The UNIDO [color=rgb(0, 137, 201) !important]World Productivity Database compiled by Anders Isaksson at the UN Industrial Development Organisation. Contrary to my initial assumption (oh, bliss), this data does not refer to manufacturing, but to the aggregate economy. Coverage is from 1960 to 2000 for 112 countries. The website is mainly a tool to compute TFP, so you don't really get access to the 'raw' data.

Michael Clemens (CGD) and Lant Prichett (HKS) have produced an interesting alternative measure to per capita income/GDP: 'income per natural' — the mean annual income of persons born in a given country, regardless of where that person now resides. The data is a cross-section for 2000 and the related paper is [color=rgb(0, 137, 201) !important]here. I copied that data into an [color=rgb(0, 137, 201) !important]excel filefor ease of use.

The World Bank [color=rgb(0, 137, 201) !important]PovcalNet is an interactive tool to calculate poverty lines and compare them across countries.

[color=rgb(0, 137, 201) !important]Wikiprogress is the official platform for the OECD-hosted Global Project on "Measuring the Progress of Societies" and[color=rgb(0, 137, 201) !important]Wikiprogress.Stat allows users to upload their data and metadata, and to navigate through a robust database of progress indicators. Themes on the website include Ecosystems Condition, Human Well-Being, Economy, Social and Welfare Statistics and Peace. There's a wealth of indicators here (sometimes cross-sectional or limited to a few time-series observations) and the data sources are clearly identified. Available for download to Excel. [Thanks to Angela Costrini Hariche, OECD Development Centre and Statistics Directorate and Project Manager of Wikiprogress]

The World Bank [color=rgb(0, 137, 201) !important]Doing Business project 'provides objective measures of business regulations and their enforcement across 181 economies and selected cities at the subnational and regional level.' The raw data for these surveys (run from 2004 onwards but with varying coverage for individual countries) is available via [color=rgb(0, 137, 201) !important]summary reports, which can then be accessed in excel.

Many of the above are featured on the resource website [color=rgb(0, 137, 201) !important]Macro Data 4 Stata which homogenises several commonly used macroeconomic datasets and imports them into Stata. The project is run by Giulia Catini, Ugo Panizza and Carol Saadeand started uploading .dta files fairly recently. The library at present includes data from the Penn World Table and the Groningen Growth and Development Data Centre. The [color=rgb(0, 137, 201) !important]AAA Codes dataset looks particularly handy for anybody doing cross-country analysis  [thanks to Aid-man [color=rgb(0, 137, 201) !important]Nic Van de Sijpe for pointing me to this resource].Without wanting to sound patronising, I applaud anybody's attempts to make data more widely available, so congratulations to a new upstart called Google, offering access to some World Bank, Eurostat and US data on their [color=#089c9 !important]website. Don't try and google "Google data" as you won't find it that way ;-) This resource is useful primarily for their data visualisation tool - for individual variable country series can be graphed as lines over time, bars or with the use of maps [thanks to Paddy Carter at Bristol for the pointer].

Funded by the IADB, the Oxford Latin American Economic History Database ([color=rgb(0, 137, 201) !important]OxLAD) contains statistical series for a wide range of economic and social indicators covering twenty countries in the region for the period 1900-2000. Its purpose is to provide economic and social historians worldwide with a systematic recompilation of available statistical information in a single on-line source. The website also provides other resources including a long list of references, many of them in Spanish, and detailed discussion of the methodology of data construction. Downloads are in csv format.

A useful resource to learn about how macro data is collected (among other things): the UK Economic and Social Data Service (ESDS) has produced [color=rgb(0, 137, 201) !important]'Countries and Citizens: Linking international macro and micro data'. This is "an interactive training resource with online tutorials, activities, study guides and videos, designed to show how to combine socio-economic data from country-level aggregate databanks (macro data) with individual-level survey datasets (micro data). It comprises five units, each of which was written by a subject specialist and has been designed as a self guided learning resource. Though specifically for postgraduates and researchers, it may also be of interest to undergraduates." Unit 2 seems quite useful.

Country-Specific Macro dataJohn Muellbauer and Janine Aaron at CSAE run a research project on 'Structural Macro-Modelling of the South African Economy' ([color=rgb(0, 137, 201) !important]SMMSAE). They provide a number of indicators and indices they constructed, including FLIB (financial liberalisation) and trade openness indicators in excel/CSV format. The SMMSAE website also links to the papers they have written on macro-modelling for SA.
Marc Muendler at UCSD has brought together a number of useful tools for the analysis of Brazilian [color=#089c9 !important]data (and some data, too). This includes various price indices, sectoral FDI (1980-2000), tariffs and exchange rates.

