Linear Dynamic Panel-Data Estimation using Maximum Likelihood and Structural Equation Modeling

Published on Jun 1, 2018in Stata Journal1.80
· DOI :10.1177/1536867X1801800201
Richard Williams18
Estimated H-index: 18
(ND: University of Notre Dame),
Paul D. Allison46
Estimated H-index: 46
(UPenn: University of Pennsylvania),
Enrique Moral-Benito18
Estimated H-index: 18
(Bank of Spain)
View in Source
Panel data make it possible both to control for unobserved confounders and to include lagged, endogenous regressors. Trying to do both at the same time, however, leads to serious estimation difficulties. In the econometric literature, these problems have been solved by using lagged instrumental variables together with the generalized method of moments (GMM). In Stata, commands such as xtabond and xtdpdsys have been used for these models. Here we show that the same problems can be addressed via maximum likelihood estimation implemented with Stata’s structural equation modeling (sem) command. We show that the ML (sem) method is substantially more efficient than the GMM method when the normality assumption is met and suffers less from finite sample biases. We introduce a command named xtdpdml with syntax similar to other Stata commands for linear dynamic panel-data estimation. xtdpdml simplifies the SEM model specification process; makes it possible to test and relax many of the constraints that are typically embodied in dynamic panel models; and takes advantage of Stata’s ability to use full information maximum likelihood (FIML) for dealing with missing data.
  • References (14)
  • Citations (8)
#1Sebastian Kripfganz (University of Exeter)H-Index: 4
22 CitationsSource
32 CitationsSource
#1Jushan Bai (CUFE: Central University of Finance and Economics)H-Index: 40
28 CitationsSource
#1Kenneth A. Bollen (UNC: University of North Carolina at Chapel Hill)H-Index: 69
#2Jennie E. Brand (UCLA: University of California, Los Angeles)H-Index: 18
112 CitationsSource
#1Daniel A. Newman (PSU: Pennsylvania State University)H-Index: 25
348 CitationsSource
#1Cheng Hsiao (HKUST: Hong Kong University of Science and Technology)H-Index: 41
#2M. Hashem Pesaran (University of Cambridge)H-Index: 70
Last.A. Kamil Tahmiscioglu (UWM: University of Wisconsin–Milwaukee)H-Index: 2
view all 3 authors...
256 CitationsSource
#1Craig K. Enders (UM: University of Miami)H-Index: 35
#2Deborah L. Bandalos (NU: University of Nebraska–Lincoln)H-Index: 10
2,093 CitationsSource
#1Seung C. Ahn (ASU: Arizona State University)H-Index: 15
#2Peter Schmidt (MSU: Michigan State University)H-Index: 56
692 CitationsSource
#1Manuel Arellano (LSE: London School of Economics and Political Science)H-Index: 32
#2Stephen Bond (University of Oxford)H-Index: 38
16.9k CitationsSource
#1Christopher Cornwell (WVU: West Virginia University)H-Index: 21
#2Peter Rupert (WVU: West Virginia University)H-Index: 1
99 CitationsSource
Cited By8
#1Vitor Miguel Ribeiro (University of Porto)H-Index: 1
#2Celeste Varum (University of Aveiro)H-Index: 11
view all 3 authors...
#1Matthew Thomas Clement (Texas State University)H-Index: 8
#2Nathan W. Pino (Texas State University)H-Index: 11
Last.Jarrett BlausteinH-Index: 5
view all 3 authors...
#1Thomas J. Mowen (BGSU: Bowling Green State University)H-Index: 10
#2Benjamin W. Fisher (University of Louisville)H-Index: 7
#1Michael J. Zyphur (University of Melbourne)H-Index: 24
#2Manuel C. Voelkle (Humboldt University of Berlin)H-Index: 14
Last.Ed Diener (UVA: University of Virginia)H-Index: 133
view all 11 authors...
#1Michael J. Zyphur (University of Melbourne)H-Index: 24
#2Paul D. Allison (UPenn: University of Pennsylvania)H-Index: 46
Last.Ed Diener (UVA: University of Virginia)H-Index: 133
view all 11 authors...
4 CitationsSource
#1Enrique Moral-Benito (Bank of Spain)H-Index: 18
#2Paul D. Allison (UPenn: University of Pennsylvania)H-Index: 46
Last.Richard Williams (ND: University of Notre Dame)H-Index: 18
view all 3 authors...
5 CitationsSource
#1Marco FerrettiH-Index: 6
#2Salvatore FerriH-Index: 2
Last.Alessandro SapioH-Index: 8
view all 5 authors...
3 CitationsSource