Excitement over broken windows

Oliver Hartwich of the NZ Initiative revisits the broken window fallacy:

“Natural disasters and wars never generate prosperity. They always destroy it, by definition.”

He is absolutely right. It’s good to see this revisited. Even though the “seen benefits, unseen costs” principle was articulated by Bastiat in an 1850 essay. Wikipedia article on the broken window parable here.

The Canterbury earthquakes led some to say it will be good for economic growth. In truth there has been limited impact on the national economy. Production shifted elsewhere and insurance money helped offset the wealth shock for many. Much more so than I expected. But there have been real economic costs for those who have lost jobs and their housing costs have surged.

Oliver, rightly points to the direct effects:

…as a direct result of the quakes, hundreds of firms went out of business, tax revenue was lost, and the government’s budget was pushed deep into the red.

And these are obviously only the economic costs of the natural disaster, not counting the loss of lives or the physical and mental health effects.

And Oliver points to the displacement effect:

If the earthquakes had never happened, there would not have been a need to deal with them. All the resources now devoted to cleaning up and rebuilding would have been employed elsewhere.

To add to his narrative, the impact is still unfolding. Take two aspects of usual life: jobs and housing.

The jobs market is still dire. According to the Household Labour Force Survey, the change in jobs since late 2010 and now is:

1)      16,000 lower total employment

2)      Construction and logistics jobs have surged by 10,000

3)      job losses of 26,000 in other sectors. Retail and hospitality has been hit the hardest.

Unless you are in the construction industry, the economic costs of the earthquakes are real.

The housing market shows the impact of a sudden reduction in supply. Prices have surged. According to MBIE data, rents in Canterbury TAs have risen by 8%-10%pa since the earthquake, while the national average has risen by 3%pa. House prices have risen in line with rents.

Unless you are an existing home owner, unaffected by the quakes, your cost of living went up sharply.

As someone who grew up in Canterbury, it is all too easy to see that the place has changed significantly because of the quakes. The costs are borne by the many, the benefits by the few.

  • JC

    Not to count the national effect of much increased insurance premiums and the loss of the replacement policy to all of NZ. Thus we all lose several hundred million in increased premiums, a future, possibly catastrophic loss to some who will inevitably under estimate the cost of rebuilding their homes plus have a more aggressive insurance industry scrutinising claims.

    On the good side.. an event hundreds of km away finally made me buy a good survival kit and a chemical toilet..


  • Bert

    Say’s law appears to be alive and well… “All the resources now devoted to cleaning up and rebuilding would have been employed elsewhere”. That’s a fairly strong assumption. The reconstruction that is now taking place has been one of the few sources of output growth over the past two years. Take a look at the labour markets in other provinces – its even worse.

    • James

      How is that an assumption? Resources were being used before the earthquake where people wanted them. The earthquake came, resources were lost. Now resources have been diverted to bring Christchurch back up to at least pre-earthquake levels. Were people all hoarding their resources in case of an earthquake and now they’re finally able to use them? The only assumption here I can see is that an economy prospers from renewed demand, or should I say “output growth”.

      Oh, but if you’re buying into Keynes’ interpretation of Say’s law, of course you’d feel obliged to defend broken window ‘prosperity’. As long as it boosts aggregate demand, no?

  • deepred

    Isn’t disaster capitalism just a more recent variety of the Broken Window Parable?

    • I don’t understand the question – what is disaster capitalism? And where is this disaster capitalism occurring, given the state accounts for half of economic activity in countries that experienced the sharpest financial crises (such as the UK) 😉

      • deepred
        • Yea I knew that, I was just aiming for a specific definition! Hence the second part of my statement – where I say it is a bit sneaky to blame a concept like “capitalism” when it is large integrated states that have been hit!

          Also, Kleins thesis rests on a series of misinformation and downright lies – she is hardly my favourite person tbh, given that she is more interested in selling her brand than trying to actually discuss issues of concern.

          However, even if we were going down the road of the state using fear to push through policies that benefit its cronies instead of the public at large – I have no idea how it is really related to the Broken Window Parable, which as I note above is about the fact that the “activity” due to breaking something we value is not a “good” in the sense of individuals and communities having more things. It is merely replacement for what we have destroyed.

