The following are some of the most commonly used datasets in economic growth research. The first set is country-level, but following those are some links to micro-level data (firms, individuals) that are becoming more and more widely used. For a colossally thorough set of links to available datasets, see the AEA site.
Country-level data:
- Groningen Growth and Development Centre. Probably the single most useful site. Includes the newest Penn World Tables data (8.0), Maddison’s historical series, and industry/sector level data from a wide range of countries.
- Barro/Lee education data. Years of education and enrollment rates.
- World Bank Development Indicators. Includes a variety of economic data (GDP, external balances, labor force indicators) as well as health, education, and inequality data.
- World Bank Doing Business. This is probably the most up-to-date source for what you might call “institutions” data, at least as they related to frictions facing firms.
- Macro data for Stata. Catini, Panizza, and Saade have provided a great public service and standardized a wide set of country-level data so that you can combine data sources easily. Leans towards open economy topics (exch rates, financial flows), but has quality of governance indicators.
At the micro-level, the following are probably the most widely used
- IPUMS. This started as the online home of historical U.S. Census individual level data, and now contains individual-level census data from a range of countries. Great ability to search for variables, and they’ve standardized nearly all data so it is comparable across countries.
- Demographic and Health Surveys. Relatively consistent set of surveys done in a wide set of developing countries, often multiple years. Heavily weighted towards (not surprisingly) fertility, health, and mortality data, it also has some economic information.
- World Bank Living Standard Measurement Surveys. Very detailed surveys done in a set of developing countries, with a lot of fine-grain information on time allocations to various economic activities, asset ownership, demographics, health, and more. Some countries actually have panels available. May require you to contact national statistical agencies for permission.
Firm-level data is getting a lot of attention. It is still early enough that there is no consistent set of firm data across countries that is the standard. The World Bank does have what it calls Enterprise Surveys, which contain a lot of firm-level financial data and information on frictions faced by firms. The surveys only cover formal firms, and some of the earlier years are not quite representative, so be sure to read the documentation before using.
The Growth Economics Blog 
The Enterprise Surveys are actually representative after 2006/2007. They pursue a common methodology, and unlike the LSMS cited above can actually allow for cross-country comparisons. They are also the only Firm-level surveys that are both comparable across countries and ask certain questions about the business environment specifically regarding corruption and regulatory burden. Of course, they are only representative at the sector of survey stratification.
The methodology is clearly stated here:
http://www.enterprisesurveys.org/Methodology
The representative nature of the surveys is fairly accepted in academia. For a list of working papers and publications based on the data, please see the excel sheet in the righthand column of the page below under “additional research.”
http://www.enterprisesurveys.org/Research
Please contact me if you have any queries about the surveys.
Asif – thanks for the information. My knowledge of the Enterprise Surveys is a few years old, so I will check out the methodology and make an update on the blog. Thanks. DV
That would be great. And any feedback or recommendations would also be much appreciated.
Best regards.