Blog by Nate Archives: Foreign Direct Investment Data Note (July 9, 2013)

[The blog migration continues to my new WordPress site.  More data updates to follow.  I hope.  I can’t promise.]

Foreign Direct Investment Data Note

I get a large number of questions from people looking for high quality data on foreign direct investment.  My quick answer is generally, “let me know if you find it.”

Here is a quick guide to some cross-naional FDI data.

1)  The most commonly used FDI data in political science is from the World Bank’s World Development Indicators(or alternatively the IMF International Financial Statistics).  I use FDI/GDP in a 2003 International Organization paperand Quan Li and Adam Resnick use FDI in dollars in their 2003 International Organization paperHere and here are two studies that explore how outliers affect our results.

The major problem with this data is that it is highly aggregated at the country-year level.  We know nothing about the sector of investment, the form of entry (greenfield, brownfield, joint venture, etc) or the country of orgin of the investor.

Another concern is that we know very little about the quality of this data.

2)  An alternative data source is OECD inflow and outflow data which also breaks down the data by sector.

One advantage of this data is that it also includes FDI stock data, not just the flows of foreign investment and this is broken down by sector.  This data can be used to look at OECD inflows, or outflows from the OECD.

In a little working paper we looked at this data and we’re surprised by a number of data quality issues.  First, many of the OECD countries have missing inward and outward data.  Second, many of the observations don’t match.  (This is also discussed in this UNCTAD report). The amount of FDI France says that it sends to Germany doesn’t match the amount of French FDI that Germany recorded as inflows.  We try to deal with this issue in old working paper that needs updating, but it seems like a major problem.

3)  Less commonly used is the UN Conference on Trade and Development FDI data.  This includes data on stocks and flows of FDI.  See Eddy Maleksy’s QJPS article for an example.

These are the three main sources of cross-national data.  I’ll write up another blog post that documents some other sources including US BEA data, Japanese foreign affiliate data, and some other sources of FDI.