Methodology

Internet affordability determines how much time people have to work to afford a stable internet connection. The internet’s affordability directly impacts its accessibility. A more affordable internet has a positive effect on digital wellbeing, while low affordability has a negative effect.

Indicator 1: Work time it takes to afford the cheapest mobile internet (seconds) (50% weight of the pillar)

We determined the factor value of mobile internet affordability by dividing the price of the cheapest mobile data package by the average hourly wage in a specific country. Then, we derived the index value by dividing the value of the most affordable mobile data package globally by the country’s mobile internet affordability value. It was then multiplied by 0.5 to weigh it in the internet affordability pillar, which consists of two indicators.

In the DQL'23 edition, the mobile internet price data source was changed from Cable.co to ITU (the United Nations specialized agency for information and communication technologies) to compare mobile vs. fixed internet pricing more accurately. Cable.co pricing includes not only 1GB of mobile data but also texts and voice calls, while ITU gives data-only broadband (minimum speed of a broadband connection is 256Kbps, relying on 3G technologies or above) mobile package pricing.

Indicator 2: Work time it takes to afford the cheapest fixed internet (minutes) (50% weight of the pillar)

We determined the factor value of broadband affordability by dividing the lowest broadband internet package price by the average hourly wage in a specific country. Then, we derived the index value by dividing the value of the most affordable broadband package globally by the factor value. It was then multiplied by 0.5 to weigh it in the internet affordability pillar, which consists of two indicators.

Internet quality measures how fast and stable internet connectivity is in a country and if it’s improving. The quality of internet connectivity greatly depends on its speed and stability. Slow and unstable connections inhibit daily use and diminish work efficiency, while fast and stable connections allow for better communication, high-quality content, and more.

Indicator 3: Mobile speed (Mbps)

To index global mobile internet speed, we first took the 12-month average of the mean monthly download speed in any specific country from April 2023 to April 2024. We then divided it by the highest global yearly average download speed value and weighed it by 0.2.

Indicator 4: Broadband speed (Mbps)

We measured the broadband speed index in the same way as the mobile speed index.

In the DQL'21 edition, the source for broadband speed data was changed from Cable.co to Speedtest.net as mobile internet speed is collected from Speedtest.net.

Indicator 5: Mobile internet stability (indexed value)

We determined mobile internet connection stability by comparing monthly download speed changes. We compared the percentage change in the mobile download speed of each month to that of the previous month. If the sum change during the 12 months was positive or equal to zero, we considered countries to have a stable internet connection and assigned them a value of 1. If the change was negative, we deduced the percentage modulus difference from the max value of 1. Then, we weighed the stability factor value by 0.15.

Indicator 6: Broadband internet stability (indexed value)

We measured the broadband internet stability index in the same way as mobile internet stability.

Indicator 7: Mobile internet speed growth (indexed value)

We determined the mobile internet speed growth by calculating the slope of mobile internet speed from April 2023 to April 2024. Then, we derived the index value by dividing the country’s slope value (Mbps/month) by the highest global slope value. If the slope value of mobile download speed was negative, countries were assigned a value of 0. Then, the growth factor value was weighed by 0.15.

Indicator 8: Broadband internet speed growth (indexed value)

We determined broadband internet speed growth in the same way as mobile internet speed growth.

Electronic infrastructure determines how well-developed and inclusive a country’s existing electronic infrastructure is. Highly functional e-infrastructure enables people to use the internet daily for many purposes, such as studying, e-commerce, entertainment, banking, etc.

Indicator 9: Individuals using the internet (per 100 inhabitants) (50% weight of the pillar)

We derived the internet use factor by dividing the number of individuals using the internet in a specific country by the number of individuals using the internet in the country with the highest value. Then, the factor was weighed by 0.5 as one of the two indicators of the e-infrastructure pillar.

In the DQL'21 edition, the source for individuals using the internet data was changed from the UN to Internet World Stats to have the latest data. In the DQL’24 edition, the source was again changed to ITU since the previous source is no longer updated.

Indicator 10: Network readiness

The Network Readiness Index (NRI) was determined by weighing the NRI value from the Portulans Institute (which hosted the Digital Transformation Dialogue Series) by 0.5. The NRI aims to evaluate how ready a country is to take advantage of the opportunities provided by information and communications technologies.

In the DQL'21 edition, the network readiness index data source was changed from the World Economic Forum to Portulans Institute to have the latest data.

Electronic security measures how safe people are online. E-security shows a country’s readiness to counter cybercrimes and its commitment to protecting online privacy.

Indicator 11: Cybersecurity (index)

To account for cybersecurity in the examined countries, the National Cyber Security Index (NCSI), developed by the e-Governance Academy Foundation, was used for its viability and the extent of data inclusivity. The NCSI index was weighed by 0.5 as one of the two indicators of the e-security pillar.

In the DQL'21 edition, the source for cybersecurity index data was changed from ITU to NCSI to have the latest data.

Indicator 12: Data protection laws (indexed value)

We indexed the data protection laws indicator by assigning a value from 0 to 5 depending on the existence and completeness of legislation regarding personal data protection in a country. Data protection quality was benchmarked against the EU's General Data Protection Directive as the best example of personal data protection legislation.

The values were assigned in the following order: 0 — no specific laws or data (none); 1 — some data protection law(s); 2 — independent authority and law(s), 3 — partially adequate laws, 4 — adequate laws, 5 — GDPR-level data protection. To index the assigned values, we divided them by the highest value of data protection (5) and weighed them by 0.5.

In the DQL'21 edition, the source for data protection level was changed from Commission Nationale de l'Informatique et des Libertés (CNIL) to UNCTAD and back to CNIL in 2022 to have the latest data.

Electronic government determines how advanced and digitized a country’s government services are. Better e-government helps minimize bureaucracy, reduce corruption, and increase transparency within the public sector. It also improves the efficiency of public services and helps people save time, influencing the quality of their digital lives.

Indicator 13: Online Service Index (index)

To determine how user-friendly e-governments are, we used the United Nations Online Service Index (OSI) for its inclusivity. The OSI is a composite part of the UN’s E-Government Development Index (EGDI). The OSI is based on the extent of a country’s government’s online presence. To weigh up the OSI index, it was multiplied by a factor of 0.5.

Indicator 14: AI readiness (index)

AI readiness indicates a country’s capacity to harness the potential and efficiency of artificial intelligence. The indicator was created using the Government Artificial Intelligence Readiness Index 2023, which was developed by Oxford Insights and the International Development Research Centre. The index was normalized to fit the DQL logic by dividing a country’s AI index score by the highest AI index global score and weighing it by 0.5.