality control.
Rigorous, advanced data analysis
Robust quantitative and computational methods, from causal inference and small-area estimation to machine learning on survey and satellite data.
Interpretation and visualization
Findings written for decision makers, with the maps, figures and dashboards that let others check the work.

Selected counterparts

We work with ministries and municipal governments, multilateral banks and UN agencies, companies, foundations and community organizations. The method is the same whatever the size of the question.