Productivity claims alone don't justify NBN

One of the murkiest and most problematic areas of support for the NBN is the implicit assumption that better broadband access will boost our national productivity. This, in turn, invariably seems to spawn problematic attempts to quantify these improvements — and conflict over what are often optimistic assumptions piled upon others.

An attempt to do just that has met with resistance from some New Zealand economists, as a Berl Economics report has suggested that the country's Ultra-Fast Broadband (UFB) broadband roll-out could boost national GDP by 7 per cent to 9 per cent by 2025 (that's not per year, but compared to growth over that time without the UFB). A report from the firm, weighing the GDP benefits of an early UFB roll-out to Wellington, predicted that such a roll-out "could add around 11 per cent to GDP growth by 2026", and that a delayed roll-out would pull this back to around 8.8 per cent. The report envisioned the benefit to Wellington as being higher than that of New Zealand as a whole, where it envisioned additional GDP of 8.9 per cent through 2026 in an early roll-out scenario and 7.2 per cent under a delayed roll-out.

These are ambitious numbers — and have of course been skewed towards the optimistic because they've been paid for by a special interest group. Consider for a moment that, in real terms, New Zealand's entire economy only grew at 3.7 per cent per annum between 1995 and 2000, and 1.5 per cent per annum between 2005 and 2010. For the UFB to be finished in 2013 and fuel an additional 11 per cent cumulative growth to New Zealand's GDP by 2026, it would have to contribute an average of 0.85 per cent additional GDP growth per year.

That's a high benchmark, especially in the context of a Victorian study — cited by the report's authors — that suggests investment in ICT boosts productivity by 0.23 per cent annually, equivalent to a 0.27 per cent GDP boost. Remember also that many organisations have already made their ICT investments, so we cannot assume that incremental benefits will be as large as when technology was first introduced.

Furthermore, that figure relates to all types of ICT, and not just access to broadband. So where, then, can we justify numbers like those used in Berl's latest report, which suggests Bay of Plenty dairy farms could boost their output by 16 per cent just by getting better broadband? That suggests a supercharged industry that could totally reinvent itself just by getting access to faster broadband; if the economy as a whole could expect this, then everybody would be rushing to boost their broadband.

For the UFB to be finished in 2013 and fuel an additional 11 per cent cumulative growth to New Zealand's GDP by 2026, it would have to contribute an average of 0.85 per cent additional GDP growth per year. That's a high benchmark, especially in the context of a Victorian study that suggests investment in ICT boosts productivity by 0.23 per cent annually, equivalent to a 0.27 per cent GDP boost.

The jump from dial-up to slow broadband is cited as an indication of the potential benefits — but many of these ventures have some form of broadband already; can we assume that a boost to broadband speeds will provide such a dramatic productivity boost again?

We face similar challenges of definition in Australia, where fevered debate about the NBN has been polarised between those who demand hard numbers to justify its substantial investment and those who, having taken the project's merits on faith, gild their statistical liberties with procedural assumptions to back their foregone conclusions.

Stephen Conroy, for all of his merits as a subduer of Telstra and an instigator of structural change, often falls into this trap: he happily espouses the productivity benefits of telework enabled by the NBN, for example, as though they are both massive and equally applicable across the entire economy.

Of course, Labor needs these kinds of talking points to justify its investment in the NBN. But while telework can be a real boon to many people, it's utterly irrelevant in many sectors of the economy. There also seems to be an implicit assumption that an increase in productivity necessarily translates to increases in GDP: help people work from home, and they'll work better, faster, smarter and more productively. This is directly reflected in repeated calls by Malcolm Turnbull — himself a fan of NZ's UFB — for the NBN to be reviewed by the Productivity Commission, where, he seems confident, it will fail on all conventional measures of productivity.

But can we always link improved productivity with economic output? Sure, analytical bodies have attempted to correlate the two — for example, "optimistic" OECD figures cited in the Berl report suggest that boosting labour productivity by 1 per cent can boost GDP by 1.54 per cent — but can it be reliably said that improving access to broadband will boost labour productivity?

