Saturday, 19 May 2012

Maybe we needed a bigger earthquake

Paul Krugman credits the Japanese earthquake/tsunami with high current Japanese growth rates:
Wait, what? Japan as star performer? What’s that about?

Actually, no mystery. From Bloomberg:
Japan’s economy expanded faster than estimated in the first quarter, boosted by reconstruction spending that’s poised to fade just as a worsening in Europe’s crisis threatens to curtail export demand.
So Japan, which is spending heavily for post-tsunami reconstruction, is growing quite fast, while Italy, which is imposing austerity measures, is shrinking almost equally fast.

There seems to be some kind of lesson here about macroeconomics, but I can’t quite put my finger on it …
I'm not a macroeconomist, so there are reasonable odds I've got things wrong. But I would have thought we'd have needed to think a bit more about how a fiscal push gets funded and about any likely reaction from the reserve bank.

Oughtn't there be a reasonable difference between a tax or debt-funded fiscal expansion and one paid by reinsurance inflows? I'm not sure that we can jump from "the Japanese government is spending a lot" to "Japan's growing" without looking at where the money's come from. A tax-funded spending programme will take money out of other parts of the economy; a debt-funded one might induce people to offset government spending with greater savings in anticipation of future taxes. Even if you think it's worthwhile, the effect of spending will be smaller than it would be where the money came as windfall in exchange for wealth reduction. A fiscal push paid for by domestic insurers selling off assets and by inflows from international reinsurers ought to look a little different from one funded by tax or debt; the GDP effects of tsunami-scale spending programmes in Italy or Greece might be comparable to Japan's if they were paid for by selling off little used assets, like some of Greece's uninhabited Adriatic islands.

New Zealand has its own earthquake rebuilding project: Christchurch. While 2012Q1 figures aren't out yet, fourth quarter 2011 had a year-on-year growth rate of 1.1%. And iPredict's picking quarter-on-quarter growth rates between 0 and 0.5% for most quarters all the way through to December 2013, with reasonable risk of bad outcomes for 2013Q4 - one chance in three of growth rates lower than -0.5%. New Zealand's perhaps making it look hard because of the morass of bureaucracy and insurance that's holding things up, combined with lingering worries that a one-in-four chance of another really big aftershock might mean we're best advised to wait on any big building projects anyway. I doubt that the problem was that the quake just wasn't big enough.

Further, any earthquake-related fiscal push also needs an accommodative central bank for expansionary aggregate effects. Here, a fast rebuild push would very likely push up wages and prices in related sectors, at least until we started being able to bid workers back from Oz. We'd also be pushing rents up for temporary workers, which would feed through into national rental markets as displaced Christchurch folks facing a short term vertical housing supply curve moved elsewhere. And then the RBNZ might start having to put the thumb down.

Bottom lines from a non-macroeconomist:*
  • Earthquakes are hardly sufficient for macroeconomic stimulus. There's evidence that Canterbury's seeing decent growth relative to the rest of the country, but aggregate national figures are hardly rosy and construction hasn't been particularly helping things. December quarter 2011 had a 2.5% increase over the prior quarter, but only after substantially negative results for the prior three quarters (ie from the quake onwards). See Table 2 of the Excel sheet. Table 3 tells us construction's contribution to GDP is up in 2011 on 2010, but is still lower in real terms than in any year from 2006 through 2009. I suppose earthquake construction booms also operate with long and variable lags.
  • Running a big fiscal push doesn't help anything if the Reserve Bank is then just induced to offset. And if you have to get the monetary authority onside, why not just run it as a monetary expansion in the first place? Cowen says the potential for monetary fixes is weakening, but I'd expect the same case to hold against broad fiscal AD pushes as well. And recall that what Krugman means by "austerity" is perhaps not what is commonly understood.
  • It makes a lot of sense for New Zealand to borrow heavily to fund the earthquake rebuild and to divert some money from other useful projects, with longer term tax increases and spending shifts to make up the difference. That the coming budget is almost certain to do nothing about longer term structural issues like the retirement age isn't going to help make room for earthquake-rebuilding debt. 
  • The quickest way to make room for the RBNZ on interest rates is getting legislation giving effect to the Productivity Commission's recommendations around housing and land use regulation. We're seeing housing prices ramping up again; Auckland median values are now well above the 2007 peak and aggregate values aren't far below it. Price inflation in non-tradeables like housing and fear of setting off another housing bubble could be constraining RBNZ against interest rate cuts. In the last housing run-up, from 2003 through 2007, year on year CPI measures in the "CPI less housing non-tradeable items" series is about twenty percent lower than the full CPI. RBNZ will have more room to accommodate where housing costs are less of an issue. This obviously matters for any monetary push but matters too for any fiscal stimulus because RBNZ moves last and has to offset fiscal moves that look set to push medium-term CPI above 3 percent.**
* Full disclosure: I was very seriously wrong about macro policy in early 2008 when I thought RBNZ was cavalier about a persistently high CPI when instead they had better foresight about the coming maelstrom; I'd calibrated around RBNZ responses to inflation rates circa 2005-6. So discount as you reckon appropriate.

** They've also been looking to other potential tools for damping housing price run-ups.


  1. Your error is thinking of reconstruction spending as a fiscal stimulus. If a disaster destroys capital in a no-government economy, that capital will be reconstructed to the extent that the market values it. Whether that's stimulatory depends on (a) the prior state of employment in the economy and (b) how able the economy is to import resources. If the economy was at full employment with closed borders - no stimulus, just redistribution. Reinsurance payments are imported resources, so they can be stimulatory.

    Then introduce the government, let it borrow (stimulatory) or tax (not) if it wants to rebuild more than the market does, and think about where the offsets lie.

    *Then*, having worked out how much stimulus is actually going on, introduce the central bank. Decide whether it targets just inflation or activity too, and with how much discretion. Think about how it will react to the stimulus you've identified.

    And that's how you think about a natural disaster.

    1. I don't disagree, fibby. There are really big differences between stimulus as we normally think about it and earthquake rebuild expenditures. We shouldn't expect the two to yield comparable outcomes.

  2. Surely this is just a rehash of the broken window fallacy. GDP measures economic activity. Following a major disaster, economic activity necessarily increases. That doesn't mean the country is becoming wealthier. This is the same Krugman who thinks an alien invasion scare is just the thing to sort the economy out, right?

  3. jeez, Eric, and its good for the Germans to pick up Greece,it could keep our $NZ dollar respectable, suck here, and its good for Aucklander to salvage Christchurch, pay here you bastard Aucklander, and my broken window repair man has never hear of Bastion, buts its ok,
    : now it best if you spend more time with women and children in household Eric, do not think too much,