Does intensification increase traffic congestion?

Earlier this week, I took a look at the relationship between congestion and density. I was investigating geographer Phil McDermott’s claim, based on some dodgy data comparing between cities, that increasing density would increase congestion.

Economists know that it is difficult to make inferences about causality using cross-sectional analysis. Simply looking at variations between different cities doesn’t allow you to form robust conclusions about how those cities got to where they are.

One of the ways in which economists seek to strengthen their understanding of causality is to look at changes over time. For example, if you observe that increases in density tend to be followed by increases in congestion, then that is stronger (although not necessarily conclusive) evidence that there is a causal relationship.

With that in mind, it is worth asking: How have congestion and density changed in New Zealand cities over time? Unfortunately, we don’t have enough data points to conduct a robust econometric analysis, but we do have enough to start painting a picture of recent changes. We can draw upon two relevant sources:

I’m going to focus on NZ’s three largest cities – Auckland, Wellington, and Christchurch – as two of the three experienced big increases in density between 2001 and 2013. (MoT hasn’t collected data on Hamilton and Tauranga for as many years.) If these increases in density coincided with rising congestion, it may be an indication that intensification can lead to increased congestion.

Here’s the data. It shows that density has risen 33% in Auckland between the 2001 and 2013 Censuses, 17% in Wellington, and a mere 3% in Christchurch:

Changing density in large NZ cities chart

And here’s the Ministry of Transport’s Congestion Index, which measures the average minutes of delay per vehicle-kilometre, relative to totally free-flowing conditions. This is a bit of an unrealistic comparison, as a 2013 NZTA research report by Wallis and Lupton shows. The only way that you can totally avoid all queuing or stopping at traffic lights is if there are no other cars on the road. So it wouldn’t be realistic to say that we could speed up the average Auckland trip by half a minute per kilometre. However, this is still a useful indicator for changes from year to year.

While the Index bounces around a bit from year to year, the overall trends are clear. Levels of congestion are flat or falling in Auckland and Wellington, which experienced big increases in density over the last decade, and rising in Christchurch, which hasn’t gotten denser. In particular:

  • Average delay for Auckland drivers was 25% lower in 2013 than it was in 2003
  • Average delay for Wellington motorists fell 5% from 2004 to 2013
  • Average delay for Christchurch drivers rose a staggering 31% between 2004 and 2010. Unfortunately, MoT’s monitoring seems to have been disrupted by the earthquake, but anecdotal evidence suggests that congestion has worsened since then.

MoT Congestion Index chart

In short, data on changes in density and congestion in New Zealand cities contradicts the notion that intensification will necessarily cause worse traffic congestion. If anything, it suggests that rising density may do the opposite, by making it more feasible for people to walk, cycle, or take public transport.

Do we need to treat this data with caution? Most certainly. As I noted earlier in the week, there are a number of omitted variables that influence congestion, such as such as changing consumer preferences, macroeconomic changes, and significant investments in both roads and public transport over the last decade. But it does suggest that wild claims about the negative traffic impacts of new apartment buildings should be taken with a significant grain of salt.

What do you make of this data?

Arguing for sprawl with “strategic misrepresentations”

A number of recent posts have taken a look at some of the “strategic misrepresentations” that people have used to argue for a sprawled-out, roads-focused Auckland. We’ve taken aim at some of the common fallacies, including:

A while back someone sent me an article by geographer Phil McDermott that really hits the trifecta of fallacies. He argues that building apartment buildings on arterial roads – precisely where they will have the best access to frequent public transport services on Auckland’s New Network – is a bad idea because it will lead to increased congestion on the roads.

McDermott’s argument is long on subjective judgments (young people may want apartments but old people downsizing from big suburban homes never will!) and short on quantitative analysis. Here’s his key piece of evidence that constructing apartments on arterial roads will inevitably lead to more congestion:

Congestion – the elephant in the apartment

That might be just as well because mindlessly boosting residential development on arterial roads promises simply to compound Auckland’s congestion problems.

We know higher densities are associated with higher congestion. Auckland’s geography means it already performs poorly on this count. The Tom Tom Congestion Index confirms this.

When the 2013 congestion index for 65 American and Australasian cities is plotted against population density (sourced from the Demographia website) Auckland sits among the worst performers – Vancouver, Sydney, Los Angeles, and San Francisco (Figure 2).

Figure 2: Population Density and Congestion

McDermott density and congestion

This is not a serious piece of analysis – it is an insult to econometricians. McDermott makes three elementary errors in this short excerpt alone.

First, he uses bad data that misrepresents levels of density and congestion in these cities. Matt has previously taken a look into the guts of the TomTom Traffic Index and found that it is not a useful measure:

It measures the difference in speed between free flow and congested periods. That means cities with lots of all day congestion there isn’t as much of a difference between peak and off peak times and therefore they get recorded as having less congestion.

Likewise, I’ve done some empirical work on population densities in New Zealand and Australian cities that has showed that Demographia’s statistics are similarly meaningless. Demographia measures the density of the average hectare of land in the city, rather than the density of the neighbourhood in which the average person lives. Nick has shown how badly these figures misrepresent the actual density of New Zealand cities:

Auckland Wellington Christchurch charts_Page_1

Second, McDermott omits important variables and makes inappropriate inferences about causality. While he observes a correlation between two variables, that’s hardly sufficient to prove that building apartments will increase congestion. The causality could very easily run the other way. For example, it could be the case that the presence of congestion creates an incentive for people to live closer to employment and amenity. If that’s the case, then McDermott’s preferred policy of banning apartment developments would make Aucklanders much worse off by preventing them from minimising their travel costs.

