OECD hits back at broadband stats inaccuracy claims

Stuart Corner
Thursday, 17 May 2007 06:42

IT Industry - Market

A good deal of the report focuses on broadband at speeds lower than 256kbps. The OECD makes every effort to ensure that the data we receive from national regulators adheres to our strict reporting guidelines. By way of background the OECD created the 256kbps benchmark to exclude services such as ISDN (128kbps). This was an important criterion for data collection, in the early provision of broadband services, but has been made largely irrelevant by developments in the market place. In France, for example, Market Clarity claim authorities collect data on services down to 64kbps.

However, broadband offers slower than 256kbps aren't even available in the market. France Telecom offers DSL services at 512kbps, 2Mbps, 8Mbps and 18Mbps. Smaller firms in France offer DSL services up to 28Mbps. Notwithstanding this the OECD regularly confers with operators to see if any sub-256kbps offers are available and if they would make a material difference to the reported data.

Market Clarity is inconsistent itself with the treatment of these connections. For example, the report creates an indicator limited to 256kbps and above that actually includes sub-256kbps connections for some countries (eg using FCC data at 200kbps for the United States) while excluding sub 256kbps connections for others (eg Canada, Norway and the United Kingdom). This breakdown offered by Market Clarity suffers from serious methodological and data consistency problems.

Finally, some of the findings of the report are highly dubious and are likely the result of computational errors and/or a flawed methodology. For example, the final table of the report (Table 10, right column) seems to be an attempt to summarise the broadband penetration rate strictly among households (excluding business connections). The results show New Zealand and Turkey as the two leading countries in the OECD in terms of households with broadband subscriptions. Surprisingly, Turkey's reported broadband penetration rate limited to households (62.8 percent) is almost double that of long-time broadband leader Korea (38.7 percent).

These results are in stark contrast to data on individual, household and business broadband penetration supplied to the OECD from official government sources. Perhaps Market Clarity was unaware that the OECD publishes these usage statistics in addition to subscriber data.

Many of these flaws could have been avoided simply by contacting the OECD and talking to us about our methodology and data sources. The result is a report that suffers from data problems and methodological errors, rendering it incomparable with the OECD's published statistics.