Every weekend we dig into the archives. This post by Peter was originally published in June 2016.

In the 1990s, in the early years of the information technology revolution, economist Robert Solow famously commented that “you can see the computer age everywhere but in the productivity statistics.” Two decades on, that still rings true. Social life has been profoundly transformed by new technology: It has altered the way we communicate with friends and family, how we entertain ourselves, and even how we date.

how people meet romantic partners chart

When I read Douglas Adams’ Hitchhiker’s Guide to the Galaxy in the early 2000s, the titular device still seemed like a fantastical idea: a handheld device you could use to access information (much of it inaccurate or incomplete) on anything, from anywhere.

Now, we all have smartphones. But productivity growth has stubbornly failed to take off over this period. Does this mean that technological progress has failed to deliver?

Journalist Ezra Klein (Vox) recently reviewed the current debate over technological progress. One perspective he discusses is that the benefits of information and communication technologies (ICTs) have largely accrued to consumers rather than producers:

Measures of productivity are based on the sum total of goods and services the economy produces for sale. But many digital-era products are given away for free, and so never have an opportunity to show themselves in GDP statistics.

Take Google Maps. I have a crap sense of direction, so it’s no exaggeration to say Google Maps has changed my life. I would pay hundreds of dollars a year for the product. In practice, I pay nothing. In terms of its direct contribution to GDP, Google Maps boosts Google’s advertising business by feeding my data back to the company so they can target ads more effectively, and it probably boosts the amount of money I fork over to Verizon for my data plan. But that’s not worth hundreds of dollars to Google, or to the economy as a whole. The result is that GDP data might undercount the value of Google Maps in a way it didn’t undercount the value of, say, Garmin GPS devices.

As Klein goes on to observe, ICTs have transformed our leisure time more than our work time – in large part, by giving us many more choices about where to dine, what television shows to watch, and who to talk to.

Interestingly, what’s true for technology might also be true for cities. The conventional narrative about agglomeration economies – the economic benefits of scale and density – is that their main effect is to lift productivity. But, as Stu and I have discussed in the past, there’s an increasing body of evidence that suggests that agglomeration also has significant benefits for consumers.

In recent years, economists have used micro-data on household consumption patterns to build a much richer picture of the impact of city size and structure on consumption choices. In short, larger cities don’t always offer lower prices – as you’d expect if higher productivity made it cheaper to produce goods and services. But they do offer a much greater variety of goods and services, which in turn translates into higher wellbeing for households.

A 2015 paper by Jessie Handbury and David Weinstein uses barcode data on retail sales in 49 large US cities to analyse prices and product varieties. They find that:

There are approximately four times more types of grocery products available in New York [metro population 21 million] than in Des Moines [population 456,000].

Because people in larger cities tend to buy a wider range of goods, including more expensive products, a naïve comparison of average retail prices would suggest that larger cities are more expensive. But Handbury and Weinstein’s analysis shows that, after accounting for product variety, prices in large cities are no more expensive than smaller cities. If anything, they tend to be lower:

When we use the data to construct a theoretically rigorous price index that corrects for product, purchaser, and retailer heterogeneity and accounts for variety differences across locations, we find that the price level is actually lower in larger cities. Consumers spend less, on average, to get the same amount of consumption utility in larger cities.

Moreover, what’s true in grocery stores is also true in restaurants. In a 2012 paper, Nathan Schiff took a look at the impact of city size and population density on restaurant markets in 726 urban places in the US. His key finding is that:

For the 182 cities in the top quartile by land area of my data (mean population 331,000), a one standard deviation increase in log population is associated with a 57% increase in the count of unique cuisines. A one standard deviation decrease in log land area–which increases population density without changing the size of the population–is associated with a 10% increase in cuisine count, equivalent to increasing the percentage of the population with a college degree by one standard deviation and larger than the effect of increasing the ethnic population associated with each cuisine by one standard deviation.

In other words, cities that are larger or denser offer people more choices about where and what to eat. Density is especially crucial in large cities, as people generally don’t travel long distances to dine. (Incidentally, relatively open migration policies are also an important enabler of restaurant choice in cities, as migrants bring new cuisines with them.)

Distances travelled to jobs and retail in Auckland

What does this mean for urban policy? I think there are two main lessons.

The first is that although agglomeration economies in production are important to long-run economic outcomes, we might be looking for the benefits of cities in the wrong places. They might not always appear in productivity statistics or price indices, but in the consumption choices that cities offer people. Measuring variety – and how people respond to it – is therefore crucial to understanding agglomeration economies.

The second is that conventional urban policy might be based on false premises. Ever since the “dark Satanic mills” of the Industrial Revolution, policymakers have assumed that cities are good for businesses but bad for people. Accordingly, they designed transport systems and planning policies that aimed to disperse the city and to separate people from their workplaces and from each other.

That made sense when cholera was a major cause of death, but it’s increasingly illogical in today’s world. Urban disamenities such as air quality, crime rates, and infectious diseases are all improving, and the evidence increasingly shows that the consumer choices offered by cities (and dense urban places) have benefits for households. In this context, policies that enable urbanisation are likely to have larger benefits than commonly assumed.

What do you think about the role of consumer choice in cities?

Share this


  1. Just wildly speculating, but if 3d printers ever become tools instead of toys, physical devices might slowly give way to designs downloaded from the web plus raw material. You would probably have specialist fabricators for different types of goods. One for furniture, one for 3d printing houses, another for jewelery. That would change how we would consume and transport goods in cities. Cities would have the best fabricators.