Another possibility is that the relationship between density and congestion is mediated through other factors. Both may be caused by a third variable that McDermott has omitted, or there may be an intermediate step between density and congestion. (Or, as noted above, the measures themselves might be rubbish.)

A while back, CityLab’s Eric Dumbaugh provided an excellent illustration of the complex nature of congestion. He looks at data on US cities and finds that higher congestion is associated with higher, rather than lower, levels of productivity:

As per capita delay went up, so did GDP per capita. Every 10 percent increase in traffic delay per person was associated with a 3.4 percent increase in per capita GDP. For those interested in statistics, the relationship was significant at the 0.000 level, and the model had an R2 of 0.375. In layman’s terms, this was statistically-meaningful relationship.

Dumbaugh congestion and productivity chart

Such a finding seems counterintuitive on its surface. How could being stuck in traffic lead people to be more productive? The relationship is almost certainly not causal. Instead, regional GDP and traffic congestion are tied to a common moderating variable – the presence of a vibrant, economically-productive city. And as city economies grow, so too does the demand for travel. People travel for work and meetings, for shopping and recreation. They produce and demand goods and services, which further increases travel demand. And when the streets become congested and driving inconvenient, people move to more accessible areas, rebuild at higher densities, travel shorter distances, and shift travel modes.

In light of these counterintuitive relationships, the simple two-variable OLS regression that McDermott is relying upon is almost certainly misleading.

Third, McDermott fails to recognise that people are less exposed to congestion in denser, mixed-use cities. It’s simple: when people have better transport choices – i.e. access to frequent bus services and rapid transit, and safe walking and cycling networks – it doesn’t matter as much that the roads are congested. Increasing Auckland’s density by constructing apartment blocks and terraced housing on arterial roads will make it easier for people to have those choices, because the arterial roads are where the frequent bus services under the New Network will go:

Frequency is freedom

Frequency is freedom

Furthermore, density allows people to be closer to where they want to go. I find it odd that McDermott (and others) underestimate the importance of physical proximity in cities, even as people are paying high prices for the privilege. Building more homes in the areas that are accessible to jobs and amenities will allow more people to choose proximity over long commutes. (Without preventing others from making a different choice.)

A question for the readers: Would you rather have a 40 kilometre commute travelling at 80 km/hr, or a 5 kilometre commute moving at 30 km/hr? Show your work…

Fact-checking hyperbolic arguments for more roads

Following the gridlock on the roads last Saturday, the NZ Herald published several perspectives on how Auckland should cope with disruption to its transport networks. Matt weighed in with an excellent piece on the need to build Auckland’s long-awaited rapid transit network, which would give people an alternative to congested roads. However, the Herald “counterbalanced” it with some arrant nonsense about the need for more motorways by University of Auckland associate professor (and prominent climate change denialist) Chris de Freitas.

I use the term “nonsense” for good reason. The article was rife with factual errors that undermined the points that it was trying to make. Let us count the mistakes.

One: Congestion does not cost the Auckland economy billions each year.

De Freitas contends that:

The cost to the region’s economy of traffic delays is estimated to be many billions of dollars a year, which does not include the mental anguish caused to frustrated and angry drivers.

He does not provide any citations for this figure. However, I am aware of the relevant research, including a 2013 NZTA research paper by Wallis and Lupton that found that a more realistic figure for the cost of congestion in Auckland was a mere $250 million:

Including all congestion cost components, we concluded that the costs of congestion in Auckland are approximately $1250 million per year when compared with free-flow conditions, or $250 million per year when compared with the network operating at capacity.

In other words, the only way we could achieve that hypothetical $1.25 billion saving in congestion costs would be to build a network far, far in excess of what is required to move vehicles. Furthermore, Wallis and Lupton’s estimates are derived using NZTA’s Economic Evaluation Manual procedures, which explicitly account for non-monetary values such as travel time and driver frustration. The actual financial costs of congestion are likely to be an order of magnitude lower – i.e. closer to $25-50 million. That’s just not a lot compared to Auckland’s regional GDP of $75 billion.

Two: Auckland is not adding a Dunedin worth of population every 3-4 years.

De Freitas asserts that:

Given that the region’s population continues to expand by the size of Dunedin every three to four years, the vulnerability to traffic snarl-ups will grow exponentially.

According to the most recent Census data, Dunedin has a population of roughly 120,000 people. Between 2001 and 2013, Auckland’s population increased by approximately 255,000 people, or roughly 21,000 people per year. For those who like numbers, that means one new Dunedin every six years, not every three years. De Freitas seems to think that Auckland is growing twice as fast as it actually is.

Furthermore, the Ministry of Transport’s Congestion Index shows that travel time delay actually fell by one-quarter between 2003 and 2013. This contradicts de Freitas’ claim that congestion will increase “exponentially” as population grows – why hasn’t it increased over the past decade?

Three: Rapid transit networks are well-suited for regions with natural choke-points.

