We all know that this government is pretty keen on building motorways – certainly more keen on them than investing in public transport it would seem. Generally this is justified on the basis that building motorways generates economic growth and development. If we look at Steven Joyce’s answer to a written question from David Shearer about why the government’s continuing with promoting the Puhoi-Wellsford “holiday highway” even though it’s cost-benefit ratio is only 0.8, the line that it will boost economic growth and development gets trotted out to “save the day”.
Question: Why is the Minister continuing to progress the Puhoi-Wellsford road of national significance when its cost-benefit ratio is only 0.8?
Answer Text: Traditional transport benefit-cost ratio analysis is based on calculations of accident cost savings, travel time benefits, vehicle operating cost benefits and journey time reliability. However,the objectives of all the roads of national significance projects are to grow the national economy by lifting productivity in New Zealand’s largest cities and surrounding regions. The Puhoi to Wellsford project’s objectives are:
1) To enhance inter regional and national economic growth and productivity
2) To improve movement of freight and people between Auckland and Northland
3) To improve the connectivity between the medium to long term growth areas in the northern Rodney area
4)To improve the reliability of the transport network through a more robust and safer route between Auckland and Northland.
Consequently, taking into account the wider economic benefits the project will deliver, the BCR is 1.1 at a discount rate of 8% rising to 1.5 at 6% and 2.0 at 4%. The traditional benefit-cost ratio calculation does not take account of these other benefits, including examining freight movement on the road and tourism benefits to the Auckland and Northland area.
The mantra of “boosting economic growth through building roads” was also referred to extensively in a speech given by Finance Minister Bill English today:
One of our goals is to ensure that New Zealand’s competitive industries are not hamstrung by infrastructure bottlenecks.
For this reason, the Government has increased its capital allowance by $7.5 billion over the next five years. Much of this is earmarked for infrastructure projects such as the roll-out of ultra-fast broadband, better hospitals and more modern schools.
In addition, the Government will spend $7 billion on improving State Highways in the coming five years and $3.3 billion to strengthen the national electricity grid.
Combined, this amounts to productive investment of about $3.5 billion a year for the next few years.
This is supporting thousands of jobs across New Zealand – quite apart from the wider benefits to the economy. This year, construction will begin on the Te Rapa Bypass in Waikato, the Christchurch Western Corridor and the Te Atatu-Lincoln section of Auckland’s Western Ring Route.
We will continue work on the Christchurch Southern Motorway, the Victoria Park Tunnel, the Kopu Bridge, the Hawkes Bay Expressway extension, along with others.
And this year we will complete the Manukau Extension and the Manukau Harbour Crossing in Mangere and progress electrification of the Auckland commuter rail network.
At least rail got a brief mention in there I suppose. But I come back to the main question – does building motorways really boost economic growth and development? I suppose that in the short-term it creates jobs for people on diggers, and certainly when unemployment is rising it is good for that to happen. However, that’s not really the point of spending public money on infrastructure projects or we’d have tonnes of people employed breaking rocks in half for the sake of it.
A lot of the justification for the link between building roads and economic growth was outlined in a Treasury “Facts and Issues” paper relating to infrastructure investment, released in September last year. That “Facts and Issues” paper used an OECD study into the relationship between transport investment and economic growth to provide the foundation for the claim that building roads and motorways generates economic growth. In general that OECD study makes little sense to me, but there’s a series of graphs quite near the end of the document which detail the “correlations” between investment in road (non-motorway), motorway and rail networks with economic growth for a variety of different countries. It makes for interesting, if somewhat complex reading:
Now, to make some sense out of this we can first look to find New Zealand on each of the three graphs (indicated by NZL). Each country has two bars on each of the three graphs – an orange bar and a black bar. As detailed in the “note”, orange bars indicate negative coefficients (non-correlation) while black bars indicate positive coefficients (actual correlation) across a number of different measures.
This means that for New Zealand we have three findings:
- For “roads”, there’s a “positive correlation” of about 0.5 and a negative correlation of about 0.2
- For “motorways”, there a positive correlation of about zero (you can’t really see it) and a negative correlation of 0.5.
- For railtrack length, there’s a positive correlation of about 0.65 and a negative correlation of about 0.05 or 0.1.
Now I am certainly not a statistician or an economist, so the numbers don’t mean altogether that much to me. But I have discussed this graph with people who really do understand what it means, and – to cut a long story short it means that while investment in roads and rail does correlate with economic growth, there are certainly no clear signs that investment in motorways does the same.
So the very same report which is consistently referenced as justification for pouring money into motorways to generate economic growth actually says quite the opposite – that to generate economic growth from investing in transport infrastructure what you should really be doing is upgrading arterial routes, rural roads and railway lines. How very interesting.
Processing...
The answer on the BCR is the perfect example of what I always say: BCR calculations are not the scales of justice, they are not impartial measurements. As soon as politicans don’t like the outcomes, they twist them until they fit. I will use this example the next time somebody tries to claim that something must be / cannot be built because of the BCR.