The Institute for Applied Economic Research (Ipea) in Brazil provides a range of [color=#089c9 !important]macro data for the country and its regions. The link is for the Portuguese site, there's also an English version. [Thanks to[color=#089c9 !important]Manoel Bittencourt, Senior Lecturer at the University of Pretoria/South Africa, for the link]

Chinese macro and micro data: when researching provincial FDI I frequently made use of the [color=#089c9 !important]China Data Center at U Michigan. Much of the more recent data (primarily statistical yearbooks for various topics as well as provincial statistical yearbooks) is downloadable as Excel worksheets, whereas the earlier data is available in pdf format. There are also the China Survey Data Network and various census datasets. Researchers at Universities may find that their institution has forked out for the annual subscription fee and that they can access these data without additional cost. Statistical Yearbooks for 1996-2001 were freely accessible at the time of writing.

The Socio-Economic Database for Latin America and the Caribbean ([color=#089c9 !important]SEDLAC) provides statistics on poverty and other distributional and social variables from 25 Latin American and Caribbean countries, based on microdata from households surveys. [Masa featured the new site on his excellent [color=#089c9 !important]Devecondata website]


Cultural and Social Norms and Value, Faith and ReligionThe World Values Survey represents 5 waves of data from the early 1980s to the late 2000s, covering survey data on social norms and values from 87 nations. The [color=rgb(0, 137, 201) !important]data is provided in SPSS, STATA and SAS formats. Variables related to individuals' happiness, how they feel, what is important in their lives, qualities their children should learn etc.

Fulvio Castellacci and Jose Miguel Natera have created a balanced panel dataset for cross-country analyses of national systems, growth and development ([color=rgb(0, 137, 201) !important]CANA) hosted by the Norwegian Institute of International Affairs. The originality of this dataset (which draws on a variety of sources) is in that the gaps in the data have been filled, using a methodology of multiple (and repeated) imputations by two political scientists, Honaker and King (2010). I have not looked at the [color=rgb(0, 137, 201) !important]Castellaci & Natera paper describing the data construction and robustness checks in detail, but am a priori quite sceptical about imputations: these macro variables are likely to be integrated, so imputations could be rather misleading. On the other hand, missing data is a serious problem for a lot of the dimensions they consider: (1) Innovation and technological capabilities; (2) Education and human capital; (3) Infrastructures; (4) Economic competitiveness; (5) Social capital; (6) Political and institutional factors. There are a total of 41 indicators for 134 countries over the period 1980-2008. The data is in excel format and well-documented. I'd say keep an eye out for reviews and applications of this dataset.

Robert Barro and Rachel McCleary have compiled a cross-country dataset on the share of religious people in the population. "Adherence fractions of population are shown for 10 religion groups and non-religion (incl. atheists) in 1970, 2000, and 1900 (from Barrett)." Data is available for download in excel format from Barro's Harvard [color=rgb(0, 137, 201) !important]data page. His working paper page offers a considerable number of papers on the topic of religion and growth. [via Masa Kudamatsu's [color=rgb(0, 137, 201) !important]DevEconData blog]

Human Capital (i): Formal Education/Schooling
The seminal dataset on educational attainment, compiled by Robert Barro and Jong-Wha Lee (the 'Barro-Lee data'), is available from a new dedicated [color=#089c9 !important]website. The data is available for download in full for 146 countries by 5-year age group or 15 years, 25years, and over in 5-year intervals for the period 1950-2010 (in xls, csv, or dta format). The site also links to some previous versions of the dataset and other resources, including Soto and Cohen (2006) and a few select academic papers [Thanks to [color=#089c9 !important]Adrian Wood for the pointer to the new site.]

The Washington-based Education Policy and Data Center ([color=#089c9 !important]EPDC) "provides global education data, tools for data visualization, and policy-oriented analysis aimed at improving schools and learning in developing countries." They say they have "the world’s largest international education database with over 3.8 millon data points from 200 countries. The data comes from national and international websites including household survey datasets as well as studies and reports." This is not just macro data, but also household surveys and census data; another very useful thing they do is to provide Stata do-files to construct indicators from the hh data.