  • Miguel Sanchez

    I’m surprised by the number of otherwise smart economists who are content to shout “Broken Window Fallacy!” without thinking through what it actually means. The insight is that repair work which benefits one party shouldn’t be extrapolated as being good for the whole economy, because it comes at the expense of more productive things that could have been done with those resources. But it’s a closed economy argument; when we say “the economy” we’re really talking about the global economy, and the beneficiary in this case could be “Christchurch”, and I don’t think it’s a stretch to even extend that to “New Zealand.”

    The key factor in where the benefit lies is the degree of insurance coverage. The BIS did some research on this last year; see http://www.bis.org/publ/work394.htm. They found that uninsured natural disasters are unambiguously costly, but when there’s a high level of insurance – as is the case in NZ – the effect on the long-run level of GDP can be neutral or even slightly positive. Quibble about the modelling if you like, but I find this far more sensible than the BWF brigade, who conveniently sidestep the issue of insurance by treating all damage as self-inflicted. I somehow doubt that nuking yourself would be covered by any insurance policy.

    • Seamus Hogan

      Miguel. You are right that there is a possibility of a natural disaster resulting in a higher present value of a country’s output if a) there is over-insurance in the sense that insurance replaces assets that are better (maybe just by being newer) than what was destoryed, and b) that fraction of that insurance that comes from within the country is smaller than the fraction of payouts that are over insurance. But I don’t think the criticism of the BWF brigade is right here, for two reasons:
      First, it is very clear in Christchurch that insurance has not been sufficient to restore many of our assets without additional top-ups, and in any event a lot of the insurance was held by EQC which did not reinsure overseas, so empricially, I don’t think the conditions for net benefit to NZ are met.
      Second, the possibility of over-insurance is not what the “disaster’s are good for GDP” people are thinking; the BWF brigade are attacking that analysis that focuses on dollar income rather than real resources.

      • Shamubeel Eaqub

        Seamus, thanks for the thoughtful reply. I have been out of action with some deadlines. I think the main issue, perhaps one I should have explained further, is that the GDP test is not necessarily the right one. When we look at the other costs, such as the displacement etc, its pretty clear that natural disasters are not good for the economy.

      • Miguel Sanchez

        Hi Seamus – as I read it, the BIS results weren’t dependent on either over-insurance (whatever “over” might mean) or offshore reinsurance. By the way, your comment on EQC is not quite right – reinsurance will cover up to $5bn of their estimated $12bn liability.

        On the second point, I imagine the reason that people focus on the dollar income is because the real resources story seems patently absurd to them. We’re clearly not in full employment now, and we weren’t before the quakes, so what basis do we have for claiming that “all the resources now devoted to cleaning up and rebuilding would have been employed elsewhere”? My point is that economists are too content to simply make this assertion without actually demonstrating it.

        • It is a compositional effect – economic activity associated with capital investment that is replacing lost capital doesn’t provide “welfare” in the same way consumption of that income would. The essence of the broken window idea isn’t the level of GDP per se, it is the fact that this aggregate measure of activity involves a lower level of welfare than it would have otherwise – I think that was the main point the post was going for as well, and perhaps we are all talking past each other 🙂

          “We’re clearly not in full employment now, and we weren’t before the
          quakes, so what basis do we have for claiming that “all the resources
          now devoted to cleaning up and rebuilding would have been employed
          elsewhere”? My point is that economists are too content to simply make
          this assertion without actually demonstrating it.”

          We given monetary policy an active role to ensure we expect “head back to potential” ex-ante – that is the factor that allows us to make this argument. The fact that there was underutilisation of workers and capital ex-post doesn’t invalidate that, especially if the reason for this was unforeseeable.

          We are all claiming it on those grounds.

        • Seamus Hogan

          Hi Miguel. I had remembered discussion at the time of the Sept 04 quake about how EQC had moved away from reinsuring overseas, but I guess that had been only partial. For over insurance, I suspect we are using the terms a bit differently. I only meant that when capital stock is replaced by insurance, it will typically be better, if only becuase it is newer than what it replaces.
          As to the more substantive points about whether there can be a beneficial impact even when the insurance liability is local, and whether less-than-full-employment renders the broken-windows fallacy non-fallacious, these are interesting points that I thought deserved their own posts. See http://offsettingbehaviour.blogspot.co.nz/2013/09/broken-windows-part-i-measured-gdp.html and a follow-up post tomorrow.

    • Shamubeel Eaqub

      The context of course is that parts of the economy are non-traded. That was the reason for highlighting the impact on construction and non-construction jobs.