Tariff data to use in Stata

[Magnus Vattoy]

Hello,

I am new to this forum and econometrics in general so im not sure this is a appropriate line of questioning in this forum. I am constructing a dataset to use in Stata/SE 15, where my aim is to look at the effect of tariffs on U.S. manufactuing sector.

I am using the aggregation created by Li (2018), found here: https://www.card.iastate.edu/china/trade-war-data/ (point C). The data is allready aggregated to GTAP commodity level with corresponding trade values as weights. My goal is to further use this data to create two monthly variabels.

What I did was:
1. remove GTAP commodites not related to manufacturing, e.g. remove all GTAP sectors except 27-47 (https://www.gtap.agecon.purdue.edu/d...iledsector.asp). This was done using VLOOKUP in excel. Then I summed all the tariff increases and divided on number of rows. From this I got two variabels: 1) tariff increase that the U.S. put on import from the rest of the world and 2) tariff that is placed on exports coming from the US.

When I ran these against my dependent variable (production) in Stata I got that the "export tariff" has a positive effect, which sounds unlikely. I tried just summing the tariffincreases without dividing by number of rows (since they are already weighted) without any luck.

Does anyone have a way to use the data found on CARD to create monthly independent variables for tariff increases.

Would appreciate any help and thank you in advance.

PS: I am studying on a bachelor level, not yet very advanced in Stata so the regression I will use is very basic. Therefore a model like used in the reasearch paper here "Disentangling the Effects of the 2018-2019 Tariffs on a Globally Connected U.S. Manufacturing Sector" by Aaron Flaaen and Justin Pierce will be out of reach.

[Phil Bromiley]

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

Remember the we are not mainly from your area. Indeed, it is quite possible none of the active people on the listserv are from your area. Consequently, it is essential that you explain what you're doing in general terms that people conversant with statistics will understand.

First, I would generally try to dump things into Stata and do all the programming in Stata. You want to use the do file editor. This is because you then have a replicable set of code and can for example download new data and run it through the same program.

In Stata, we refer to variables and observations, not rows and columns.

I suspect you are making this difficult because you are using changes in tariffs rather than the level of tarrifts themselves. When you use a change variable on the right-hand side, the coefficient is then the change in predicted value for a one unit change in the change of the variable. What you really want is a change in predicted value for a one unit change in the variable.

With any date variable, the first thing you need to do is get Stata to recognize it as a date. If you formatted as a date in Excel, Stata may recognize it as a date. If not, you will need to convert it to a Stata date – look in the user guide that is part of the PDF documentation (under help, PDF documentation, you will see a long list of available documentation that comes with Stata). Once you make it a Stata date, then you can extract year and month easily. Alternatively, if there may be more than one change in a month, you can collapse the data by year and month.