De Freitas argues that geography is destiny, and that Auckland’s skinny shape makes it a natural for roads:

Public transport itself will not ease the region’s traffic crisis. Auckland’s geography, history and politics make it a unique case for infrastructure planning. Its long, thin shape led to the earliest transport routes developing along a narrow north-south axis. Strategic arterial roads followed this pattern.

He correctly observes that road networks become less efficient when they are forced through natural choke-points like harbours and portages. However, these choke-points actually make public transport more efficient, not less. Putting more cars on a single road causes congestion and makes that road less efficient, but putting more buses or trains on a single right-of-way increases efficiency by allowing them to share costly infrastructure.

Four: Auckland’s motorway network already has alternative routes.

De Freitas contends that the Auckland motorway network lacks redundancy:

The result is a highway system that is not yet part of a fully integrated network. It is linear with no alternative routes around major bottlenecks. Traffic that would want to bypass the city is forced through Spaghetti Junction, adding to the vulnerability of the system to gridlock.

He has apparently not noticed that NZTA has almost finished building a bypass of Spaghetti Junction at a massive cost of $3.6 billion – the Western Ring Route. Perhaps he hasn’t been out west in the last decade, but if he had he would have noticed the construction of SH18 and the Upper Harbour Bridge, major expansions of the SH16 causeway, and the in-progress construction of the Waterview Connection to link SH16 with SH20.

Do we have to cover the whole region in asphalt to satisfy the man?

Five: A major earthquake in Auckland is extremely unlikely.

De Freitas raises the spectre of a Christchurch-esque quake:

The region’s most strategic arterial roads are vulnerable during earthquakes. Older multi-span bridges and abutments along motorways such as around Spaghetti Junction would be most vulnerable to damage from ground liquefaction. Even minor damage to these would bring city traffic to a halt.

Now, I’m no geologist… but both of my parents are geophysicists who started out researching Auckland’s rocks. They do not believe that Auckland faces serious risks of earthquakes. Volcanoes are a stronger possibility, of course, but volcanic activity doesn’t cause soil liquefaction. Here is a map from the British Geological Survey of every major earthquake in New Zealand since 1843. Notice the total absence of any recorded earthquakes anywhere near Auckland. Unlike Christchurch, we are not close to NZ’s fault lines:

Source: BGS

Source: BGS

Six: More roads are not a good solution for disaster readiness.

De Freitas argues that more roads are needed to evacuate Auckland:

The vulnerability of a city is to a large extent a function of the adequacy of preparedness planning. How soon could Auckland be evacuated?

There is limited motorway access out of the isthmus that is the Auckland urban area, so there few alternative exits. Main feeder roads head for one major harbour crossing and easily become congested.

Some American cities that are vulnerable to regular natural disasters have tested the “more roads” approach to evacuation. So here is Houston, attempting to evacuate on one of its eighteen-lane freeways during Hurricane Katrina in 2005. Not a lot of people actually made it out of the city:

Houston-HW_600

We could devote endless hectares of increasingly valuable land attempting to repeat the same solution that failed Houston. Or, if we think that natural disasters are a serious risk, we could invest in disaster preparedness and civil defense to ensure that the city’s residents will still have access to food, water, and health care services, regardless of what happens. That’s likely to be a much more practical, cost-effective solution.

Finally: The Herald needs to get better at fact-checking, or print a retraction.

While de Freitas’ article was printed in the op-ed page, that is no excuse for its blatant errors and omissions. Auckland only has one newspaper of record, and its credibility and usefulness to its readers is undermined when it prints this sort of gibberish.

How traffic engineering standards can break our cities

Many of you will know me as just a facetious blogger who spouts off about random things from time to time. This is indeed true.

What is also true, however, is that in a previous life I worked as a transport engineer. While nowadays I work primarily as a transport planner and an economist (roles in which I have more of a policy focus), from time to time I still find myself getting down and dirty with the application of basic traffic engineering principles.

Over time these experiences have led me to form the opinion that the traffic engineering profession in New Zealand is, shall we say, pretty much broken. It’s broken largely because “standards” have been progressively used as a substitute for thinking. And we’ve basically chosen the wrong standards. We’ve chosen traffic engineering standards that 1) fail to acknowledge basic scientific/economic principles, such as minimum parking requirements and 2) prioritise vehicle mobility ahead of other more important socio-economic outcomes. Many of these standards have profoundly negatives impacts on our cities.

In this recent post Matt presented one case study of a road re-design in Washington, in which he contrasted the designs for a corridor put forward by traffic engineers versus transport planners. Needless to say the options presented by the latter (illustrated below) appealed more to all of us here at TransportBlog.

Washington Freeway replacement Planner options

Not only do the options developed by transport planners have greater aesthetic appeal, but three of them also provide for additional land use development within the transport corridor. Indeed, traffic engineering practises that are currently applied in New Zealand (and to a greater degree Australia) do not even stop and consider the value of land. Land is so dam valuable, and the efficiency with which we use it determines, to a large degree, the productivity and amenity of our urban areas.

It would not be too factitious to suggest that many traffic engineering standards seem to presume that land is free. It’s as is if there are dutch pixies at the bottom of the garden who are manufacturing land from the sea.

One example of such a standard is the concept of the “design vehicle”, which I will focus on for the remainder of this post. Of course there are many other examples of traffic engineering standards, such as minimum parking requirements, which have been discussed before on this blog and that also have hugely negative consequences. The reason I want to focus on the “design vehicle” concept is because it does not receive much attention. And also because it has a fundamental impact on so many things.

For those who are not familiar with the “design vehicle” concept let me briefly explain. The “design vehicle” is a phrase that typically describes the largest, heaviest (per axle), and/or least maneuverable vehicle that is expected to use a particular part of the road network. Naturally, the physical footprint required to accommodate this design vehicle subsequently defines most aspects of the physical road geometry, such as turning radii and pavement design. For this reason, the shape of our road networks is very much defined by the design vehicle that is chosen.

You can read up on some of the design vehicle standards recommended by the NZTA here. The design vehicle for the standard street is typically some form of medium rigid truck, such as what is commonly used to move furniture. I’ve illustrated the physical dimensions of this vehicle below.

Medium rigid truck

The choice of design vehicle can have a massive impact on the degree to which a particular road supports, or more commonly undermines, socio-economic outcomes in urban areas. Working with a large design vehicle effectively puts paid to the types of narrow lanes and tight intersections that are ubiquitous in European cities (as an aside I’m writing this from Amsterdam, having just traveled through Paris, Porto, and Barcelona). Here’s a photo of my bicycle and I in a narrow lane in Barcelona.

DSCN4527

Alternatively, if you can convince your traffic engineer to use a smaller design vehicle then you can reduce the physical footprint of the road network.

In my experience, however, the traffic engineering profession has developed a “gotcha” for anybody who dares suggest a smaller design vehicle be used. The “gotcha” is service vehicles, such as garbage trucks and emergency service vehicles. That’s right, the humble garbage truck, they argue, needs comprehensive access to every urban nook and cranny. In turn, our urban nooks and crannies are designed around the needs of the garbage truck, which is – perhaps needless to say – rather large.

In this way, it is actually relatively difficult to argue for tight lanes and turning circles in many new developments in Auckland, by virtue of the need to provide access for service vehicles. Now I would have less issue with this standard if someone, somewhere had actually sat down and considered what the benefits and costs of such a standard were. Typically, traffic engineering standards require more land, which is a cost. The benefit, I presume, is increased mobility. Hence it should be fairly straightforward to undertake some form of benefit-cost analysis of the regulation to work out

Now take a look at the photo below, which shows a relatively common streetscape from Amsterdam.

Compressed

You will notice a cycle lane running from the bottom left of the figure towards the right hand side of the figure. If you follow this cycle lane closely then you should be able to make out the back of a small vehicle that is parked in the cycle lane just outside the shop. This is, my friends, a rubbish truck.

Which brings me to the point of this post. Amsterdam, like many European cities, designs their urban areas to deliver a broad range of socio-economic outcomes, such as walkability. This in turn requires narrower lanes and tighter intersection footprints. In response, they have effectively had to “down-scale” their rubbish trucks.

In this way I think the traffic engineering profession in New Zealand and Australia has put the garbage cart before the community horse. More specifically, instead of designing the communities we want and then selecting the vehicles that can integrate with that design, we choose the vehicle first and subsequently design our communities around their needs. I suspect our approach is very, very economically inefficient insofar as it increases the physical footprint of the road network. Remember, in cities, space is always expensive!

Supply curves slope upwards. Demand curves slope downwards. Are we accounting for this?

In a number of recent posts I’ve taken a look at both the demand for new roads (tends to be lower than expected) and the cost to build new roads (tends to be higher than expected). This has led to a bit of trouble for road builders in many developed countries – it’s getting harder and harder to find good projects.

This wasn’t always the case. Fifty years ago, people were flocking to new roads, even toll roads like the Auckland Harbour Bridge. But the transport market has shifted over the last decade. I would argue that we’re seeing signs that the market for driving is saturated. New Zealand’s road networks are mostly complete, and they have already succeeded in attracting the highest-value trips. Further road expansions will, by and large, serve lower-value trips.

I’m not saying anything revolutionary here – just restating some concepts from Economics 101. It’s worth going back to the basics of supply and demand to consider what might be happening in New Zealand’s road markets.

First, supply curves tend to slope upwards. In the case of transport investment, this reflects the fact that it tends to be increasingly expensive to build more road capacity as the network expands. Building the first ten kilometres of road may be cheap – the next 1,000 kilometres, not so much.

supply-demand-1

Second, demand curves tend to slope downwards. Effective, what this means is that the people with the highest willingness to pay tend to be at the front of the queue. The first people to use a new road will tend to be the people who place the highest value on travelling. The people who come later – perhaps as the road is being expanded – will tend to value travel less.

supply-demand-2

So far, so good. In a typical market, the quantity supplied would be determined by the intersection of the two curves. If transport agencies applied similar thinking, they would seek to build roads to the point at which the marginal willingness to pay for more driving was equal to the marginal cost of the next road. That’s shown in the green circle below.

supply-demand-3

However, transport infrastructure is not a typical market – it’s provided free of charge by monopoly infrastructure builders that must estimate (rather than observe) the amount of demand for their products. And rather than specifying a downward-sloping demand schedule, NZTA’s Economic Evaluation Manual (EEM) simply presents a standardised average figure for the value of time for travellers. (More precisely, a set of different averages for different types of roads and types of users.) The result of that – shown in the graph below – is effectively a flat demand curve.

supply-demand-4

If estimates of future demand are based on the value of time for the average current user rather than the marginal new user, they will significantly overestimate people’s desire to drive on new roads. In theory, the EEM includes procedures to correct for this – principally, their application of the “rule of half” to scale down transport benefits for new users. In practice, a lot of analysis seems to be done on the basis that future users will value driving in much the same way as previous users. Furthermore, demand forecasts can be developed in a fairly crude way that doesn’t take into account people’s marginal willingness to pay for new road capacity.

