[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: machado mata stata
ChangHwan Kim <firstname.lastname@example.org>
Re: st: machado mata stata
Wed, 15 Jul 2009 10:37:08 -0500
Thank you for your superior googling. So, where is the code? I also
googled it, but I can't see the code.
Melly's Stata code is for his new decomposition which is not the exactly
same as Machado & Mata, although it is said to yield essentially
identical results. Many articles explain the procedure to conduct
Machodo & Mata with appreciation to contributors who provided them the
If it is a rule to write my own code, that's fine, I will do that. I
just wish I can appreciate someone when (s)he shows me the code like the
authors in other papers I googled.
btw, "http://lmgtfy.com" looks cool.
Schaffer, Mark E wrote:
> I can't resist using Austin's reply below to introduce the list to the wonderful web site "Let me Google that for you:
>> -----Original Message-----
>> From: email@example.com
>> [mailto:firstname.lastname@example.org] On Behalf Of
>> Austin Nichols
>> Sent: 15 July 2009 15:37
>> To: email@example.com
>> Subject: Re: st: Date: Wed, 15 Jul 2009 09:17:21 +0100
>> A Google search of "machado mata stata" turns up a lot of
>> useful results, including
>> with Stata code by Blaise Melly (2007)
>> (see also
>> lly ) http://ftp.iza.org/dp3428.pdf
>> in which the authors "thank Yuriy Gorodnichenko and Klara
>> Sabirianova Peter for providing
>> us with their Stata routine of the Machado-Mata methodology
>> of wage decompositions."
>> which says (in fn.4):
>> This description of the simulation procedure follows Machado
>> and Mata (2005). Standard errors for the estimated quantiles
>> of the counterfactual distribution that is generated in this
>> way can be found by bootstrapping, as described in Machado
>> and Mata (2005), or by applying the asymptotic results given
>> in Albrecht et al. (2009). An alternative procedure for simulating
>> Ft,s(y) is as follows:
>> (i) Estimate βs(q) for a grid of values, e.g., q = 0.01, 0.02, etc.
>> (ii) Multiply each estimated quantile regression coefficient
>> vector by each x in year t’s empirical distribution of observables.
>> Variations on this alternative procedure have been used by
>> Albrecht et al. (2003), Autor et al. (2005) and Melly (2007).
>> Standard errors for the estimated quantiles of the
>> counterfactual distribution that is generated using this
>> alternative procedure can again be found by bootstrapping or
>> by applying the asymptotic results given in Melly (2007). The
>> results presented in this paper were generated using a STATA
>> program written by Blaise Melly. This program implements the
>> procedure described in this footnote and gives bootstrapped
>> standard errors. We have replicated our results using the
>> STATA program written by Aico van Vuuren.
>> This program implements the original Machado and Mata (2005)
>> algorithm and gives the asymptotic standard errors derived in
>> Albrecht et al. (2009). The results using the two different
>> programs are essentially identical.
>> On Wed, Jul 15, 2009 at 4:18 AM, Stephen P.
>> Jenkins<firstname.lastname@example.org> wrote:
>>> Amadou, if you find the paper that provides the code,
>> please post the
>>> reference/URL to the list
>>> I've just seen this thread.
>>> I recall a paper doing a similar decomposition using quantile
>>> regression and they provide a stata code at the end (in the case of
>>> I'll try to find the paper and send it to you privately.
>>> 2009/7/14, ChangHwan Kim <email@example.com>:
>>>> OK. Here is the full reference.
>>>> José A. F. Machado and José Mata, "Counterfactual Decomposition of
>>>> Changes in Wage Distributions using Quantile Regression", /Journal
>>>> Applied Econometrics/, Vol. 20, No. 4, 2005, pp. 445-465.
>>>> Thanks, CH
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/statalist/faq
>> * http://www.ats.ucla.edu/stat/stata/
* For searches and help try: