Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Re: RE: Re: RE: questions about stationarity test results using IPSHIN


From   Jeannette Wicks-Lim <janetlim@econs.umass.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: Re: RE: Re: RE: questions about stationarity test results using IPSHIN
Date   Fri, 19 Nov 2004 09:28:59 -0500

Hi Nick,

One other follow-up question. Why do you say that the original paper is needed to judge the significance of the t-tests? I assumed that the critical values provided are enough and the paper was only needed if you wanted the exact p-values for the t-statistics estimated.

I did follow your suggestion and just graphed a couple of states' series. The plot of -lngap- does look to be stationary with a deterministic trend. And -lnmin1- appears to be nonstationary because it is basically a series with a bunch of structural breaks (each break being a legislated raise in the minimum wage), not the typical erratic movement that I think of when I think of nonstationary series. I guess because of that the ratio between -lnq5- and -lnmin1- is stationary...that's my guess for now!

Thanks again,
J

----- Original Message ----- From: "Nick Cox" <n.j.cox@durham.ac.uk>
To: <statalist@hsphsun2.harvard.edu>
Sent: Thursday, November 18, 2004 5:59 PM
Subject: st: RE: Re: RE: questions about stationarity test results using IPSHIN



My comment is very simple: put another way,
on the t-bar statistics -lnq5- and -lngap-
are behaving relatively similarly, whereas -lnmin1-
differs from both. No more than that, and no
less.

The help for -ipshin- makes it clear that
to judge significance you need to get
results from the original paper.

Nick
n.j.cox@durham.ac.uk

Jeannette Wicks-Lim

Maybe I'm misunderstanding the results ( I am a real novice
re: time series
issues). I thought that the t-bar being further from zero for lngap
indicates that it is stationary (more negative then the cv1). Is that
incorrect?
Nick Cox

> Your commentary seems at odds with your results
> in that t-bar is further from zero for -lngap-
> than for the other variables.
Jeannette Wicks-Lim

>> I've conducted the IPSHIN test on two variables, one of which
>> appears to be
>> nonstationary (log of the minimum wage, or "lnmin1") and the
>> other appears
>> to be stationary (log of the 5th wage percentile, or "lnq5").
>> When I create
>> a third variable (log of 5th wage percentile - log of minimum wage,
>> or"lngap"), the IPSHIN test indicates that it is stationary.
>> How can it be
>> that the ratio of a stationary and nonstationary variable is
>> stationary?
>> (Some background info: the panels in this dataset are US
>> states -- all 50,
>> the time points are 6 month intervals over 20 years).
>>
>> Here are my results:
>>
>> . ipshin lnmin1 if gestcen~=53, lags(17) trend
>>
>> Im-Pesaran-Shin test for cross-sectionally demeaned lnmin1
>> Deterministics chosen: constant & trend
>>
>> t-bar test, N,T = (50,40)         Obs = 1593
>> Augmented by 17 lags (average)
>>
>>     t-bar     cv10      cv5       cv1   W[t-bar]    P-value
>>    -1.456   -2.320    -2.360    -2.440        .         .
>>
>> . ipshin lnq5 if gestcen~=53, lags(17) trend
>>
>> Im-Pesaran-Shin test for cross-sectionally demeaned lnq5
>> Deterministics chosen: constant & trend
>>
>> t-bar test, N,T = (50,40)         Obs = 1593
>> Augmented by 17 lags (average)
>>
>>     t-bar     cv10      cv5       cv1   W[t-bar]    P-value
>>    -3.087   -2.320    -2.360    -2.440        .         .
>>
>> . ipshin lngap if gestcen~=53, lags(17) trend
>>
>> Im-Pesaran-Shin test for cross-sectionally demeaned lngap
>> Deterministics chosen: constant & trend
>>
>> t-bar test, N,T = (50,40)         Obs = 1593
>> Augmented by 17 lags (average)
>>
>>     t-bar     cv10      cv5       cv1   W[t-bar]    P-value
>>    -3.585   -2.320    -2.360    -2.440        .         .
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index