3 things to watch for when measuring broadband costs around the world

A Broadband Pricing League Table was recently released by cable.co.uk and subsequently cited in a post by a contributor to the World Economic Forum (WEF). The original release highlights the most and least expensive countries for fixed broadband. The study however, illustrates the problems of international comparisons. Here's why.

1. Averaging multiple broadband plans skews prices

Averaging multiple plans on offer results in the price of broadband being highly skewed. As per the publicly available methodology notes, “the most basic package at each speed level” was selected. In some instances, it is typical for operators to advertise the theoretical maximum speed per technology.

Let us take Sri Lanka as an example. The incumbent Sri Lanka Telecom (SLT) advertises all ADSL plans with download speeds of “up to 21 Mbps” and all fiber-based plans with download speeds of “up to 100 Mbps”. The difference in price is vast. The most basic ADSL plan is LKR 450.00 (‘Web Light’, approx. USD 2.89 without tax for 7GB / month) and the cheapest fiber-based package is LKR 1490.00 (‘Web Family Plus’, approx. USD 9.57 without tax for 90GB / month). Thus, entry-level pricing would be USD 2.89 (without taxes), but cable.co.uk reports this as USD 19.93.

If the purpose of the study is to compare entry-level prices, or, as the research analysts at cable.co.uk state, “the cost of getting online,” then why are prices being averaged?

When posed the question via email, consumer telecoms expert for cable.co.uk, Dan Howdle, said that various price points were averaged as there was no reasonable comparison for the “cost of entry”. However, a decision on what constitutes ‘entry level’ could have been reasonably drawn from the methodology adopted by the International Telecommunication Union (ITU) for example, where the selection is based on 1 GB/month at a minimum of 256 Kbps, This would have made for a fair and more realistic comparison.

2. The exclusion of taxes

The methodology used does not address the tax issue explicitly – the data sheet and example that is publicly available has no mention of it. Ideally, the ranking would consider the actual cost incurred by the end consumer. Among the South Asian countries alone, telecom tax percentages range from 0 to 32. This contributes significantly to the final price borne by consumers and ought to be factored in if meaningful comparisons are to be made on “the cost of getting online”.

3. Comparing dollar for dollar

The cost of a commodity in one country cannot be compared to the same commodity in another, based simply on dollar for dollar. The result would be highly misleading. Hence the need for principles such as purchasing power parity (PPP), the BigMac index, etc. The methodology adopted by cable.co.uk simply converts the average price of multiple fixed broadband plans on offer in each country to a USD value.

For further illustration, see Table 1 which presents the cable.co.uk ranking of South Asian countries based on an average of various broadband prices.

Table 2 presents a ranking based on the ITU ICT price basket methodology. The cost of entry-level broadband plans (one plan per country with a minimum 1GB / month at 256 Kbps, based on what is offered by the service provider with the largest market share) is presented as a percentage of Gross National Income (GNI) per capita (World Bank, 2016, Atlas method). The prices include the respective taxes in each country and are a better reflection of the cost incurred by consumers. This simple comparison levels the playing field and, at least makes it fit for comparison on a global scale.

Through this method we see that Maldives (ranked by cable.co.uk as the most expensive in South Asia) has the 2nd cheapest entry-level fixed broadband package among South Asian economies, after Sri Lanka. Nepal (which has the cheapest according to the cable.co.uk ranking) moves down to the 5th most costly in South Asia.

Indicators can be highly subjective and are often flawed because arriving at a one-size-fits-all composition is near impossible. Trade-offs are often made when designing indicators, that would entertain a lesser or greater methodological flaw. In this case however, it is clear that there were measures that could have been taken to allow for a more realistic comparison that would have increased the legitimacy of the ranking.