Have you received a reply? I haven't seen one. RSiteSearch("panel
data unit root test") produced another question on this but no answers
that I saw. RSiteSearch("unit root test") produced 75 hits with many
useful. RSiteSearch("panel data analysis") produced 75 hits, the first
of which suggested the nlme package
(http://finzi.psych.upenn.edu/R/Rhelp02a/archive/26962.html). You could
construct nested models to compare the corAR1 correlation structure with
first differences.
hope this helps.
spencer graves
jukka ruohonen wrote:
Hi,
The question is as follows: has anyone coded panel data unit root tests
with R? Even the "first generation" tests (see e.g. Levin & Lin 1993;
Pesaran, & Smith & Im 1996; Maddala & Wu 1999) would be sufficient for
my needs. To my understanding, these are rather easy to code, but as I
have taken just my first steps in coding with R, existing code would
save me from a lot of trouble & time.
With regards,
Jukka Ruohonen
University of Helsinki
References:
Levin, A. & Lin, C.F. (1993): Unit Root Tests in Panel Data.
ftp://weber.ucsd.edu/pub/econlib/dpapers/ucsd9356.pdf
Maddala, G.S. & Wu, S. (1999): A Comparative Study of Unit Roots Tests
Panel Data and a New Simple Test. Oxford Bulleting of Economics and
Statistics. Special Issue 61, 631-652.
Pesaran, M.H. & Smith, R. & Im, K.S. (1996): Dynamic Linear Models for
Heterogenous Panels. In: Matyas, L. & Sevestre, P. (eds.): The
of Panel Data: a Handbook of the Theory with Applications, second