Perhaps you have slow broadband at home, and fast broadband at work. You may be able to access information faster at work, but does that make you more productive? Or does it just give you more free time for coffee breaks, creative thinking, and extended lunchtime visits to the gym? And even if it does let you check emails and pore over PowerPoints faster, does that directly translate into a boost in GDP? Just because the NBN lets you spend less time in rush-hour traffic doesn't mean that you'll be spending that time on GDP-boosting activities; you may just use it to play with your young children, or write a song on your piano.

IBRS analyst Guy Cranswick, who has spent more than a bit of time contemplating these issues, is on the record disputing assertions that such boosts can be measured. "As the ABS does not measure it, it's not official," he points out, "so it can never be in official GDP figures."

Rationalising productivity

This is why productivity figures always have to be taken with a grain of salt: they're often forged in a hot oven of optimism, stoked with the flames of political expediency. They crack after being doused in the water of economic rationalism. If an investment will not directly translate to GDP, goes that theory — and it's a perennial favourite of Tony Abbott's Liberal ideology — it is not worth making, no matter how many incidental social benefits it provides.

In the harsh light of economic rationalism, big business has its own ways of boosting productivity — and they have nothing to do with broadband. Consider Qantas, which this week put 1000 families into dole queues in an effort to increase productivity by trimming what is seen, in management circles, as an oversupply of labour. This move will, of course, make the remaining employees more productive — but that's only because they are desperately scared of losing their jobs, too, and will work harder than ever to make sure that does not happen.

Somehow, one suspects that providing those employees with better broadband would have little benefit.

Figures from the Grattan Institute seem to support this idea: interestingly enough, comparing Australian GDP per hour worked with US GDP per hour worked suggests that we, as a country, are currently only around 85 per cent as productive as our American counterparts. Even more interesting: this figure is down from around 91 per cent as productive in the late 1990s.

In other words, despite increases in broadband and overall technology use over the past decade, by these measures we are actually less productive now than we were in 1998. Perhaps we have failed to take advantage of the technological developments of the past decade, as well as the Americans have. Perhaps, some have posited, we pay more for productivity-enabling technology so our net returns are lower. We must also consider that the Americans, punished by a struggling economy, are working themselves to death in an effort to keep their jobs.

Despite increases in broadband and overall technology use over the past decade, by these measures we are actually less productive now than we were in 1998. Perhaps we have failed to take advantage of the technological developments of the past decade as well as the Americans have ... If a bit of broadband hasn't boosted our productivity, then can more broadband do the trick?

If a bit of broadband hasn't boosted our productivity, then can more broadband do the trick?

My point is not to develop a grand theory of labour force productivity, however. It is simply to point out that productivity is a subjective thing, and cannot always be reliably traced to any single cause. And while most people concede that better access to information helps people work better, we must also ask the question of whether spending $X on speeding up their access to information will produce a commensurate benefit in terms of national productivity.

We must also ask whether this really matters. Perhaps, rather than holding ourselves to productivity forecasts that will inevitably prove too conservative or too optimistic, we should simply ask ourselves how the lack of the NBN is affecting our productivity. It may not matter whether you have 12Mbps or 100Mbps broadband, for example, but if your current dodgy broadband is preventing your business from functioning smoothly, many would see that as justifying the investment, even if you're not sure how much actual economic benefit you could obtain from it.

Here's the thing: nobody really knows how the NBN will affect national productivity. The bloody-minded Coalition wants numbers to support Labor's contention that the NBN will indeed boost productivity, and Labor is doing its best to boost its business case for the NBN by pointing to all of the things that it says will flow from the investment.

Will the NBN produce those benefits? Probably, but in different forms and to different extents than anybody might be able to predict now. And it may take longer than we think; input prices are rising, and those costs will have to be passed on, potentially impacting net GDP growth in the short term. At this point, the real effect of the NBN is everyone's guess, and onlookers who accept modest predictions will bristle at figures that seem overly optimistic — especially if they're trumpeted to support political goals.

Does this mean that the NBN should not be built? Of course not. There are many good reasons for upgrading our nation's core communications infrastructure and producing a workable, egalitarian and competitive environment to replace the skewed, unequal one that we have — but holding proponents of next-generation networks to arbitrary productivity predictions runs the very real risk of making everyone involved look like a goose.

What do you think? Do we need a hard economic case to justify the NBN? Are assumed productivity benefits enough? Or should we be moderating our investment until we can get some productivity forecasts that we all agree on?