Re: matching software (fwd)

From: Daniel Feenberg <feenberg_at_nber.org>
Date: Sat, 15 Mar 2008 11:18:56 -0400 (EDT)

Caroline Hoxby writes that she has a need for name matching software, but
the software is expensive and would be licensed for all users on a
computer. Any family member who might wish to use such software (installed
on a Unix machine at 1050) is invited to write to me with a sentence about
the quantity of material, and what sort of material, you might have, and
we will further investigate the feasibility.

Daniel Feenberg

> On Thu, 13 Mar 2008, caroline hoxby wrote:
>
> > Dear Dan, Mohan, and Jean,
> >
> > I have a routine need for matching software (e.g. that matches people
> > from two datasets on the basis of their names and birthdates, that
> > matches firms from two datasets on the basis of the names and
> > locations, that matches colleges, and so on...). According to several
> > people with whom I've spoken, the software that does such matching has
> > improved greatly in recent days and saves a great deal of time on what
> > is otherwise a very bothersome task.
> >
> > I asked Juan Saavedra to do some research into what the best software
> > is, and he has come up with one well-reputed software package that
> > could be used by everyone at the NBER (can be put on the server with a
> > perpetual license). See attachment. He is also pursuing inquiries
> > with two other software companies. I think that all of these software
> > packages have trial versions so that we could ensure that we were
> > getting something very useful.
> >
> > Thus, my question is:
> > 1) Do we already have matching software on the NBER server or with an
> > NBER member whose license could be extended to the family?
> > 2) Before buying software, would you all need the need to do a survey
> > of NBER server users to see who might use it?
> > 3) Is there interest in buying matching for the NBER? Buying it is
> > pretty expensive for a single user who has the kind of volume that I
> > and other applied micro people have. On the other hand, it seems
> > cheap on a per-person basis for NBER users.
> > 4) Do you know of other matching software into which we should look?
> > I understand that SAS has a matching sub-component that you can buy,
> > but I've also heard negative things about it.
> >
> > Caroline
> >
> > --
> > Caroline M. Hoxby
> > Scott & Donya Bommer Prof. of Economics and Senior Fellow of the
> > Hoover Institution
> >
Received on Sat Mar 15 2008 - 10:18:56 EST