Assessing the Performance of Matching Algorithms When Selection into Treatment is Strong

By B. Augurzky, Jochen Kluve. Social Science Research Network, 2007

56 citations recorded by the source corpus.

This paper investigates the method of matching regarding two crucial implementation choices, the distance measure and the type of algorithm.We implement optimal full matching – a fully efficient algorithm – and present a framework for statistical inference. The implementation uses data from the NLSY79 to study the effect of college education on earnings. We find that decisions regarding the matching algorithm depend on the structure of the data: In the case of strong selection into treatment and treatment effect heterogeneity a full matching seems preferable. If heterogeneity is weak, pair matching suffices.

Subfields: Economics and Econometrics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/f14f867c8ef5d3bd618cb4b2a53c3ae515f31d7f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Statistics and Probability

Citations in the corpus

The performance of estimators based on the propensity score (2013, background): “See, for example, Augurzky and Kluve (2007) and Hansen (2004) for some discussion of these issues.”
Evaluating continuous training programmes by using the generalized propensity score (2012, methodology): “Removal of bias by using the generalized propensity score Research in programme evaluation in recent years has made comprehensive use of matching methods (see inter alia the overview in Augurzky and Kluve (2007) and a symposium on the econometrics of matching in The Review of Economics and Statistic”
Optimal Full Matching and Related Designs via Network Flows (2006, methodology): “That algorithm is implemented by Gu and Rosenbaum (1993), and by Augursky and Kluve (2004), whose findings indicate that the absence of restrictions on treatment and control subjects’ relative proportions, in particular, is problematic, so much so that the latter authors reject it in favor of a…”
Some Practical Guidance for the Implementation of Propensity Score Matching (2008, unknown): “…estimator which works backwards and rearranges already matched treated individuals if some specific treated individual turns out to be a better (closer) match for an untreated previously matched individual (see Gu and Rosenbaum (1993) or Augurzky and Kluve (2007) for detailed descriptions).”