“What BCR will it have?”
“What BCR would you LIKE it to have, Minister?”
Building motorways through semi-rural areas near cities tends to result in a lot of lifestyle blocks and whatnot getting sold off (district plans permitting, I guess) as commute times decrease. Creating mini-property booms (aka “sprawl”) around motorways is, I think, the underlying reason why successive governments have built so many of them.
Maybe I can help explain how to interpret that OECD work.
Firstly, the best figure to look at is Figure 9 in that report. It is basically looking at the effect of different types of infrastructure spending on economic growth. So it models economic growth (I’ll abbreviate this as GROW) as a function of infrastructure spending, separating out electricity generation (ELEC), road (ROAD), motorway (MWAY), telephone line (TL), telephone subscribers (TS) and rail (RAIL).
So you have a model that looks something like GROW = c0 + c1*ELEC + c2*ROAD + c3*MWAY + c4*TL * c5*TS + c6*RAIL.
This is a basic regression model where the numbers c0,…,c6 are the coefficients. c0 gives the basic rate of growth, where c1,…,c6 determine the effect on growth of infrastructure spending of the different types. So focus now on c2, the coefficient of (non-motorway) road spending. If this is a positive number, money spent of roads will increase economic growth; if it is a negative number, money spent on roads detracts from economic growth (this is based on the idea that the money is there to be spent and would have been best applied in another part of the economy).
So Figure 9 gives their best estimates of what these coefficients are, with the blue dash showing their mean estimate and the vertical line showing the extent of the 90% confidence interval, so showing the range that they are pretty certain the coefficient lies within under that model.
The coefficients for ROAD and RAIL are both strongly positive, so investments there have shown themselves to have a positive effect on economic growth (c2 is a bit above 2 and c6 is a bit below 2 but given the uncertainty, are roughly the same), while the coefficient for motorway length is indistinguishable from zero so has had no noticeable affect on economic growth. The model I give here is far simpler than the one the authors use but gives the basic idea.
The authors actually tried several different models and estimated coefficients for each. Figure 10 PANEL A that you have reproduced here shows whether or not the coefficients for ROAD, MWAY and RAIL (called c2, c3 and c6, respectively, in my model) are significantly different from zero under the different models they tried where “significantly different from zero” means that the confidence interval does not include zero. This graph shows that for NZ, the coefficient for road length is positive in about .5 of the models, negative in about .1 of the models and near zero in the remaining models. That for MWAY is negative in about half of the models and near zero otherwise, while that for RAIL is positive about .6 of the time and barely ever negative.
That means that spending on rail infrastructure in NZ can most reliably said to have had a positive effect on economic growth, general roading is a near second (and according to figure 9, may have larger multiplier) whereas increasing the motorway length has been money largely wasted, for whatever reason.
Hope that helps. (and please have look at what “correlation” means – you’ve misused it horribly here!).
Thanks for that further detail. I realise that my understanding of how we get to the conclusions is pretty limited, but in the end those conclusions hold true: that building motorways clearly does NOT generate economic growth. Right?
Yes, those conclusions seem right for New Zealand. Clearly there is some strongly country specific stuff going on here. It may be that rail and road length are increased only when there is a clear need, while motorway length increases for non-economic reasons (Puhoi-Wellsford anyone?).
“How are we going to get city transportation into better balance? The difficulties are not technical; a
wealth of ingenious and practical technical innovation in mass transportation remains in limbo because
of lack of development and construction funds. And the funds are not forthcoming, not because we are
a poor country, and not because we have nothing to lose in losing the hearts of our cities, but because
we have gone bankrupt in another realm – in the realm of planning policy for transportation.
“Perhaps I have been too hard on our Epaminondas traffic commissioners. After all, they are only doing
what they know how to do, in a policy vacuum. New York’s Traffic Commissioner Wiley is reported in the
press as having said last week that his job is ‘to operate the street systems and highways as efficiently
as we can. My job’, he went on to say, ‘has nothing to do with over-all city planning’. Of course he is
wrong, because what he does is drastically shaping the future of the city. But in a sense, he is right too,
because there is no policy, no goal for the city behind his work, which is not his fault.
“How are we to get transportation planning policies which control the flow and number of vehicles to the
extent that they are not able to disintegrate and destroy the downtowns? How do we get policies which
emphasize and support superior, comfortable and speedy rail transit or express bus service into the
heart of the city, and promote the utmost convenience for bus use within the city’s heart? We will get
them only by recognizing, and recognizing in time, that our present course of destructive expediencies
for traffic is total folly, and, ultimately, total futility.”
Quotes from Jane Jacobs, Address to the New York State Motorbus Association, Nov 10, 1958!!!!
Yes 1958 – yet these comments are just as applicable to us here in NZ today. Have we learnt nothing over the last 50 years?
I have read Jane Jacobs’ works too. They are both enlightening and depressing. Enlightening in that she understands and recognises the causes of many urban problems far more effectively than just about anyone else. Depressing, because she was saying this 50 years ago and we are still stuffing our cities up.
