The True Cost of the NBN

Senator Conroy stated in a press release today that the latest OECD statistics, available here, reveal that Australians pay relatively high prices for the Internet compared to other countries.  That is certainly true for connections below 2.5mbps and less true for what he defines as “high speed connections” — that is, connections between 15 and 30mbps — where Australia’s prices are the 14th highest out of 33 countries measured.  As ABS statistics show, this is probably a more important measure, with 73% of Australians accessing speeds between 1.5mbps and 24mbps.

Nevertheless, there is no doubt our broadband performance is not good enough relative to other countries and there is an especially important concern about those who are stuck on speeds of less than 2.5mbps.  Everyone agrees they should be brought up to speed as soon as possible.  But rather than investigating the cheapest and quickest way to do so, Senator Conroy will build an entirely new network asking some residents to wait until as late as 2020 before they are upgraded.

However, Senator Conroy bypasses all logic by declaring:

“The NBN will provide Australia with world-class broadband infrastructure. It will open up a genuine choice of services and drive competitive prices for consumers, whether they live in a capital city or in regional, rural or remote areas.”

Economics 101 tells us that state-owned monopolies are not the most efficient models to “drive competitive prices” as Senator Conroy assumes.  In fact, the main reason the Government has legislated against Telstra and Optus using their HFC networks to compete against the NBN and other operators from “cherry-picking” NBN customers is because it fears that facilities-based competition will deliver lower prices in the city, undermining the NBN Co’s ability to charge higher prices (economist Josh Gans explained on The Drum why protecting a cross-subsidy model as proposed by the NBN leaves everyone worse off).

But this is not mere theory.  As the NBN Co Corporate plan shows (available here), more than 70% of users will opt for speeds of less than 25mbps (p.129) with only a small movement up the speed chain by 2020.  For people purchasing lower speeds, prices are forecast to stay steady in nominal dollars (see p.101 and graph below) while people willing to pay higher prices will see their costs come down over time:

There are many things wrong with this.  First, as economist Robert Kenny has pointed out (read his paper here), the Government has not made the case of why lower-income households should be subsidising higher income households, presumably with lots of kids, who can afford multiple devices in multiple rooms all of which are using high bandwidth applications simultaneously. These are not the productivity-enhancing e-health, e-learning and business applications the Government likes to talk about but, in fact, mainly entertainment.  I note that p.130 of the corporate plan outlines what applications will be used by households with a 50mbps plan — it assumes they will have two smartphones, use advanced internet, engage in casual gaming, video calling, online storage and IPTV, presumably all at once to chew up that bandwidth. Elsewhere Mr Quigley has stated, as is the fact, that it is video applications which are driving bandwidth demand.

But most concerning, however, is that the NBN will reverse a long-term trend of very rapidly falling prices.  OECD statistics show that between 2005 and 2008, DSL prices in Australia fell by 45% (because competitors had been kept out of Telstra’s exchanges for so long, you could argue that prices were artifically high in 2005 — but nonetheless, the figures reveal what happens when competition is introduced).

Why is all this so important?  Because for all of Senator Conroy’s rhetoric about eliminating the tyranny of distance, the biggest barrier to Australians taking up broadband is its cost.  ABS figures show (see table three) that there is a disparity in access to broadband between households in major cities and households in remote areas — 75% compared to 62% (the national average is 72%).  But the disparity is much greater when household income is taken into account: 94% of households earning more than $120,000 a year have access to the internet at home compared to only 43% of households earning less than $40,000 have access to the internet.

Rather than spending time putting out press releases about how the NBN will deliver cheap connectivity and drive down prices, there is one way Senator Conroy could convince everyone that the model he has chosen is indeed the best for consumers and taxpayers — submit the project to a rigorous cost-benefit analysis by the Productivity Commission.  Until he does so, his continued claims about the NBN should be treated with extreme suspicion.