The graph below shows this dilemma. Effectively, if we don’t take account of the fact that demand curves slope down, we run the risk of massively overspending on new roads.

supply-demand-5

But that’s just the theory – has this actually happened in reality? There is some evidence that it has. For example, the failures of private toll roads in many developed countries suggest that road-builders may not have accounted for people’s declining marginal willingness to pay for travel time savings. Broader support for this hypothesis is provided by the fact that vehicle kilometres travelled in New Zealand have flatlined over the last decade, in spite of major investments in new roads:

ResizedImage600304-FutureDemand-Diagram1

Why won’t it grow? We thought it would grow! (Source)

Can we expect the same thing to happen in public transport or cycling networks? Eventually, yes, it will. But right at the moment the fact that these networks are incomplete means that improvements may attract new users faster than expected. For example, the City Rail Link is a three-kilometre tunnel that will give Auckland Transport the ability to double train frequencies throughout the entire 100km+ network. Doing that will make rail a lot more useful throughout Auckland and attract new users who otherwise wouldn’t have taken the train. Recent improvements to PT seem to have had this effect – i.e. attracting users more rapidly than expected.

CRL Times Western Line

But that’s a story for another day…

Engineers versus … everyone else?

How do you define “optimal conditions” in a transport sense? The graphic below paints a stark difference in opinion on what constitutes “optimal conditions” for different professions:

LOS

Image from Patrick Kennedy, D Magazine, http://streetsmart.dmagazine.com/2014/11/07/age-of-enlightenment/

Now I’m both an engineer and an economist, so feel I have some insight into the motivation for whoever developed this graphic.

It highlights a very important issue: Engineers tend to measure performance using indicators that measure mobility, whereas economists tend to measure performance using indicators that measure value.

Now don’t get me wrong: I strongly believe that mobility has a value. I write this having just fanged up and down the northern motorway taking some visiting Australian guests to Tawharanui. But it’s not the only thing I value. Indeed, tonight I will take them to Fort Street for dinner at Ima’s.

I accept that what constitutes “optimal conditions” will change depending on the context, and that in some contexts mobility has a very high value. What I can’t fathom is why the transport/traffic profession have developed such simple and inflexible performance indicators to guide their work.

I’m going to say this right now: Level of service tells you *nothing* about what people value. Why? Because it’s measured independently of the costs associated with being able to move freely. We could, for example, enjoy great LOS if we bulldozed the entire city and replaced it all with twenty-lane motorways connecting to vast carparks, with the occasional office building or house dotted amongst the seas of asphalt. But I think it’s obvious that would be a really, really bad outcome for almost everything else that we value.

From an economic perspective, congestion in cities can be seen as a good thing. It’s an indicator that lots of people are using the city in lots of different ways – going to work, travelling to see friends or family, going shopping, visiting sports games or art museums etc. Congestion has costs, of course, but eliminating it entirely would be even more costly.

Rapid transit has passed the acid test

I recently ran across a New Zealand Herald article from 2000 on the region’s plans to start building good rapid transit infrastructure. (Which, as Patrick highlighted in a recent post, is exactly what is holding Auckland back relative to its peer cities.) I noticed three things from the article:

  • We’re still having to scrimp and save and struggle to get good public transport projects built
  • This is in spite of the fact that the projects that have been built (against the odds) have been runaway successes
  • Many of the people who were urging caution back then are still around, but they haven’t acknowledged the evidence and changed their position.

On to the article:

The North Shore busway, allowing buses to travel faster than cars, will be the acid test for Auckland’s grand public transport schemes.

Planners are pinning their hopes on around $1 billion of rapid transit services running every five minutes along dedicated corridors as one answer to congestion.

The $130 million busway, a carriageway alongside the Northern Motorway, is likely to be first out of the blocks. It is being eyed to see how it fares for funding in about three months – and how many people it will coax out of their cars when it starts picking up passengers in three to five years.

Of course, the Northern Busway wasn’t actually completed until 2008, and the rest of the plan is still a glimmer in Auckland Transport’s eye.

Stephen Selwood, then of AA and now heading the NZ Council for Infrastructure Development, was quoted extensively in the story:

The region’s Passenger Transport Action Plan set targets of doubling and tripling public transport numbers in several key areas by 2011.

Yet the Automobile Association’s northern regional manager, Stephen Selwood, is not convinced they will be reached.

“The key test will be the busway, because that is the one where we know there’s congestion and thousands of people go over the bridge. If we can’t make that one work, nothing will.”

What actually happened? Although the busway was constructed late, it worked like crazy. By 2012, actual patronage on the busway was almost double what the patronage forecasts indicated:

Busway Patronage vs projections Graph

More prognostications from Mr Selwood:

The Passenger Transport Action Plan’s market-share goals for the number of commuters headed towards the central business district range from 15 to 45 per cent, and Mr Selwood claims this shows an improved public transport system would cater only for a minority.

