Life expectancy at birth
Documentation – version 9
Summary documentation of v9
Period 1800-1970, main source: v7, by Mattias Lindgren
We use the data from our own version 7 below, which is based on 100 sources compiled by Mattias Lindgren, who also assessed the fatal impacts of the biggest disasters in history and made rough guesstimates showing in the charts as sudden dips in life expectancy.
Period 1970-2016, main source: IHME
We used Global Burden of Disease Study 2016, from IHME, the Institute for Health Metrics and Evaluation, University of Washington, Seattle, published September 2017. Accessed October 7, 2017. Data for every five years can be found in the Appendix Table 13A published with this Lancet article, zip file downloaded here using the file named IHME_GBD_2016_MORTALITY_1970_2016_APPENDIX_TABLE_13A_Y2017M09D14.XLSX which has data for every five years. Annual estimates were downloaded from the GBD results tool: https://vizhub.healthdata.org/mortality/results.
Period 2017-2099, main source: UN
We used UN forecasts from World Population Prospects 2017 published in the file with Annually interpolated demographic indicators, called WPP2017_INT_F01_ANNUAL_DEMOGRAPHIC_INDICATORS.xlsx, accessed on September 2, 2017.
How these three sources were combined
Whenever IHME has data we use it, which is for almost all countries and years between 1970 to 2015. the UN source has data between 1050 to 2099, but before 1970 we use Gapminder’s version 7 instead, because the UN data is “smoothed out” and doesn’t show clearly which year a dip in the curve occurred. For years before 1970, we have taken Gapminder v7 and adjusted it during 50 years so that it meets IHME data the first year it exists, as in this example for China:
After 2016, we have extended the IHME series with the UN forecasts, as you can see with the Nicaragua graph below. But we have extended it not with the actual UN number. We extend it with the UN expected change. In the example of Nicaragua IHME estimate Life Expectancy in 2016 to be 78.2 years. That is almost three years higher than the UN estimate: 75.4 years. this means that we use the expected increase from UN for year 2017 and beyond, but we apply it to the higher estimate from IHME.
When we show data for 2017, that number doesn’t come directly form IHME, but it’s the 2016 number from IHEM with the percentage change in UN forecast extending it to 2017.
In many cases the new estimates from IHME improved a lot from the previous releases. This means that many countries moved there position in our chart when we updated the data. changed drastically because they improve their method of modelling HIV and other diseases. In previous Gapminder v7 we used the IHME 2015 data, which had a life expectancy for Botswana on 48.7 in 2015. The new IHME data puts it on 64.6, which is one of the largest changes. This shows up in this example, where you can see the blue line of Gapminder v7, far below the updated lines.
Global and regional averages
To global total life expectancy is using the IHME numbers from 1970 to 2015. For years before and after this period, we calculated the average weighted by population. The proper way to calculate it would be to account for the number of people of different ages, but that would require much more work. Our weighted average, 2016: 72.9 years, is extremely close to the properly calculated number form IHME, 2016: 72.9 years. Regional averages were also calculated weighted by population size.
Detailed documentation and feedback
For transparency we provide the files we used to calculate this data here, but we haven’t had time to clean them up and document all the steps in the process, see our data method overview. Any questions about the data and suggestions for how to improve it are always very welcome: in our data forum.
» Go to the google-spreadsheet that contains a copy of the data
» Go to the google-spreadsheet that contains a copy of the data, including projections
» Download reference files for each country from Prof. James C Riley’s bibliography for Average life expectancy at birth (zip-archive)
» Visualise indicator (Line qraph with colors illustrating the data quality).