  2. “policymakers have assumed that cities are good for businesses but bad for people. Accordingly, they designed transport systems and planning policies that aimed to disperse the city… Urban disamenities such as air quality, crime rates, and infectious diseases are all improving,”

    The transport systems and planning policies they used to disperse the city worked, creating a city that is ‘bad for people’. Low levels of physical activity, together with the consequent levels of obesity, are contributing to many of the diseases that are affecting our people, and costing us dearly in the public health system.

    Our population would be far better off if politicians at all levels understood this and assisted the city to focus on correcting this poor urban form.

  3. Heidi
    Dispersal of the city can not be blamed on just our policymakers. Many young people still believe that they need a house with its own piece of land; and the piece of land that they can almost afford happens in many cases to be somewhere near Kumeu, or Wellsford, or Papatoetoe. So the policymakers facilitate this by saying, we’ll build you a park and ride so you can live wherever. And if someone wants to drive all the way they can afford to because they don’t have to pay the full cost of the roads that they travel.

    If Auckland wants to reduce congested roads then the change needs to start right now. The NZ Herald says that a briefing to Minister Twyford last year outlines that North Shore motorways are reaching a looming crisis, and that we are too late to build our way out of it soon enough (a rail solution).

    It seems that the answer is cheaper public transport to increase demand; and road, or vehicle charging mechanisms to decrease car trips.

    We also need to be conscious that there are alternatives for those forgoing car trips. I have studied my own area and find that by altering the configuration of current services frequency can be changed from 30 minutes to 10 minutes at no extra cost – and probably less cost. A turn up and go service where it is hard for motorists to argue, “I had to drive, there was no bus.”

    Much of what Auckland has done in the transport area either hasn’t worked, or has worked poorly. As Lester Levy says, time for transformational change.

    1. There is no sensible reason for Heavy Rail to the Shore.

      Light Rail yes, but Heavy Rail is a total waste of time and money.

      As for “alternatives”; I’m not swapping a one hour drive in absolute total peak congestion for a three leg public transport journey that takes two hours. Just because it’s possible doesn’t mean it’s feasible as an alternative.

  4. Buttwizard, I agree that Heavy Rail to the Shore is not a sensible solution.

    I would love to know your PT peak journey that takes two hours. Maybe the answer is, how can NZTA/AT improve that journey time? Or maybe you will be one of the people that pays for the increasing cost of carbon emissions or road congestion charges? I get that – not everyone will will be able to use PT, or indeed want to use it.

    When I work in South Auckland it takes me about 85 minutes to get there – some of it due to the poor connection of the NEX to Britomart. At peak driving it can take me 90 minutes. The first I find a pleasant often productive journey as I can read files, or catch up with news on the net. I accept that many still do the latter while driving and they might not find the drive stressful.

    1. John – I’m going from Massey in West Auckland to Greenlane; as infuriating as it is, it’s the simple act of getting to a train station just down the road at Sturges Road. I can be halfway to work in the morning in the time it takes me to get to the Park & Ride, and that still involves driving the first leg. PT-wise, it’s quicker to catch the bus to Constellation Station, then to the City, then train to Greenlane. Still, that offers me a travel time of 93 minutes + the walk at the other end. It’s a 30 minute drive in the morning door to door.

      The CRL and North West Light Rail will probably halve this time; or at least make it competitive with the ‘travel 10km in the opposite direction option which is somehow quicker’, which is why I find the delays to the LRT project at NZTA so frustrating; more so given the additional development burden my area is taking on relative to other areas that have better transport connections already. It’s also why I favour going straight to LRT along SH18 between Westgate and Constellation Station.

      1. Buttwizard
        I completely agree with you that the first step is “going straight to LRT along SH18 between Westgate and Constellation Station.” We build roads ahead of demand and so why not PT particularly as the environment has changed so dramatically and carbon emissions are a huge focus. While many might think that electric cars are the answer it is clearly not the answer to congestion.

        I also believe that the first step, Constellation to Albany, should be light rail and yesterday I wrote (OIA request) to NZTA asking whether they have investigated this in the light of their report to the Minister about the looming North Shore crisis – their words not mine.https://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=12210993

        I sympathize that transport options other than driving are so poor out west. We were out that way just last night and I couldn’t believe that the NW motorway seems to have lanes farther than the eye can see. Despite this colleagues of mine talk of travel times of 90 – 120 minutes in the morning.

        1. I’d be keen to know the outcome of that OIA. It’s just mind-bending to know how reliant we are on a single onramp like Royal Road (which I think has now been worked on for two years with no end in sight) for the entire Northwest Quarter, with thousands of new houses going in within a 5km radius and no firm plan for rapid transit in the area. This week there was an accident on the onramp and a 30 minute commute blew out to 90 minutes for my partner. Traffic backs up onto Don Buck Road and paralyses the entire suburb. It’s chaos, but it’s out of sight and out of mind because it’s out west, so no one cares.

      2. Is it possible for you to cycle to the Sturges Road station or go further on the NW Cycleway to say Mt Albert Kingsland station, lock the bike up and train from there? Which side of the NW motorway are you?

        1. It’s possible, sure, but coming from the top of Don Buck Road means it’s going to be a lot quicker one way than it is going home. That being the case, I generally need to be at my desk by about 7am each day which rules out cycling 10km+ to get to a train station further down the line.

Leave a Reply

Your email address will not be published. Required fields are marked *