Broadband in The Age (Day 2)

Today’s broadband piece in The Age is on whether broadband is a public good.

From a taxpayer’s perspective, high-speed broadband is a high-odds gamble

Joshua Gans, The Age, 11th November 2008

YESTERDAY I argued that, for most people, the economic value of higher speed broadband just isn’t there. It is true, there are some parts of Australia where commerce and people are concentrated and in these areas, they value higher speed broadband and, not surprisingly, it has arrived.

But with the Federal Government proposing to spend $4.7 billion of taxpayers’ money so 98% of Australia has fibre-to-the-node connectivity, it is natural to query the likely public benefit.

It must first be acknowledged that the Government considers this an investment and expects a return, it does not consider it expenditure.

In my mind, it is unclear why and how the Government can earn such a return where private investors could not.

One possibility is that they intend to accompany the investment with a regulatory regime that restricts competition and thereby allows the National Broadband Network to exercise monopoly power.

However, it would seem that the Government would gain little from politically-unaffordable but universally-available broadband.

The more likely possibility is that it actually is expenditure, but the Government is justifying it on the basis of public, rather than private, benefits.

Both the current and the previous government argued that higher speed broadband investment would bring $30 billion in benefits to the economy annually. Telstra has also quoted that figure.

However, that estimate is based on a 2001 Accenture study that itself utilised US estimates on the value of basic broadband. That value involved moving from very low broadband availability to 90%; which is what we already have in Australia. So that value is likely to be a vast over-statement.

Moreover, Telstra (which often argues that it is best placed to assess the value of these types of investments) has recently claimed that not investing in a fibre-to-the-node network is costing $200 million per month in lost gross domestic product. While I am unsure of the basis for that estimate, what it does do is place the value of broadband at less than a 10th of the previously quoted figure.

That said, there may be additional benefits to increased broadband availability; such as in education and health care. The idea of a surgeon in Melbourne operating remotely on a patient in an isolated area is the picture of what could be provided. But does that require us to hook up all households in the country or just to ensure there is a high-speed connection to town centres?

Moreover, we know that Australia is a laggard in broadband speeds. Many countries have near universal availability of high-speed connections. Indeed, Japan and South Korea have connections with 100Mbps (more than five-times higher than the Government’s target). Surely we should be able to identify and measure the public benefits by looking at those countries?

The surveys show that those benefits simply are not there yet. To be sure, residents in those countries enjoy faster video downloads and more connectivity for internet gaming but these are not public goods. And it is true that small businesses in those countries do not have to worry as much about location if they require high-speed connections. But even here, a design export business can still operate, it just may not be able to operate from anywhere it chooses.

Some might argue that this scepticism represents short-term thinking and that the benefits will eventually arrive, that even if they are not currently defined, we will nevertheless be ready for them. But we have to remember who is footing the bill.

For starters, the Government’s proposed service will likely cost as much for consumers as existing broadband. Even today, over 2 million subscribers remain on dial-up, despite the widespread availability of ADSL. Their revealed preference is not to pay for that. Moreover, because you need a state-of-the-art computer to take advantage of higher speeds, many middle to low-income households will miss out (25% of all households do not even have a computer). The problem is, as taxpayers, they will still pay for broadband while others (including myself) will get the subsidised benefit. It is Robin Hood in reverse and, from that perspective, an unfair gamble.

Joshua Gans is an economics professor at Melbourne Business School. He maintains a blog on these issues at

10 thoughts on “Broadband in The Age (Day 2)”

  1. “Some might argue that this scepticism represents short-term thinking and that the benefits will eventually arrive, that even if they are not currently defined, we will nevertheless be ready for them.”

    Well exactly. 20 years ago, very few of us could have predicted the immense impact of that new-fangled thing called the Internet. And it ain’t over yet. It will be to our benefit to expand online access beyond the minimally adequate capacity we currently have.

