Panel data model for policy intervention

[Mikhail Alekseev]

Research question: I am writing a master thesis where I am analyzing the impact of a policy on the flow of investments from country A to country B.

Data: I have highly(!) unbalanced yearly panel data on investment flows for 9 years (policy were implemented on year 5), i. e. microdata for the investment flows of every company resided in country A and investing in country B.

Research design: My independent variable is individual investment flows in US dollars. I have a main independent variable, which is Policy coded as dichotomous dummy variable (0 = before policy implementation, 1 = after policy implementation). Plus, I have several control macroeconomic variables, such as GDP growth in a receiving country B, market size in a country B, etc.

Statistical model: That's where my questions begin. In the last weeks, I read several articles and statistics books on panel data analysis and decided that the most suitable model for my design and data available is One-way fixed time effects model (with LSDV as estimator for time dimension). Unfortunately, it is not possible to do two-way fixed effects as each year there are firms which exit and enter the investment market (that's what I meant with "highly unbalanced").

My questions:

1. Is that a right approach? Can this data be used for the model described?
2. I also want to analyze how this policy shift impacted investments in various economic sectors (e.g. finance, agriculture, energy, etc.). In the dataset, I have the information which investment goes to which sector. Can I create a categorical variable with sectors to see how the sector affiliation impacts the flow? Can the sector category be Interpreted as individual effects allowing for the complete FE model?
3. I also want to see how different components of the policy influenced the investments. E.g. for example there is a tax deduction for agriculture, there is low interest rate for energy, and both for finance.
4. Maybe, I want too much and have to consider more than one model?

I would be glad to receive your advice and any other hints.

[Phil Bromiley]

Welcome to Stata list. You will increase your chances of a useful answer by following the FAQ on asking questions – provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

Your primary input on what model to analyze should be coming from your thesis advisor. Your advisor and related faculty will know on awful lot more about this than we do.

It's important to think about why those firms are disappearing or coming back. Are they truly missing data or are they really zeros? If they are actually zeros – no investment flow between those two countries for this specific company in that specific year – then treating them as missing is incorrect. It also make your life much more difficult. Most of the difference in difference approaches to this problem assume you have observations on the same entities before and after the policy change.

It is not clear what your primary fixed effects variable would be. Are you intending fixed effects for each firm? If you have more than one observation for a firm, I'm not sure why you can't include year dummy variables along with fixed effects for firms. You do lose all the firms that only appear in one year.

If you want to understand how the shift influenced investment in different sectors, I suspect you need to do an analysis by sector – that is if investment is your dependent variable and you want to talk about investment in agriculture, then investment in agriculture seems like it should be the dependent variable. This naturally would result in a separate model for each economic sector.

[Mikhail Alekseev]

Thank you for your response! I will consider producing several models for each sector, respectively.

I am now looking into other options how do conduct the analysis since I came to realize that it is actually not a panel data but rather time series cross section, that is a register of all investment transactions between countries. But if I wanna pack everything in one regression, is it theoretically viable to make a fixed effects variable from the sectors after having made some conceptual assumptions?