A.1. Annex 1: Methodology

The primary goal of the YDI is to provide an evidence base on the condition of youth around the world, focusing on opportunities for their development. The theoretical framework for the development of the YDI is derived from the work of Sen (1985) and Nussbaum (2000, 2003) on capabilities.

The 2023 iteration of the YDI has three main aims.

1. Assess the data context for updating the global YDI in 2023, including acknowledging the admission of Gabon and Togo into the Commonwealth.
2. Update the global YDI database of 27 indicators across six domains with more recent data, where available and taking into account additional countries for which data is now available.
3. Provide an analysis of the situation.

The calculation and indicators remain the same as in the YDI 2020 report (Commonwealth Secretariat, 2021). The YDI is designed to measure youth development based on six domains:

• Education
• Employment and Opportunity
• Equality and Inclusion
• Health and Wellbeing
• Peace and Security
• Political and Civic Participation

These domains, and the indicators within each domain, were decided upon through consultation with the YDI Expert Panel. To capture youth development within each country across all domains, 27 indicators were sourced. Table A1.1 presents the indicators selected to capture these domains and includes information on their sources and the year of their most recent update.

Imputations

The 2023 YDI’s methodology has been designed to be in line with other prominent global development indices, and substantial effort has been made to populate the index with the best available country data. However, consistent and comprehensive datasets covering youth development continue to remain scarce.

Effort has been made in the design of the methodology to include as many Commonwealth countries as possible, the required threshold for inclusion being 50 per cent availability of the needed data. Even with the high imputation threshold, for certain smaller-island countries there is simply not enough data available to justify their inclusion. Table A1.2 details data availability for these countries, indicating for which countries data coverage is above the threshold and for which additional data is needed.

Data availability has been a particular challenge for the Equality and Inclusion and the Political and Civic Participation domains. Constructing the YDI has therefore highlighted gaps in youth development data and the need for further improvement in data collection.

A major challenge to developing a harmonised composite index lies in attempting to overcome the significant variation in data across very diverse countries around the world, not just in terms of demographic and geographic characteristics but also in terms of socio-economic characteristics, which often times can affect data collection and quality.

The issue of data gaps is a common challenge to creating an index. The Organisation for Economic Co-operation and Development (OECD) et al. (2008) recommend several statistical techniques for dealing with data imputation to fill in data gaps. Table A1.3 lists the approaches used in the YDI. Using a combination of these techniques, the YDI represents the use of the best possible data without an overly complex methodology.

The banding process

In order to aggregate the incommensurable indicators, all indicators have been banded (normalised). This means each indicator is scaled to a score ranging between 0 and 1, relative to the initial global range. Appropriate minimum and maximum values are, therefore, chosen for each indicator so that any values below the minimum are assigned 0 and values above the maximum are assigned 1. All other values are scaled between 0 and 1, equivalent to their position in the original minimum-maximum range. Depending on the nature of the data, the banding process can take slightly different forms.

For example, for the literacy rate indicator, a higher score reflects a more desirable situation. Therefore, in this case, the banding process has assigned the largest data point a value of 1. Conversely, the lowest data point in the indicator has been assigned a value of 0, while all other data is scaled relative to these two points. This process is referred to as forward banding. On the other hand, a lower score on the mortality indicator reflects a more desirable situation. In this case, the data is reverse banded, so the lowest value is assigned 1 while the highest is assigned 0.

Therefore, for year y, Equation 1 calculates a forward banded score for indicator i. A reverse banded score is calculated using Equation 2.

An integral part of the banding process is to set appropriate minimum and maximum cutoff values for the banded scores. Some data has a normal distribution and therefore outliers can be easily defined as those greater than three standard deviations from the mean. However, other datasets do not follow the bell-curved distribution trend. A number of considerations are therefore essential in choosing the appropriate technique: the nature of the data, the distribution, the purpose of the index, the information to be conveyed and so on. When investigating global datasets for the YDI, very few can be classified as having a normal distribution. The presence of outliers defines the variance, skewing both the minimum and the maximum values. To account for this, IEP has set artificial minimum and maximum values to ensure that results are not too heavily influenced by outliers. Table A1.4 outlines the data distribution and bands for each indicator. The upper and lower bands are indicated when they are not based on the distribution of the data but rather have been input manually by IEP.

Weighting indicators and domains

Table A1.5 shows the indicators and respective weights applied in the YDI. The YDI assigns a higher weighting to three domains: Health and Wellbeing, Education, and Employment and Opportunity receive 22 per cent each, as these domains are considered key to youth development and data quality and availability are higher here. Equality and Inclusion is weighted at 14 per cent, while the Peace and Security domain and the Political and Civic Participation domain are weighted at 10 per cent each.

Within each domain, indicators are weighted by their importance relative to the other indicators in the respective domain. In some instances, they are weighted equally, indicating that they together comprise the core features of the respective domain and are equally essential. Across the three core domains, which comprise 66 per cent of the overall index, three indicators are considered primary: mortality rate, literacy rate and NEET. These primary indicators are weighted slightly higher than others in the index and therefore have a big impact on domain scores. In some cases, they grant countries a more pronounced domain score regardless of their overall rank in the YDI.

Aggregation and calculation

Once the data has been banded and weights have been assigned, the final stage is to multiply each banded indicator with its corresponding weight and add each country’s performance to arrive at an overall YDI score. Final scores are calculated by combining scores for the six individual domains into the overall YDI score, as demonstrated in Figure A1.1.