    I’m sure we hsd a similar debate about horses and automobiles. After all, a horse is a perfectly adequate form of transport.


  2. Australia is never going to compete with China in manufacturing. We need to go high tech and be able to offer first class online services and technology to the world.
    Broadband seemed to be the only technology that got slower. Broadband was 10Mbit/sec technology when conceived, the fact that ISPs where allowed to call 256K broadband was a mistake.
    Give Australians a 100Mbit connection to each other and the world and we will do amazing things beyond your imagination. Forget your 20th centry thinking where we all drive a car to a concrete building and sit in a cubical all day. High speed data communications allows us to operate business and united people from anywhere. Its time people start looking forward to this countries and the world’s future.

    The government has no issues bailing out a failed car manufacturing industry for the sum of 6.5 billion but giving us the infrastructure for the future seems too hard for them.


  3. “It will be to our benefit to expand online access beyond the minimally adequate capacity we currently have.”

    12Mb/s from the existing ADSL2+ is not “minimally adequate” especially when, as Joshua points out, the bottleneck is the international links from Australia to the rest of the world, over which it is difficult to get more than 2Mb/s.

    There certainly are applications that require much higher speeds than that, which is why Australian universities have collaborated to sponsor development of AARNet with current backbone speeds of up to 10Gb/s. Some commercial applications in health, engineering design, movie production and the like could certainly also benefit from faster links than the domestic internet backbone supplies, but the small number of users like this are no justification for wiring the entire country at Gb/s speeds.


  4. “20 years ago, very few of us could have predicted the immense impact of that new-fangled thing called the Internet … I’m sure we had a similar debate about horses and automobiles.”

    The most extravagant predictions about the future are not always correct. We need to make a best assessment based on the data. That is what Josh is trying to do. Comparisons with the past are useful, but not conclusive.

    10 years ago, a surprising number of people thought the Internet would totally transform commerce and entertainment by 2008, wiping out activities like physical shopping and TV. 50 years ago, many people thought 2008 would be all about flying cars and/or regular flights to the Moon.

    At some point in the future, 100Mb/second broadband will no doubt become basic service. But timing matters. Josh is arguing that we not need ultra-fast broadband to every corner of the country right now, any more than we needed a six-lane highway to Sydney in 1910.

    “Give Australians a 100Mbit connection to each other and the world and we will do amazing things beyond your imagination.”

    A few nations have close to this connectivity already. What is interesting, as Josh points out, is how little is being done with it.

    You can argue that the scarce resource isn’t physical bandwidth – very 20th-century thinking, you could claim – but the creativity to make better use of the plentiful bandwidth we already have.


  5. oh dear, this is getting more and more technologically uninformed. You do _not_ need a state of the art computer to enjoy the benefits of 10 megabit broadband. 10baseT Ethernet cards have been around since before the turn of the century and any minimally specced computer, including an olpc, intel classmate, hacked 1st gen xbox, Asus eee, my old amiga 500 or indeed anything upwards of a 486 equipped with a video card and a sound card will happily connect to and do anything you would care to do on broadband Internet, excepting maybe the latest h.264/mp4 decoding tech.


  6. Another excellent article. What resonates with me the most about this issue is a story Josh mentioned a while back, at least I think he did. He talked about setting up ethernet cabling in his house only for it to become obsolete in a few years as wireless network technology improved.

    With technological improvements happening so fast who knows what the right decision will be? My guess is the private sector is better placed to make that call, rather than a government relying on reports from 2001.

    I’d be interested to see if the fast broadband available in Japan and South Korea extends to their rural areas (yes, they do have some rural areas).


  7. What a great article. People are far too quick to assign the label of a public good without rigorously checking the facts. Good to see that someone is bringing sense to these debates.


  8. There is another way of financing the development of the nation’s communications infrastructure that does not involve the government trying to guess the best technologies to support.

    Last year the M3 money supply increased by $170 Billion. This means that loans of $170 billion in new currency were created. Of course many more billions of loans were given using existing currency.