In asking the question Jarbury you are lending weight to the assumption that economic growth is be all and end all. Is it?
Economic growth could occur on the back of severe environmental damage (Canterbury Plains). Economic growth might simply mean a further distribution of wealth upwards (Income averages rather than medians). Economic growth goes up with the number of car crashes!
Until we have useful statistics that differentiate between “good” and “bad” economic growth, pursuing increased growth is dangerous and empty goal.
That’s certainly not my assumption Scott. My question simply is to wonder whether the government is correct in saying that motorways promote economic growth. From this report it appears they are not correct.
I think it is fairly well established that *constructing* motorways leads to economic growth is as much as they create employment for workers, consume resources and materials, stimulate the banking sector though financing and generally contribute to the economy. Given that these projects cost anywhere for a few hundred million to several billion dollars, this economic boost is no doubt considerable… but then again the same could be said for building railways, hosptials, school or even spending a billion dollars digging holes and filling them with concrete.
I actually believe Joyce is smart enough and good enough of a businessman to know the above to be true but he simply doesn’t care, the big party backers and hard righters want it…
Scott, I agree that there are a lot of problems with narrow definitions of economic growth (like bad BCR estimates) but, like Josh says, these are the measures that supposedly decide whether or not projects go ahead. And on that score, motorways sem to have failed. It’s not true, though, that things like lots of car crashes really show up as a positive for economic growth in anything except the very short term – a car crash stimulates economic activity but that activity is being diverted away from more beneficial areas. The same goes for motorway construction as Nick points out.
I’m not sure that Joyce really has an idea of the benefits or otherwise of road building – “businessmen” can be tricked by magical thinking like the idea that somehow a big motorway to the north will alleviate problems up there just as easily as anyone else can. And it’s especially easy when a broad electorate likes cars and equates good roads growth and prosperity.
I noted that in the Parlimentary Questions for Answer on the CBT boards the indication was that the main beneficiaries of the proposed motorway north would be the forestry and tourism industries.
Considering that logs and timber products are most easily shipped by rail, and that tourists are the mainstay of NZ’s intercity rail network, shouldn’t we be looking at improvements to the existing rail line north instead?
Surely an NAL upgrade, the Marsden Port rail link and an Tranzscenic service to the Bay of Islands would be a lot cheaper and therefore have a much greater benefit for the cost.
Also one wonders if they must be referring primarily to domestic tourism, i.e. Aucklanders going up north for a long weekend. While this might benefit Northland isn’t going to have any impact on the national economy. For that we would need to attract and retain more international tourists, and I’m not sure a freeway is the best way to do that. A scenic rail service to Opua would do that better.
“I actually believe Joyce is smart enough and good enough of a businessman to know the above to be true but he simply doesn’t care, the big party backers and hard righters want it…”
I disagree. I think Joyce and Key really believe that this is what makes for a strong, economically powerful nation. They may (in secret) acknowledge the downsides of their course (fuel/car dependence, sprawl etc…) but consider that those are worth it.
If we’re really relying on the “tourist industry” and forestry to justify spending $1.4 billion on the Puhoi-Wellsford road I really do think Joyce is scraping the barrel to bring the cost-benefit ratio up to what it is.
Funding this highway might be good politics, but from the point of view of a transport economist, it’s rotten.
Here’s why. The merits of benefit-cost analysis are very simple: they allow a limited pot of money for new work to be allocated out on a reasonably consistent basis, so that no-one whose pet project doesn’t get funded can claim that the system is being “unfair”. It is true that B-C analysis does not do a good job with evaluating non-road projects such as rail investment, because the cast in the New Zealand approach is (and always has been) to concentrate on the benefits of road investment *to road users*, not the wider community. In the New Zealand system this focus is heightened because of the road funding mechanism, which ties what governments spend on the road network to what motorists and road users put in. The other virtue of benefit-cost analysis, is that its results tend to promote small but cost-effective projects on the roading system – and not the big-ticket items that the politicians like.
Which brings us round to the Puhoi project and for that matter Transmission Gully in Wellington. The gaping flaw in the government’s defence of the work, is that if they allow these extra ‘soft benefits’ for the Puhoi project, in the interests of consistency you then have to allow them for anything else which is popular but which can’t be justified on the current rules – which would include a lot of smaller projects as well – so, in reality, the case for Puhoi isn’t helped, especially when money is limited. It is a very expensive fix for what the B/C analysis says is actually not too big a problem.
Yes exactly Ross. Adding in the wider economic benefits needs to be done consistently for all projects, which means that you’d still expect them to end up being “ranked” the same. Perhaps bigger projects would have more significant wider economic benefits than smaller ones, although often it seems that because a route is only as good as its weakest link, the little projects that seek to eliminate the weakest links (the bottlenecks usually) are those that make the biggest difference.
At least the cost-benefit analysis of the CBD rail tunnel will also be analysing its wider economic benefits – which are likely to be huge.
That helped me a lot thank you I am very greatful