Working papers

Multimarket Contact: Competition or Collusion. Evidence from the US Airline Industry. , joint with G. Arie and S. Markovich

Research Articles under Review:


Published Papers:

M. Varela, M. Viswanathan. 2020. Savings that Hurt: Production Rationalization and its Effect on Prices, Journal of Economics and Management Sciences, 29(1): 147-172 [pdf]

M. Varela. 2018. The Costs of Growth: Accelerated Growth and Crowd-Out in the Mexican Supermarket Industry International Journal of Industrial Organization, 61: 1-52 [pdf]

Awarded journal's best empirical paper of 2018

M. Mazzeo, K. Seim, M. Varela. 2018. Welfare Consequences of Mergers with Endogenous Product Choice, Journal of Industrial Economics. 66(4): 980-1016 [pdf]

L. Dafny, K. Ho, M. Varela. 2010. An Individual Healthplan Exchange: Which Employees Would Benefit and Why?. American Economic Review, Papers and Proceedings. 100: 485-489. [pdf]

L. Dafny, K. Ho, M. Varela. 2012. Let Them Have Choice: Gains From Shifting Away from Employer-Sponsered Health Insurance and Toward an Individual Exchange. American Economic Journal Public Policy. 5(1): 32-58. [pdf]

Data and Code:

Below I share personal programs, coded for Stata and Matlab, meant for an applied economists. They simplify coding and, in many circumstances save on computational time relative to standard packages


LSFEREG – Short for `Large-Scale Fixed Effect Regression’, useful to implement N-way fixed effect regressions. It is the natural extension of `areg’ when more than one group of fixed effects is to be included. For example, this program is useful if your data consists of a panel of markets over time and you wish to include fixed effects for markets and time. See the help file for more details. [ado] [sthlp]

LSFEIVREG – Short for `Large-Sacle Fixed Effect IV Regression’, this is the extension of LSFEREG for endogenous repressors. See the help file for details. [ado] [sthlp]


BY – Replicates, partially, the STATA ‘egen’ command for Matlab. Given an N-by-K set of index variables (N observations, K index variables), an N-L matrix of data variables, and a string command (e.g. `sum’, `mean’, `median’,  `max’, `count’, `sdev’, `tag’), it returns an N-L matrix of variables containing, for each group of observations defined by a unique combination of the index variables, the relevant statistic for each one of the data variables. See file for details. [file]

NELDERMEAD – Implements a Nelder Mead, gradient free optimizer. See file for details. [file]

All works © Mauricio Varela 2016.