[SciPy-user] Software for economic experiments

Guilherme P. de Freitas guilherme@gpfreitas....
Thu Dec 4 02:13:39 CST 2008


>> I was wondering if this could not
>> be done with Python.
>
> It certainly can be done with Python.
> I might be interested in helping.

Great! As for the language, I feel that not only it is possible to use
Python, but that it is also a good choice. Do you agree? Do you
suggest another language or set of languages? I know this is a Python
mailing list, so we are all biased but... anyway, it's good to check.

> Will one subject at a time interact with the software, or
> would it need to synchronously elicit responses from
> multiple human subjects? What kind of network configuration
> will you rely on in the lab?

I still have to find out about the details. See, I'm not directly
involved in the project, I'm just a grad student that works with the
professor on other things, but that saw the hacks they were doing to
run experiments and thought that "there must be a better way to do
this!".

That said, from what I've talked to the grad student that is actually
involved in the project, I think that at each stage (which lasts, say
90s) subjects make their decisions (how much to fish, demand for
fishing quotas, supply for fishing quotas, bids on the auction for
quotas). It is not clear to me what kind of "money" the subjects would
use in the auction, if its fake money distributed by the auctioneer or
if it comes from their profits. It should come from profits, but I
have to get the details.

>
> What department are you in?  What are the research goals?
> Is coauthorship a possibility?  Would the code be free and
> open source (e.g., BSD license)?  Is the existing auction
> code already FOSS licensed?  What language is it in?

I'm a PhD student in the Humanities and Social Sciences division at
Caltech. This research is led by prof. John Ledyard, who is my main
advisor at the moment. From what I understood, the goals are to set up
new regulations for fisheries in the Pacific coast that are more
efficient, environmentally "healthy" and politically viable (for
example, current fishing companies will have to be given a bigger
share of the pie). As for co-authorship, again, I'm not part of the
project, but I can ask the people involved about it.

I can't say for sure, but I firmly believe the code will be licensed
under some FOSS license. I would definitely make a lot of noise here
if it did not! The trading (general equilibrium) software is called
jMarkets, and it is written in Java. Here's the website:

http://jmarkets.ssel.caltech.edu/

It is licensed under the GPL, as you can see on the website. Prof.
Charles Plott is the main person behind jMarkets at the moment, from
what I know. There is also a jAuctions:

http://www.hss.caltech.edu/~jkg/jAuctions.html

also in Java. Prof. Jacob Goeree is the main person behind jAuctions.
Anyway, I'll ask John (Ledyard) about these details too.

> Feel free to send a couple relevant papers,
> if that might help me see into your project.
> I have no substantial fisheries knowledge,
> but I am an economist with Python programming
> experience, and I think fisheries issues are
> interesting.

Great! We'll be in touch.


-- 
Guilherme P. de Freitas
http://www.gpfreitas.com


More information about the SciPy-user mailing list