Mauro Caselli, Jörg Mayer and Adrian Wood have compiled a unique extension to the Barro-Lee (2001) and Cohen-Soto (2001) data on average adult years of schooling (attainment) using UNESCO data on literacy rates. Missing values are imputed based on a regression model investigating the link between average adult education and literacy rates in the available data and applied to countries where the attainment variable is missing but literacy rates are available.  Of the 133 countries covered, no imputations were needed for 95, imputations for some but not all years for 19, and imputations for all years for 19. The [color=#089c9 !important]link is for a zipped folder containing Excel and Stata files as well as detailed documentation. The data is applied in a [color=#089c9 !important]paper by Jörg and Adrian investigating the global impact of China's industrialisation on other LDCs' structural change. [Thanks to [color=#089c9 !important]Adrian Wood for making the data available.]

Marcelo Soto and Daniel Cohen have constructed a rival to the Barro & Lee gold standard of data on average years of schooling across 95 countries. From the abstract of their Journal of Economic Growth [color=#089c9 !important]paper (Vol.12(1), 2007): "We present a[color=#089c9 !important]new dataset for years of schooling across countries for the 1960–2000 period. The series are constructed from the OECD database on educational attainment and from surveys published by UNESCO. Two features that improve the quality of our data with respect to other series, particularly for series in first-differences, are the use of surveys based on uniform classification systems of education over time, and an intensified use of information by age groups."  [thanks to my man [color=#089c9 !important]Fabio Manca for pointing me to this resource].

Christian Morrisson and Fabrice Murtin from the OECD have constructed a historical [color=#089c9 !important]database (entry under 'A century of education') on educational attainment in 74 countries for the period 1870-2010 (decadal estimates), using the perpetual inventory methods before 1960 and then the above Cohen and Soto (2007) database. This data should be particularly interesting in combination with for instance the Maddison data.

Fulvio Castellacci and Jose Miguel Natera have created a balanced panel dataset for cross-country analyses of national systems, growth and development ([color=#089c9 !important]CANA) hosted by the Norwegian Institute of International Affairs. The originality of this dataset (which draws on a variety of sources) is in that the gaps in the data have been filled, using a methodology of multiple (and repeated) imputations by two political scientists, Honaker and King (2010). I have not looked at the [color=#089c9 !important]Castellaci & Natera paper describing the data construction and robustness checks in detail, but am a priori quite sceptical about imputations: these macro variables are likely to be integrated, so imputations could be rather misleading. On the other hand, missing data is a serious problem for a lot of the dimensions they consider: (1) Innovation and technological capabilities; (2) Education and human capital; (3) Infrastructures; (4) Economic competitiveness; (5) Social capital; (6) Political and institutional factors. There are a total of 41 indicators for 134 countries over the period 1980-2008. The data is in excel format and well-documented. I'd say keep an eye out for reviews and applications of this dataset.

Jerry Dwyer at the Federal Reserve Bank of Atlanta provides [color=#089c9 !important]data from his 2006 Economic Inquiry article with Scott L. Baier and Robert Tamura. This covers output, physical and human capital for 145 countries over a long time horizon (1831-2000); the data provides between 2 and 17 time-series observations per country, with an average of around 7. Additional variables of particular interest include average age and experience of the workforce, which allow for Mincerian wage equation-type analysis at the macro level. The data is provided in a neat excel file with additional information on variable definition and construction also provided (along with the article).

A collaborative effort by the IIASA World Population Program and the Vienna Institute of Demography (VID) has reconstructed [color=#089c9 !important]population data by Age, Gender and Level of Educational Attainment for 120 Countries over the 1970-2000 period. The authors use a method which 'backprojects' the past levels from 2000 data. The files are in excel format and there are a number of working papers with technical details, comparison with observed data, etc. [Thanks to my buddy and human capital wizard [color=#089c9 !important]Fabio Manca for the link]

[color=#089c9 !important]Rural and Urban Education data (1960-1985) by C Peter Timmer is available in Chapter 29, 'Agriculture and economic development', of the Handbook of Agricultural Economics, Volume 2, Part 1, 2002, Pages 1487-1546. The link above is for the IDEAS RePec entry of this article: this is a copyrighted publication, but if you have access to the Handbook through your library you can easily copy the data. The coverage is exclusively for developing countries (N=65), and the data offers average years of schooling per person over the age of 25 for the rural and non-rural areas. OECD data on the same topic should allow for the inclusion of developed countries in the analysis.