    One of the major areas where new loans were created was for housing. About $200 billion dollars in loans were given for housing in the last 12 months. Of these loans about 10% were for new houses and 90% were for existing houses. We have no way of knowing how much “new money” was created for housing loans because the way system works is that loans that are created using new money (that is money that is loaned without there being money on deposit) is not distinguished from money that is loaned from existing deposits.

    However, some new money will have been created and loaned to finance existing houses and this new money for existing houses would not have resulted in a new asset being created.

    When we create new money we would hope that the new money would result in new assets being created. However, we know this does not happen because we have asset inflation, general inflation and we have to import money into enterprise Australia.

    The theory and the hope in the way we run the economy is that somehow new money will result in new assets being generated. We know that this isn’t always the case and we know that much of the new money goes to inflate existing asset prices including houses. The theory of supply and demand says that if house prices increase then it becomes worthwhile for people to build new houses for lower costs and this will stop house inflation. This does not happen because the price of new land goes up, taxes increase, the size of new houses increases and house prices remain high.

    It doesn’t have to be this way and the simplest way is for the Reserve Bank to ensure that new money is invested in productive assets. Instead of giving the responsibility for the allocation of new money to the banks who tend to lend it to individuals and enterprises to purchase existing assets let us make sure that new money is spent on productive new assets. Let us use old money to purchase existing assets. We do this by the simple change in the banking rules and we take away the ability of banks to lend more money than they have on deposit. We do this because experience has shown us that they will lend new money in ways that tend to cause price inflation of existing assets.

    Of course the argument is that the Reserve Bank can’t work out the best way to allocate money as that is best done by a market. It turns out there is a way that the Reserve Bank can ensure that new money is invested in productive assets and where a market allocates the money.

    There is no reason that the Reserve Bank cannot create new money that it says has to loaned out at zero interest but it must be spent on communications infrastructure. We know about how much money is needed to be invested to give Australia a good communications infrastructure. Let the Reserve Bank create this much new money and divide it up equally and give it to every man woman and child in Australia who volunteers to accept it. However we require the money be invested in new communications infrastructure. Let us only require that the communications infrastructure interface to the existing infrastructure and let us remove most of the existing regulations surrounding communications infrastructure.

    We then sit back and let the market in communications infrastructure decide what is the most economically efficient way of providing communications.

    This will not cost the taxpayer anything because the money is newly printed and there will be plenty of private industry organisations willing to take on the challenge of building and developing the infrastructure to distribute the money and to ensure it is spent in the communications infrastructure market place.

    This will NOT cause inflation because the money remains out of circulation until is spent creating a productive asset (one that returns more than it costs). It will also turn out that if we create too much communications money then it will be inflated but it will not cause general inflation. This means that the Reserve Bank does not even need to be a good guesser. Of course we will have to create more new money than that required for communications infrastructure so let the Reserve Bank create some money for use in reducing greenhouse gas emissions, saving the Murray, helping build the education infrastructure, supplying health facilities.


  9. Nudd, let me get this straight. You want a low income household to pay $225 in contributions towards the NBN infrastructure, another $60 per month on connection fees and then another $500 while hacking a computer in order to watch videos on a 10 inch screen that they can download at high speed.

    Methinks you have the attitude of “let them eat cake.”


  10. Joshua my comments were directed solely at the “state-of-the-art” computer required to take advantage of high speed broadband, which is clearly wrong unless you define state-of-the-art as “some time well before 2000”.

    I made no other comments on the rest of your article, but I guess it would be safe to draw the inference that the “state-of-artness” of a computer is superfluous to your argument. if you can afford a home pc in this day and age, then you can take advantage of broadband, and if you fall into the lower 25% who cannot, then obviously you have no computer, let alone a state of the art one. My personal view would be to try to do something about the income inequality that this stat indicates.


Comments are closed.