By 2012, public transport accounted for 44% of all motorised travel to the city centre during the morning peak. (Walking and cycling weren’t included in the data, unfortunately, but they account for a significant share of overall trips.) Since then the PT mode share has increased even further. Public transport, including the successful Northern Busway, has accounted for all of the net growth in city centre access since the 1990s:

CCFAS Modeshare 1990-2012

One last comment from Mr Selwood:

Auckland, with its traffic growing at 5 per cent a year, cannot ignore the motoring majority and a need for more roads, he says.

That might have been true back then. But it’s not true now. The most recent Census data shows that road traffic is growing at an anemic pace while all other modes are booming:

Census Auckland marginal transport user analysis

In short, Auckland has faced the public transport “acid test”, and it has passed, with flying colours. This is even more impressive in light of the fact that:

  • The key projects that have been undertaken, such as the Northern Busway and rail electrification, have often been finished far behind schedule. Rail electrification was supposed to be done in 2011, for crying out loud!
  • The successful Northern Busway hasn’t been followed with investment in other essential rapid transit projects, such as the (planned but not yet built) AMETI busway to the eastern suburbs and the Northwestern Busway on SH16.
  • Successive governments have spent billions on Auckland’s motorway network even after it became apparent that demand was flatlining.

In light of the results, I look forward to hearing the NZCID’s strong advocacy to stop building motorways and put the funding towards good public transport projects.

Is road funding a case of “survival of the un-fittest”?

In several recent posts I’ve taken a look at people’s revealed preferences for roads (nobody’s willing to pay directly for them) and public transport, walking, and cycling (people are queuing up to get on the train). In those posts, I’ve argued that observing how people vote with their feet (or their wallets) can teach us a lot about demand for different travel modes.

BRITOMART_4116

Rail is now growing too fast to be un-fit for survival.

But as any economist knows, markets have two sides to them: demand and supply. As transport infrastructure has a lot of “public good” characteristics, it tends to be provided by government agencies such as Auckland Transport and the New Zealand Transport Agency. (These agencies wouldn’t say no if a private company turned up and offered to build a new motorway at no cost to them… but that’s not going to happen any time soon due to the fact that most recent private toll roads have failed financially.)

As a result, we have to consider how transport agencies make decisions about what to supply to the market. I’ve written a few posts on the basics of cost-benefit analysis, which is one of the tools that they use to decide which projects to build.

But is cost-benefit analysis robust, or are the results systematically biased in a certain direction? Thinking about this question led me to re-read one of my favourite papers on infrastructure costings (don’t laugh!): Bent Flyvbjerg’s “Survival of the Un-fittest: Why the Worst Infrastructure Gets Built – and What We Can do About It” (fulltext pdf). Flyvbjerg takes an empirical look at hundreds of major infrastructure projects around the world, finding that cost overruns are all-pervasive:

  • 9 out of 10 projects have cost overrun.
  • Overrun is found across the 20 nations and 5 continents covered by the study.
  • Overrun is constant for the 70-year period covered by the study, cost estimates have not improved over time.

In addition, benefits are systematically overestimated in ex-ante evaluations. The result is that a number of bad projects get built on the back of over-optimistic business cases. Flyvbjerg attributes this to “cognitive and political biases such as optimism bias and strategic misrepresentation”. (This is a polite way of saying “lying about the project to ensure that it gets built.”)

So how do New Zealand’s transport agencies stack up against Flyvbjerg’s analysis? Fortunately, we’ve got some empirical data to investigate this question with. Between 2009 and 2012, NZTA conducted and published a number of post-implementation reviews of (mainly) road projects that it funded in part or fully. Matt did an excellent job summarising the data in a post last year.

While the projects aren’t necessarily representative of all road projects, they do run the gamut from small pavement upgrades to multimillion state highway expansions. NZTA provided data comparing ex-ante and ex-post evaluations of costs and benefits for 69 projects in total. I subjected the data to some basic statistical analysis, finding that:

  • The average project had a cost overrun of 34% – a difference that was found to be highly statistically significant, meaning that there is a less than 1% probability that the observed difference happened by chance.
  • The average project had actual benefits that were 28% lower than expected – although as this difference was not statistically significant we can’t determine whether it simply reflects random chance.

In other words, NZTA and regional transport agencies seem to have had some issues accurately costing road projects. And the errors they are making are not random – they have systematically underestimated costs. This can be seen really clearly if we graph the data in histogram format.

Here’s the data on construction cost overruns, in percentage terms. The size of the bars represents the number of projects. Bars to the right of the black line indicate projects where costs were higher than expected. As you can see, costs were higher than expected for the vast majority of projects – sometimes to a quite significant degree (i.e. over 100% more expensive than planned).

Road cost overrun chart

And here’s a similar chart for benefit overruns/underruns. This shows that although estimates of benefits have in some cases been wrong by a quite large amount, most of the errors are clustered closer to the zero line. This shows that while NZTA or transport agencies often miss the mark on their estimates of benefits, the errors are sometimes positive and sometimes negative. In other words, optimism bias seems to be less pervasive when estimating benefits than when estimating costs.

Road benefit overruns chart

This data has (or should have) important implications for the way we plan and fund transport projects. It suggests that it’s necessary to be much more conservative when estimating the costs and benefits of road projects. This is especially important in light of the fact that NZTA’s funding is being devoted in large part to major motorway projects – the kind of “megaprojects” that Flybjerg identifies as posing the greatest risks for good project evaluation.

