This page collects data files and tools I have constructed while working with Vietnamese microdata.
In McCaig, McMillan, Ngoma, and Pham (2026), we construct a dataset that matches changes in administrative boundaries of wards/communes over the period 1999 through 2019. The dataset documents every split, merger, and reassignment that occurred during this period, with decree references wherever possible. This dataset can be used to construct geographically consistent boundaries at the province, district, and ward/commune level. We provide one harmonization to 1999, but you can construct your own harmonization to any reference year. You can access the dataset (Jun 2026) here. The dataset can be used with commune and ward codes in the Vietnam Enterprise Surveys (2000-2019), the Vietnam Household Living Standards Survey (2002-2018), and the population censuses (1999, 2009, and 2019). They likely work with the Labour Force Surveys as well.
Packages that include shapefiles are available here:
If you find any errors, please let me know.
Please cite our paper if you use our data.
If you do not have access to such detailed geographic data, you may still wish to create consistent provinces as best possible. As many users of Vietnamese data know, the number of provinces has changed significantly since the late 1980s. In most cases the changing of provincial boundaries was either a splitting or aggregating of existing provinces as opposed to districts being reallocated between provinces. The province codes also change within surveys and across data sources. Here I post files for creating consistent definitions of province codes across various datasets. Please email me if you notice any corrections I should make.
Consistent province codes across the 1989, 1999, and 2009 Population and Housing Censuses
Consistent province codes across the 2002-2010 VHLSSs and the 2000-2010 enterprise data
The 1992/93 Vietnam Living Standard Survey (VLSS) used a different set of industry codes than the subsequent household surveys, the 1997/98 VLSS and the 2002, 2004, and 2006 Vietnam Household Living Standards Surveys (VHLSS). The industry codes in the 1992/93 VLSS were based on an adaptation of revision 2 of the International Standard Industrial Classification (ISIC) whereas the latter four household surveys used industry codes based on an adaptation of ISIC revision 3. Below I have uploaded two documents. The first provides a description of where the industry codes listed in the surveys deviate from the ISIC nomenclature. The second provides a mapping between the two sets of codes. Please email me if you notice any corrections I should make.
Inconsistencies in industry codes
Concordance between 1992/93 VLSS industry codes and 1997/98 VLSS through 2006 VHLSS
Below, I have also uploaded copies of the industry codes as updated in 1993, 2007, and 2018. To my knowledge, VSIC1993 is the basis of industry codes used in the 2002 through 2006 VHLSSs, the 2000 through 2007 enterprise data, and the 1999 population census. VSIC2007 is used in the 2008 through 2018 VHLSSs, the 2008 through 2017 enterprise data, and the 2009 population census. VSIC2018 is used in the 2019 population census.
VSIC1993 (based on ISIC rev3)
VSIC2007 (based on ISIC rev4)
VSIC2018 (based on ISIC rev4)
I have made some attempts at concordances between the various VSICs:
3-digit VSIC2007 to 3-digit VSIC1993
4-digit VSIC2018 to 3-digit VSIC1993
5-digit VSIC2007 to 2-digit VSIC1993
2-digit VSIC2007 to 2-digit VSIC1993
VSIC 2007 - VSIC 1993 conversion
VSIC 2018 - VSIC 2007 conversion
There is a household panel component between the 2002 and 2004, 2004 and 2006, and 2006 and 2008 VHLSSs. In each panel there are some mistakes in the originally provided matches suggested by the GSO. In most cases these matches are easy to detect by comparing the gender and year of birth information of suggested matches within the household. This leads to the conclusion that some households are incorrectly matched and that some individuals within households are incorrectly matched. Using additional confidential information, I have created my own revised versions. I have made the most changes to the 2002-04 VHLSS panel. Please see here for a discussion of the likely improvement in matching for the 2002-04 panel. Please cite McCaig and Pavcnik (2015) "Informal employment in a growing and globalizing low-income country," American Economic Review, 105(5), pp. 545-50, which was my first published paper using these revised panels.
2002-2004 VHLSS household and individual panels (revised 03 July 2013)
The 2002 household identifier was created according to:
gen double hhid02 = xa*10^5 + hoso
The 2004 household identifier was created according to:
gen double hhid04=tinh*10^9+huyen*10^7+xa*10^5+diaban*100+hoso
Household panel 2004-2006 VHLSS (revised 9 August 2010)
Individual panel 2004-2006 VHLSS (revised 2 July 2013).
Household panel 2006-2008 VHLSS
Individual panel 2006-2008 VHLSS
The 2006 household identifier was created according to:
gen double hhid06 = tinh*10^9 + huyen*10^7 + xa*10^5 + diaban*10^2 + hoso
And the 2008 household identifier was created according to:
gen double hhid08 = tinh*10^10 + huyen*10^8 + xa*10^6 + diaban*10^3 + hoso
2010-12 VHLSS household and individual panels
2012-14 VHLSS household and individual panels
2014-16 VHLSS household and individual panels
2016-18 VHLSS household and individual panels
These zip files contains the cleaned household and individual panels, and a short document describing how the panels were constructed and cleaned. Please cite McCaig and Pavcnik (2021) when using this data.
"Comparison of income between the 'short' and 'long' samples of the 2002 VHLSS"
"Comparison of income between the 'short' and 'long' samples of the 2004 VHLSS"
My paper with Dwayne Benjamin and Loren Brandt provides household income estimates for 2002 through 2014.
In a series of papers with various coauthors, I have used the change in U.S. tariffs on imports from Vietnam as an export shock. Here I have provided these tariffs at the 2-, 3-, and 4-digit industry level based on International Standard Industrial Classification revision 3. Please see "Exporting out of poverty: Provincial poverty in Vietnam and U.S. market access" for details on how the industry-level tariffs were created.
2-digit tariffs (Please cite "Export markets and labor allocation in a low-income country")
3-digit tariffs (Please cite "Exporting out of poverty: Provincial poverty in Vietnam and U.S. market access")
4-digit tariffs (Please cite "FDI inflows and domestic firms: Adjustments to new export opportunities")
In McCaig and Pavcnik (2021), we construct a panel of businesses, mostly informal, run by households as reported in the 2004 through 2018 VHLSSs. You can use our business panel by following these instructions: Business panel replication. Note that the VHLSS data must be purchased from the General Statistics Office. We ask that you please cite our paper if you use our dataset.