Levin-Lin-Chu panel unit root test
levinlin varname [if exp] [in range] ,lags(numlist) [ noconstant trend ]
levinlin is for use with panel data. You must tsset your data before using levinlin, using the panel form of tsset; see help tsset.
varname may contain time-series operators; see help varlist.
Users of Stata 11+ should use the official xtunitroot llc command.
Description
levinlin estimates the panel unit root test developed by Levin, Lin and Chu (LLC, 2002). The test assumes that each individual unit in the panel shares the same AR(1) coefficient, but allows for individual effects, time effects and possibly a time trend. Lags of the dependent variable may be introduced to allow for serial correlation in the errors. The test may be viewed as a pooled Dickey-Fuller test, or an Augmented Dickey-Fuller (ADF) test when lags are included, with the null hypothesis that of nonstationarity (I(1) behavior). After transformation by factors provided by LLC, the t-star statistic is distributed standard normal under the null hypothesis of nonstationarity.
The test differs from Taylor and Sarno's (1998) multivariate ADF test, madfuller, in that the latter is estimated using the seemingly unrelated regressions (SUR) estimator (that is, with an equation for each individual) und > er the same constraint of a single autoregressive parameter across individuals. The MADF test cannot be employed on large-N, small-T panels, since T must excee > d N in standard SUR. This restriction does not apply to the LLC test.
Options
lags must be specified, and may take on any non-negative value. If a single value is provided, that lag length is used for all individuals. If a list of lags is provided (perhaps with a local macro), its length must match the number of individuals in the panel.
noconstant omits the constant term from the estimated equation (LLC case 1).
trend includes a time trend in the estimated equation (LLC case 3).
Examples
. use http://fmwww.bc.edu/ec-p/data/hayashi/sheston91.dta,clear
. levinlin rgdppc if country<11, lag(2)
. levinlin rgdppc if country<11, lag(2 2 2 3 3 3 4 4 4 4)
. levinlin D.rgdppc if country<11, lag(2) trend
References
Banerjee, Anindya. Panel Data Unit Roots and Cointegration: An Overview. Oxford Bulletin of Economics and Statistics, Special Issue, 607-629, 1999.
Levin, Andrew, Lin, Chien-Fu and Chia-Shang James Chu. Unit Root Tests in Panel Data: Asymptotic and Finite Sample Properties. Journal of Econometrics, 108, 1-24, 2002.
Maddala, G.S. and In-Moo Kim. Unit Roots, Cointegration, and Structural Change, Cambridge: Cambridge University Press, 1998.
Taylor, Mark P. and Lucio Sarno. The behavior of real exchange rates during the post-Bretton Woods period. Journal of International Economics, 46, 1998, 281-312.
Acknowledgements
We are very grateful to Gene Liang for pointing out discrepancies in this routine and providing detailed output from his routines, to J.J.A. Ewalts for bringing an issue with lag truncation to our attention, to Claudio E. Raddatz for pointing out a flaw in our error-checking logic and to Lilin Liang for detecting a problem with tempnames. Remaining errors are ours.
Authors
Fabian Bornhorst, European University Institute, Italy, Fabian.Bornhorst@iue.it Christopher F Baum, Boston College, USA, baum@bc.edu
Also see
On-line: help for est, reg3, dfuller, madfuller (if installed)