The World Bank [color=#089c9 !important]EdStats (hit Data Query link) provide access to UNESCO Institute for Statistics (UIS) data on education. It presents a large number of indicators for more than 200 countries since 1970. Indicators are organized by category, including Pre-primary, Primary, Secondary, Tertiary, Expenditure, Labor, Population, Teachers and Other. In February 2011 UIS launched [color=#089c9 !important]historical time series data for key indicators of school enrolment and completion (gross enrolment ratios, repetition and completion rates) covering pre-primary to tertiary education. They are reported on a roughly five-year basis since 1970 (some countries more frequently). As far as I could see most of this data ends in the late 1990s... but the other data provided by UNESCO ([color=#089c9 !important]UIS Data Center) begins at the same period - not sure why they didn't bring these together.

The Lynch School of Education at Boston College provides two unique [color=#089c9 !important]resources for comparative analysis of educational achievements: (i) the Trends in International Mathematics and Science Study (TIMMS), which "is the largest and most ambitious international study of student achievement ever conducted" and has data from 40 countries in 1995 and a partially overlapping sample for three more recent waves (next wave is 2011); (ii) the Progress in International Reading Literacy Study (PIRLS), which has waves in 2001, 2006 and 2011 (forthcoming), evaluating 150,000 fourth graders (9- and 10-year-olds) in thirty-five (2001) and fourty-odd (2006) countries. Some of these are middle-income countries (e.g. TTO, MAR, IND, IRN).

The same database provides IMF data on [color=#089c9 !important]public spending on education, from 1985-2000 for 147 countries (via Gunilla Petterson).  This module presents the IMF data on public spending on education from 1985-2000 for 147 developing and transition economies (excel sheets). There are two indicators in the module: (1) total public spending on education as a percent of GDP; and (2) total public spending on education as a percent of total government spending. The underlying data, in millions of local currency, are provided. The breakdown of total education spending into current and capital spending are provided when available.

Quite a number of years ago Aaron Benavot, now at SUNY Albany, and Phyllis Riddle, now at St Vincent College, PA, wrote an article entitled The Expansion of Primary Education, 1870-1940: Trends and Issues, which provides new estimates of [color=#089c9 !important]primary school enrollment rates for 126 nations and colonies from 1870 to 1940. The data is printed in the Appendix and can easily be imported into Excel. The article was published in the journal Sociology of Education, Vol.61(3), July 1988, pp.191-210. [via Masa Kudamatsu's [color=#089c9 !important]DevEconData blog]

Emma Smith at the School of Education, University of Birmingham provides a number of [color=#089c9 !important]resources and data links for educational and social research, including Afrobarometer, Asiabarometer, PISA and World Value Survey. Her website acts as a portal for all the sources of secondary data that are listed in her book ('Using Secondary Data in Educational and Social Research', OUP 2008), as well as providing links to new sources and current developments in the field of secondary data analysis.

[color=#089c9 !important]Human Capital Inequality, basically adjustments to the above Barro-Lee data, is provided on Rafael Domenech's website, covering 134 countries from 1960-1999, based on his work with Amparo Castello. This dataset was mentioned on the brilliant[color=#089c9 !important]DEVECONDATA blog.

The World Bank provides [color=#089c9 !important]GenderStats, which basically pulls out the relevant variables from the WDI database. Hit "Create your own query" to access the database. Education/schooling-related variables are often taken as a proxy for gender equality.

UNESCO has data on [color=#089c9 !important]literacy in their data centre, with data series beginning in the mid-70s or early 80s. There are also lots of variables on schooling, and public funding for schooling. Data on the number of illiterates per cohort is available [color=#089c9 !important]here for developing countries from 1970 in 5-year intervals.

The OECD provides access to [color=#089c9 !important]PISA data (Programme for International Student Assessment) for 2000 to 2009 (4 waves). The most recent data wave will be made availabe on 7 December 2010. The data is in SAS, SPSS or Text format and contains student, school and parent information/questions. This is for 30 OECD/high- and middle-income countries. There is a vast number of variables so you had better see for yourself. [via Gunilla Pettersson's [color=#089c9 !important]developmentdata.org]

Human Capital (ii): Health & Subjective Well-being
The MARA/ARMA (Mapping Malaria Risk in Africa/Atlas du Risque de la Malaria en Afrique) project has published [color=rgb(0, 137, 201) !important]extensive data related to malaria, including the MARA LITe malaria prevalence data, malaria distribution maps and estimated populations at risk (as 'raw data' and maps); also available are entomological inoculation rates and reported presence/absence of six species of the anopheles gambiae group (to you and me: mosquitoes) in Africa and islands. The website also features a wealth of resources on malaria in Africa.

