Quantitative Analysis: Econometrics & Impact Assessment

By applying quantitative methods, our aim is to support policy-makers in measuring the impacts of policies, comparing relative strengths and weaknesses of their activities, as well, as identifying how to achieve their goals. The relevant method depends on the circumstances of the problem and the availability of data. In our applied work both econometrics and calibrated methods are present.

HETFA’s impact assessment employs an interdisciplinary framework built on several subfields of social and behavioral sciences. Our main methodological modules are the following. We developed a macroeconomics-based model in order to evaluate macro-consequences of measures (e.g. impacts on regulatory environment, labor market, economic performance or taxation). A statistics and econometrics-based framework serves for the purpose of an (ex post) evaluation of causal relationships between policy measures and expected consequences. Our institutional approach based on political economy and institutional economics offers insights about initiative’s effects on incentives and institutional environment.

Impact Assessment of Public Policy Measures

Our programme evaluation applied research projects used econometrics methods to quantify the effects of policy actions. We put emphasis on the application of counterfactual methods although programme design is typically different from randomization. In case of counterfactual impact assessment settings, the quantitative impact of a given programme is differentiated from a comparative pattern to identify how the group of ’treated’ (program beneficiaries) behaves differently from the group of ’non-treated’ or from a subsample of ’non-treated’ which is the so-called control group. The identification requires the control group to be as much similar to the treated group as much it is possible in all relevant aspects that may influence the outcome variable. However, in most applied situations, it is very difficult to find a relevant control group which makes the comparison plausible.

Depending on the problem, we apply ‘difference in differences’ method or some more advanced versions of it, like probability score methods or panel data methods. Discontinuity design methods are also applied when programme discontinuity makes it possible to apply. We are conducting active research in applying panel data methods to spatial data sets by using spatial econometrics methods.

Impact assessment outcomes are valuable in at least four separate dimensions. The decision-support function of assessment helps decision makers to develop and compare alternative choices, while the system support function helps to overcome public management challenges. A communication function might address an increase in the acceptance of the assessed measure. Other important assessment dimensions is the support of policy implementation.

An important virtue of impact assessment is the capacity to carry out focuses of analysis along several dimensions. Different focuses can be identified upon thematic objectives – our assessments are able to provide knowledge about economic, social, environmental and other specific fields of impacts of initiatives. Furthermore, impact assessment can provide insights for decision makers about the size of changes, risks of implementation and possible social and political reactions to policy choices. In addition, it is capable to provide specific information about distinct groups of stakeholders (e.g. target groups, key actors, or groups of high political importance).

Macroeconomic Impacts of Major Projects and Development Programmes

We are also developing modelling frameworks to estimate macroeconomic impacts of either major investment projects or different development programmes. These projects and programmes usually concern only a few sectors of the economy and in some cases, focuses only in one region However, their volume is able to indicate the potential diffusion of their impact through the multiplication of disposable income and the effects on consumption demand. Therefore our approach is based on input-output tables and macroeconometrics combining macro-level and sector-level empirics.