Uniform Inference in High-Dimensional Dynamic Panel Data Models With Approximately Sparse Fixed Effects

Kock, Anders; Tang, Haihan

doi:10.1017/s0266466618000087

Cited by 23 publications

(44 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…21 This is consistent with the results from the simulated model, where ordinary least-squares estimates are biased in the presence of sufficiently many confounders; see Online Appendix EC.2. 22 We experimented with a version of this model that includes an even larger set of potential fixed effects and that is estimated by using the method suggested in Kock and Tang (2019). The results are quantitatively comparable and available from the authors on request.…”

Section: Discussionmentioning

confidence: 96%

Demand Estimation Using Managerial Responses to Automated Price Recommendations

2022

View full text Add to dashboard Cite

We provide a new framework to identify demand elasticities in markets where managers rely on algorithmic recommendations for price setting and apply it to a data set containing bookings for a sample of midsized hotels in Europe. Using nonbinding algorithmic price recommendations and observed delay in price adjustments by decision makers, we demonstrate that a control-function approach, combined with state-of-the-art model-selection techniques, can be used to isolate exogenous price variation and identify demand elasticities across hotel room types and over time. We confirm these elasticity estimates with a difference-in-differences approach that leverages the same delays in price adjustments by decision makers. However, the difference-in-differences estimates are more noisy and only yield consistent estimates if data are pooled across hotels. We then apply our control-function approach to two classic questions in the dynamic pricing literature: the evolution of price elasticity of demand over and the effects of a transitory price change on future demand due to the presence of strategic buyers. Finally, we discuss how our empirical framework can be applied directly to other decision-making situations in which recommendation systems are used. This paper was accepted by Omar Besbes, revenue management and market analytics.

show abstract

Section: Discussionmentioning

confidence: 96%

Demand Estimation Using Managerial Responses to Automated Price Recommendations

2022

View full text Add to dashboard Cite

show abstract

“…Finally, many empirical settings feature highdimensional fixed effects that suffer from limited variability. Some authors propose to penalize the fixed effects (e.g., Kock and Tang, 2019) with the Lasso regularization, while others do not (e.g., Belloni et al, 2016). The results in this paper suggest that penalizing fixed effects with the Lasso regularization can be problematic.…”

Section: Introductionmentioning

confidence: 88%

Omitted Variable Bias of Lasso-Based Inference Methods Under Limited Variability: A Finite Sample Analysis

Wüthrich

Zhu

2019

SSRN Journal

View full text Add to dashboard Cite

“…Fortunately, DPD methods can be used to specifically resolve these statistical problems [89][90][91][92][93][94][95]. DPD models allow direct estimation of difference equations with panel data, which resolves multiple problems that appear in the COVID-19 data [96].…”

Section: Estimationmentioning

confidence: 99%

Dynamic Panel Estimate–Based Health Surveillance of SARS-CoV-2 Infection Rates to Inform Public Health Policy: Model Development and Validation (Preprint)

Oehmke¹,

Oehmke²,

Singh³

et al. 2020

Preprint

View full text Add to dashboard Cite

BACKGROUND The Great Covid Shutdown worked when it was well implemented. Some countries completely eliminated SARS-CoV-2 while others flattened the curve. The US is more complicated due to varying social isolation policies which begs the question, when can we reopen? Reopening should be predicated on a sustained decline in Covid. Existing models of Covid-19 contagion rely on parameters such as R0 and use intensive data collection efforts, however these models use static statistical methods that do not capture all of the relevant dynamics such as varying specificity and sensitivity of diagnostic testing or asymptotic individuals unwittingly carrying the novel corona virus who are never tested. Existing Covid models use data that are subject to significant measurement error and other contaminants. Moreover, timely information is needed to improve statistical methods by extracting the information that is available from various datasets posted on websites. OBJECTIVE This study applies state-of-the-art statistical modeling applied to existing data on the internet to extract the best available estimates of the state-level dynamics of Covid-19 infection in the U.S. METHODS Dynamic panel data (DPD) models are estimated with the Arellano-Bond estimator utilizing the Generalized Method of Methods. This statistical technique allows for control of a variety of deficiencies in the existing dataset. Tests of the validity of the model and statistical technique are applied. RESULTS The results indicate 1) that the statistical approach is valid, including for determining recent changes in the pattern of infection, and 2) during the May 16-25 period the evolution of the pandemic changed with less inter-temporal persistence of the infection rate. This change represents a decline in the contagion model R value for that period, and is consistent with a ‘flattening of the curve’ but not with an imminent end to the pandemic. CONCLUSIONS Opening America comes with two certainties: 1)we will be Covid-free only when there is an effective vaccine; and 2)the “social” end is going to occur before the “medical” end of the pandemic, therefore, we need improved surveillance techniques to inform our leaders how to open sections of America more safely. DPD models can support opening America in combination with the extraction of Covid data from existing websites. CLINICALTRIAL NA

show abstract

Uniform Inference in High-Dimensional Dynamic Panel Data Models With Approximately Sparse Fixed Effects

Cited by 23 publications

References 36 publications

Demand Estimation Using Managerial Responses to Automated Price Recommendations

Demand Estimation Using Managerial Responses to Automated Price Recommendations

Omitted Variable Bias of Lasso-Based Inference Methods Under Limited Variability: A Finite Sample Analysis

Dynamic Panel Estimate–Based Health Surveillance of SARS-CoV-2 Infection Rates to Inform Public Health Policy: Model Development and Validation (Preprint)

Contact Info

Product

Resources

About