Travel cost method combining RP and SP

[Balta Sunder]

Hi
I have a dataset with info on #trips for each individual (a total of 200) to a particular recreational site during a year. This is my RP data. For each individual I also have the number of trips expected under three different quality change scenarios. This is my SP data. I am a little at a loss as how to analyze this, both in general, but particularly in Stata. My central goal is to be able to calculate what happens to the Consumer Surplus under each of the quality changes of the site as compared to the current state.

1) Can I analyze the RP and each of the SP-scenarios separately and simply compare the CS between them? For this I might use Poisson or Neg binomial models, which I know how to handle.

2) Or should the RP and SP be analyzed in combination, in a panel? And in such case
a) is there a command in Stata to make the conversion from non-panel to panel data? As I said, all the RP and SP info is currently in only one row for each individual.
b) what should the data look like when ready for analysis using panel data? Many almost identical rows for each individual except for (i) #trips and maybe (ii) an indicator for hypothetical values (ie for SP/RP, se below)?
c) how can I handle the fact that zero-trip-responses are not possible in the on-site RP (since respondents have obviously been to the site at least once) but that zero-trip-responses are possible under the contingent scenarios
d) which stata commands can be used in this panel analysis
e) should I include an indicator for hypothetical values (i e for SP/RP), and if, how should I use the info from this indicator in the regression

I am a little at a loss regarding this and would appreciate help on any of the above questions.

Best regards,
Balta

[Phil Bromiley]

Welcome to Stata list. You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions-provide Stata code in code delimiters, readable Stata output (fixed spacing fonts helps), and sample data using dataex. It is not obvious from your posting how your data is set up, etc. reshape is the function the changes from long to wide and wide too long. We can help you more than that since we don't really know what your data looks like.

Having the same values for many of the right-hand side variables is okay as long as the dependent variable varies. Note that variables that do not vary within panels cannot be estimated with fixed effects. One might need to look at a hybrid model like user written xthybrid.

You really have a question of substance on which you have to be the expert – we can help you with Stata but you have to be the expert on the theory.