The [color=rgb(0, 137, 201) !important]Global Health Observatory (GHO) database is the World Health Organization's main health statistics repository. You can find a range of health topics like mortality, the burden of disease, infectious diseases, risk factors and health expenditures. I had a quick look at the figures for 'Number of people (all ages) living with HIV' which provides full coverage of mortality rate estimates (i.e. extrapolation/interpolation, etc., distinguished by reporting confidence intervals) for 1990-2009 across a very large number of countries. [referred to in a [color=rgb(0, 137, 201) !important]paper by Paul Calu, World Bank, and Falilou Fall, Sorbonne]

[color=rgb(0, 137, 201) !important]Wikiprogress is the official platform for the OECD-hosted Global Project on "Measuring the Progress of Societies" and[color=rgb(0, 137, 201) !important]Wikiprogress.Stat allows users to upload their data and metadata, and to navigate through a robust database of progress indicators. Themes on the website include Ecosystems Condition, Human Well-Being, Economy, Social and Welfare Statistics and Peace. There's a wealth of indicators here (sometimes cross-sectional or limited to a few time-series observations) and the data sources are clearly identified. Available for download to Excel. [Thanks to Angela Costrini Hariche, OECD Development Centre and Statistics Directorate and Project Manager of Wikiprogress]

The [color=rgb(0, 137, 201) !important]Complex Emergency Database (CE-DAT) is an international initiative that monitors and evaluates the health status of populations affected by complex emergencies. CE-DAT is managed by the Centre for Research on the Epidemiology of Disasters (CRED), based at the School of Public Health of the Université catholique de Louvain in Brussels, Belgium. The data is at subnational level (building on over 2,000 surveys) and covers 1998-2010 (with gaps). It can be viewed in table format or as a map.

The WHO maintains [color=rgb(0, 137, 201) !important]WHOSIS (Statistical Information System) which has data on mortality, health services coverage, inequities in health care access among other rubrics. Time series begin in 1990 but are not annual.

The World Health Organisation (WHO) offers the [color=rgb(0, 137, 201) !important]Global Health Atlas. "In a single electronic platform, the WHO’s Communicable Disease Global Atlas is bringing together for analysis and comparison standardized data and statistics for infectious diseases at country, regional, and global levels... [The database covers] the major diseases of poverty including malaria, HIV/AIDS, tuberculosis, the diseases on their way towards eradication and elimination (such as guinea worm, leprosy, lymphatic filariasis) and epidemic prone and emerging infections for example meningitis, cholera, yellow fever and anti-infective drug resistance."

The World Health Organisation (WHO) offers the [color=rgb(0, 137, 201) !important]Global Atlas of the Health Workforce, which features two datasets: the first, aggregated dataset "includes estimates of the stock (absolute numbers) and density (per 1000 population) of health workers for up to 9 occupational categories." In the second, disaggregated dataset "estimates of the stock of health workers are available for some countries for up to 18 occupational categories, reflecting greater distinction of some categories of workers according to assumed differences in skill level and skill specialization".

The visualisation folk at [color=rgb(0, 137, 201) !important]Gapminder (including multiple Roslings) provide very convenient access to a lot of demographic and health data (HIV/AIDS, birthrates, cancer, ...) alongside other useful development data (aid, trade, employment). "Gapminder is a non-profit venture – a modern 'museum' on the Internet – promoting sustainable global development and achievement of the United Nations Millennium Development Goals... The initial activity was to pursue the development of the Trendalyzer software. Trendalyzer sought to unveil the beauty of statistical time series by converting boring numbers into enjoyable, animated and interactive graphics... In March 2007, Google acquired Trendalyzer from the Gapminder Foundation and the team of developers who formerly worked for Gapminder joined Google in California in April 2007." Poor chaps: New salary = googol*previous salary? The data commonly span several decades and are available for download in excel format (wide). [Thanks to Christoph Lakner at CSAE for the pointer.]