Alternative diff-in-diffs estimators with several pretreatment periods

Mora, Ricardo; Reggio, Iliana

doi:10.1080/07474938.2017.1348683

Cited by 97 publications

(92 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For difference-in-differences to provide an unbiased consistent estimate of the treatment effect, the treatment group and the comparison group must follow the same trend over time in the absence of treatment (parallel trends assumption). 12 We test the assumption of parallel trends using the fully flexible model for parallel paths developed by Mora and Reggio (2019). 13 We find that our sample fulfills this identifying assumption.…”

Section: Econometric Analysismentioning

confidence: 87%

Under Pressure! Nudging Electricity Consumption within Firms. Feedback from a Field Experiment

Charlier¹,

Guérassimoff²,

Kirakozian³

et al. 2021

The Energy Journal

View full text Add to dashboard Cite

Many economists and psychologists have studied the impact of nudges on households' pro-environmental behaviors. Interestingly, "private nudges" can be imagined for companies. Yet, studies focusing on nudging employees' energy use are rare. The objective of our paper is to explore this issue with the help of a field experiment conducted at 47 French companies' sites. Using a difference-indifference methodology, the effects of three nudges on employees' energy conservation are tested. The first nudge, "moral appeal", stresses the responsible use of energy. The second one, "social comparison", informs employees on the energy consumption of other firms participating in the experiment. Finally, the third nudge, "stickers", alerts employees about good energy conservation practices. Our results stress the complementarity of these nudges. When implemented alone, the three nudges have no significant effects on energy consumption. However, when the moral appeal and social comparison nudges are combined with the stickers one, they become effective.

show abstract

Section: Econometric Analysismentioning

confidence: 87%

Under Pressure! Nudging Electricity Consumption within Firms. Feedback from a Field Experiment

Charlier¹,

Guérassimoff²,

Kirakozian³

et al. 2021

The Energy Journal

View full text Add to dashboard Cite

show abstract

“…The idea behind the interaction terms is to examine the effect of the Medicaid Expansion during each year it is implemented; perhaps the largest effects are in the year of the expansion or perhaps it may take time before dropout rates are affected. 29,30 The interaction terms allow for these different treatment effects.…”

Section: Discussionmentioning

confidence: 99%

The Effect of the Medicaid Expansion on Dropout Rates

Yeung

2020

Journal of School Health

View full text Add to dashboard Cite

BACKGROUND This study investigates how the Medicaid Expansion of the Affordable Care Act affected state high school dropout rates. METHODS This study relies on a differences‐in‐differences estimation strategy that is common in program evaluation, especially in education. This method replicates in a regression framework a classic pre‐test post‐test comparison group quasi‐experimental design. The analysis is conducted at the state level, which reduces the precision of the estimates. RESULTS States that adopted the Medicaid Expansion had a 0.658 percentage point greater reduction in dropout rates than non‐Expansion states in the year of Medicaid implementation. A back‐of‐the‐envelope calculation suggests that if all the remaining non‐Expansion states adopted Medicaid, there would be a decrease of over 92,500 youths who drop out of high school, representing a drop of 11.2% in the number of dropouts in these states. CONCLUSION The Medicaid Expansion of the Affordable Care Act is more than just a health insurance program; it is a dropout prevention program.

show abstract

“…This allows, for example, for the possibility that states responded to a rise in certain mortality trends by passing telemedicine parity mandates. While we view the specification with state trends as an exercise for demonstrating the robustness of the main estimates, we note that they are also valid research designs with different identification assumptions (Mora and Reggio 2017). Second, we perform falsification tests by examining mortality rates of accidents (transport and nontransport) and homicide.…”

Section: Methodsmentioning

confidence: 99%