Unfortunately, NZTA stopped publishing post-implementation reviews in 2012, so it’s impossible to say whether agencies have used this data to refine their cost estimates. I hope they have, but there are indications that optimism bias is still running rampant. Take, for example, NZTA’s long-term forecasts of road traffic and public transport patronage, which blithely disregard the market realities. Or, more concretely, there’s the strange case of the Additional Waitemata Harbour Crossing traffic forecasts, which Matt picked up on a few years ago.

A 2010 business case for the AWHC, which would be New Zealand’s most expensive infrastructure project of all time, found that the project’s benefit-cost ratio was a mere 0.4 to 0.6. (Indicating that it costs about twice as much as it returns in benefits.) But, as it turns out, this figure was based on traffic modelling that overestimated actual traffic across the bridge in 2008 by almost 10% – in spite of the fact that the actual data was available at that point. That’s some serious optimism bias right there…

Ak Harbour Bridge Traffic volumes

Auckland Harbour Bridge Traffic volumes (actual and forecast)

Finally, it’s also worth noting that Flyvbjerg finds that cost overruns (and benefit underruns) tend to be a more serious issue for rail projects than for road projects, especially in the United States. Unfortunately, we simply haven’t completed enough rail projects to robustly evaluate whether the same holds true in New Zealand. However, there are some signs that recent public transport infrastructure projects have outperformed their business cases – as seen in NZTA’s post-implementation review of the Northern Busway and booming ridership at Britomart.

Can public transport save households money?

People sometimes argue that we should provide more public transport because it will reduce households’ transport costs. But is that actually true?

I took a look at this issue in a recent working paper on Location Affordability in New Zealand Cities that I presented at the 2014 New Zealand Association of Economists conference. In that paper, I found that:

…housing costs tend to fall with increasing distance from city centres, while commute distances, which drive variable transport costs, tend to increase. All other things being equal, higher rates of public transport use did not appear to improve transport affordability due to the fact that New Zealand’s public transport fares are comparable to or higher than car operating costs. However, car commuting is likely to be more costly in areas where parking is priced – a factor that we were not able to robustly estimate.

Car ownership rates, which drive a large share of transport costs, tend to be fairly consistent outside of city centres. One of the benefits of providing public transport and walking and cycling infrastructure is that it enables households to reduce car ownership costs. Conversely, policies such as minimum parking regulations tend to encourage higher rates of car ownership by ensuring abundant and low-priced parking.

In short, public transport can save households money, but whether it does in practice depends upon what how much they would pay for parking and whether they own a car or not. (A classic case of an economist saying “on the one hand… on the other hand…“!)

Here, I’d like to take a closer look at transport costs using a concrete example: my regular commute from Mount Eden to Takapuna. Here’s the Google Map view of the route between Mount Eden village and central Takapuna that I use on the days when I have to drive. (Note: The addresses on the map do not show where I live or where I work.) At 12.9 km, it’s a little bit longer than the average Aucklander’s 11.5 km commute:

Screen Shot 2014-11-07 at 1.48.16 PM

Time for some maths. According to data from AA’s 2013 Petrol Car Operating Cost Report, a compact car costs approximately $0.25 per kilometre to run. This figure includes the cost of petrol, oil, tyres, and regular repairs and maintenance, but excludes the cost to own the car.

As a result, I’d expect to spend around $6.45 per day commuting by car (12.9km x $0.25/km x 2).

What would the same journey cost on public transport? According to Auckland Transport’s journey planner, the best way to do this is to take the 274/277 bus from Mount Eden village to Symonds St, walk down the hill, and hop on the 839/858/875/879 service, which runs to central Takapuna. Because both buses run frequently all day, this is a really easy connection. (AT’s New Network will be adding frequent, connecting services to many more parts of Auckland – which is really great news for south and west Auckland and the North Shore!)

As I use a HOP card, which offers discounts on the cash fares and also a $0.50 discount if you transfer between services, the entire trip costs me $4.05 – or $8.10 per day to commute in both directions.

So far, driving is coming out ahead – the costs to operate a car are a bit cheaper than the cost of bus fares. But wait: we’ve forgotten to account for parking costs!

Albany P and R

How could we forget about parking when there’s so much of it in Auckland? (Photo: Albany park-and-ride)

Wilsons operates the closest parking garage on The Strand in Takapuna. They charge $11 for all-day parking. If I pay them for parking – and I don’t have many other options in the area – that means that a car commute now costs $17.45 ($6.45 + $11). That’s over twice as expensive as taking the bus!

In short, when people must pay for parking, public transport is a much cheaper option. However, a lot of people don’t pay directly for parking, due to the fact that minimum parking rules have resulted in an uneconomic oversupply of parking in many areas. (They still pay for parking indirectly – through lower wages, more expensive groceries, or higher housing costs. But these costs, while significant, aren’t as obvious to people on a day-to-day basis.)

And we haven’t yet accounted for one of the big costs of driving to work – the fixed costs of car ownership. Based on data from the AA’s Petrol Car Operating Cost Report and the Ministry of Transport’s data on the NZ vehicle fleet, I estimated that it costs around $2,900 per annum to own an average car (i.e. not a new car). This includes the cost of registration ($288), insurance ($790), and warrant of fitness ($49, twice a year), as well as the interest payments and depreciation on the car itself (assuming that the average car is worth around $8,000).

