Surveys are typically published every two to three years, but they can be irregular and infrequent in many regions. To estimate values, a multilevel, nonparametric modeling approach developed by the World Health Organization to estimate clean fuel usage was adapted to predict electricity access and used to fill in the missing data points for the time period between 1990 and the latest year. Where data are available, access estimates are weighted by population. Multilevel nonparametric modeling takes into account the hierarchical structure of data (country and regional levels), using the regional classification of the United Nations.

The model is applied for all countries with at least one data point. In order to use as much real data as possible, results based on real survey data are reported in their original form for all years available. The statistical model is used to fill in data only for years where they are missing and to conduct global and regional analyses. In the absence of survey data for a given year, information from regional trends was borrowed. The difference between real data points and estimated values is clearly identified in the database.

"High-income” countries defined by the World Bank income classification are assumed to reach universal access for the years the countries belong to the category.

In the present report, to avoid having electrification trends from 1990 to 2010 overshadow efforts since 2010, the model was run twice:

• With survey data + assumptions from 1990 to the latest year for model estimates from 1990 to the latest year
• With survey data + assumptions from 2010 to the latest year for model estimates from 2010 to the latest year