Endogenous categorization of the human development

ABSTRACT

The human development index (HDI) is one of the most well-known measures of welfare. We apply clustering techniques to endogenously determine how similar countries are with respect to the HDI, and into how many categories they can be classified. We find that, in contrast to the usual assumption in the United Nations’ Human Development Reports, the number of categories is not fixed and has varied over time, from three in 1990 to four in 2014. We also find that the countries within each category differ from the United Nation’s proposal.

KEYWORDS:

• Human development
• welfare
• cluster analysis
• agglomerative methods
• grouping

JEL CLASSIFICATION:

• I30
• D63

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Similar techniques have been applied to determine the formation of groups in other fields (see Lucotte, 2015, for example.

² At the beginning of the process, all groups are singletons of just one observation.

³ We consider these years for two reasons: the most recent data available are for 2014, and there are many missing values for dates before 1990.

⁴ This comparison is not so straightforward for 1990 since the number of levels is different.

Additional information

Funding

This work was supported by Ministerio de Ciencia under Grants ECO2011-29355 and ECO2014-53767-P; Network MOMA under Grant ECO2014-57673-REDT, Junta de Andalucía under Grants SEJ5980 and SEJ4941; and Grupo PAIDI under Grant SEJ426.