$2,900 per car per year is obviously quite a big cost for most households, and I’m sure a lot of people would rather save the money and spend it on other things. Abundant public transport and walking and cycling options can give households the option to downsize on car ownership and save thousands annually.

Here’s a summary of my calculations. As you can see, by taking public transport rather than driving I save $9.35 every day I commute to work. Over the entire working year, this adds up to a lot of money – over $2,300!

And by choosing to take the bus and not to own a car, I save even more money – over $5,200 every year in total. If I choose to save that money instead, it will add up to a large sum of money over time. According to Sorted.org.nz’s savings calculator, if I put an additional $5,200 in my Kiwisaver account every year and get a modest 6% return, I’ll have more than $200,000 in retirement savings after thirty years – which is enough to let me retire three or four years early.

In other words, our driving habit is literally squandering our lives. Sell your car and retire early!

Commute costs Mt Eden to Takapuna

Finally, it’s worth reflecting on the policy implications of this analysis. The maths on transport costs suggest that:

  • Distortions in one market (parking) can affect outcomes in an adjacent market (public transport). When parking is available for free, it makes taking the bus or train much less attractive from a financial perspective. This in turn makes PT seem less viable than it would be if parking wasn’t heavily subsidised. And it’s simply unrealistic to expect parking to be free – space is expensive in cities!
  • Councils have in the past regulated for an oversupply of parking by setting highly prescriptive minimum parking rules. In addition to forcing property owners to pay for supplying too much parking, this makes it impossible for people and businesses to price parking to reflect the true cost of providing it. This is insane, and we need to stop doing it.
  • While public transport can save households money on a day-to-day basis, the big gains arise when people cut back on car ownership. Transport and urban planning can enable households to make this choice – or prevent them from doing so.
  • When councils choose to put in new PT services or infrastructure, they should always ask: what do we need to do to turn a neighbourhood with a frequent bus route into a neighbourhood where you don’t need to drive for daily necessities? Neighbourhoods need to pass the “five minute pint test“: can you get a pint of beer or a pint of milk within a five minute walk?

So, what would you do with an extra $5,200 in your pocket every year?

Job growth in New Zealand – Auckland leading the way

In a recent post I looked at the geographic distribution of New Zealand’s economic activity, finding that the majority of the country’s GDP is produced by its three major cities – Auckland, Wellington, and Christchurch.

However, it’s also really important to look at the growth of regional economies over time. The US news site Vox recently published a great animated map that shows job losses and rebounding growth in the US following the Global Financial Crisis. It’s a great way of illustrating how severe an effect the recession had on some job markets, and how different cities bounced back at different rates.

As it so happens, Statistics NZ has just published new employment statistics for local authorities that allow us to look at changes in employment from 2000 to 2014. After a bit of wrangling with mapping software, I was able to produce a nice little map that shows the last fifteen years of changes in employment.

Blue bubbles indicate job growth, while orange bubbles indicate job losses. We can immediately see a few interesting things:

  • Auckland is without a doubt the largest and most consistent source of job growth in the country – the benefits of the city’s large, diverse labour market
  • While a lot of places struggled to get back on their feet after the GFC, the Auckland job market rebounded relatively rapidly
  • Even before the GFC, outcomes were pretty mixed for a lot of smaller councils – although it’s notable that Tauranga, Hamilton, and Queenstown have grown quite a lot since 2000.
Job-growth-in-New-Zealand-local-authorities

If animation does not load, click here.

I’ve also taken a look at the data another way – looking at net changes in jobs since the GFC. The following chart compares employment losses and gains since 2008 in Auckland and the rest of New Zealand.

As you can see, New Zealand lost a lot of jobs from 2008 to 2010, and Auckland took its share of the hit. But since then, Auckland has recovered much faster than the rest of the country. According to the latest figures, Auckland now has 22,700 more jobs than it did in 2008. This isn’t fantastic given the fact that the population’s grown by considerably more than that – but it beats the rest of the country, which is still down by 4,500 jobs.

Another way to look at this would be to say that Auckland has accounted for all of New Zealand’s net job growth since the GFC. These figures really show the benefits of large, diverse urban economies, which are:

Job growth since GFC chart

However, Auckland’s relatively quick recovery from the GFC was not matched in New Zealand’s other biggish cities. The following graph shows net changes in jobs since 2008 in Christchurch, Wellington, Hamilton, and Tauranga, which are New Zealand’s larger and faster-growing cities.

These cities have recovered more slowly – it’s only in the last year that they’ve recovered to pre-recession levels of employment. Christchurch has obviously had a rough time of it – it was hit harder by the GFC than the others, and then suffered a second blow after the Canterbury earthquakes. But it’s also made up a lot of ground in the last year alone as the long-delayed rebuild begins.

Job growth since GFC chart 2

As I’ve said before, it’s important to recognise the realities – and opportunities – of urban growth in New Zealand. Unless we do something really stupid, Auckland’s going to continue to be the main engine of job growth. It’s imperative that we make smart investments and policy decisions to enable that growth, which means:

But the data also shows that some of New Zealand’s other cities aren’t firing on all cylinders. Christchurch and Wellington have lagged behind during the recovery from the GFC, as have rapidly-growing smaller cities like Tauranga and Hamilton. I’d argue that many of the policies that are working in Auckland can serve as a model for other cities looking to boost their prospects of attracting (or keeping) that next resident or next business.

Better cities: they work.