Japan's Economic Earthquakes

See the post of 27 July on the Tohoku earthquake and the auto industry at the Autos and Economics blog It focuses on the Japanese economy, but I decided to put in a link rather than paste the article here in its entirety.

Earthquakes, tsunami and nuclear power: behavioral economics

The following are quick reactions, not tight argumentation.

At a forum on the earthquake today at Washington and Lee I was surprised by the interest in nuclear issues as opposed to the quake and its direct impact on people and the economy. So far no one has been hospitalized for radiation exposure; outside the plant levels remain below those that cause significant public health concerns as long as they are not permanent. And there's no sign of a Chernobyl type incident that would poison large swaths of the archipelago. The inside of the plant may be a different story, but since the reactors will never be restarted that's not a substantive issue. In the end, the most likely scenario is thus that no one will be made ill by the reactor problems. Ever. Nor will there be long-standing economic repercussions. And that is ironic if we think of the routine but small-scale disasters that accompany coal mining (most well out of view of video cameras) and the known impact of particulate matter and acid rain, or even the rarer but still frequent level of industrial accidents in petroleum extraction -- and then there's carbon footprint. Nuclear power does not make sense from a financial perspective; it is not cheap power. But it is safer and does make environmental sense compared to the alternatives.

Meanwhile people in the earthquake zone continue to die because of exposure, malnutrition and lack of access to needed healthcare. Virtually all (if not all) of these deaths are of the elderly, particularly those with chronic health issues. But they are far from zero. In addition, people will die because of the rolling power blackouts, due to an increase in accidents and delays in emergency services (which would be an issue even if none of Japan's plants were nuclear ones).

Even stranger, this nuclear issue seems to have drowned out the unfolding tale of roads being made passable, power lines reconnected and food being delivered. That causes no fear reaction, and so leads to none adrenalin rush that can come from sound-and-sight bites on the news. Yet the story has drama, particularly as any single scene is being replicated across a wide swath of Japan.

It even seems to have drowned out the story of 22,000 dead and missing, a number that surely will rise to over 30,000. It saddens me that people are forgetting that in their fascination over nuclear plants.

We cannot build human society trying to protect ourselves from every conceivable disaster. Earthquakes of M9 near densely settled populations are one such; it may be a generation before there is a recurrence, and tens of generations before there is a recurrence in Japan. But M7 earthquakes (1/60th the size of the Tohoku quake) are not so infrequent, and if they occur near an urban center (as did the 1995 Kobe earthquake of M6.9) are horrific enough.

They are a threat worth evaluating. Furthermore, we know how to build so as to mitigate the impact (including any associated tsunami) of earthquakes of M8 and below. Japan has done that. We need to recognize that the shake-resistant construction, the ocean-side seawalls and sirens, the emergency drills, and the disaster response infrastructure all saved untold lives. Indeed, in a "normal" earthquake disaster Japan would have seen neither a major loss of life nor even a long-lasting economic disruption to the affected region. Unfortunately the areas of the US similarly at risk – the west coast in general and California in particular; the Mississippi valley south of St. Louis; and Charleston SC – are ill-prepared. The demand for an adrenalin rush from the nightly news makes that response less likely.

Earthquakes and Debtquakes Don't Equate

Shots by an out-of-town TV crew of the Sendai train station, with the "bullet train" shinkansen tracks covered with rubble and a new train undergoing a test run still derailed a few miles away, highlight the damage of the quake away from the coast. It will be a long time before that train runs again. But they also highlight what wasn't rendered inoperable, the road network. Some places require detours, and it's slow going, but it goes. As a modern society, Japan – particularly rural Japan – relies on cars and trucks, not trains. And as a modern society – and not a developing country – roads form a network, with multiple routes in and out of all but the most isolated mountain valley or coastal cove. Getting food and emergency supplies in (and refugees out) will be a slow process, but as each day goes by it will be less so. [To track progress, see a map that Honda and Google have jointly created of roads known to be clear.]

As noted in my previous post, Japan's is a large, service-dominated economy, and while the prefectures of the Tohoku and northern Kanto region most affected by the quake have a population of 10 million people (depending on where you draw boundaries), that's in a country of 127 million, separated from Tokyo by a distance similar to that of Boston from New York or Paris to London. Again, depending on boundaries, the share of GDP ranges from 4% to 8%. That's substantial, but even a place as close to the epicenter as downtown Sendai seems to have been spared the specter of collapsed buildings. Stores don't have any way to restock their shelves, but that will happen in due course. As order is built out of the initial chaos, production will resume; the direct impact thereafter will be 1% or less of GDP, and for the country as a whole will be fully offset by expenditures on recovery.

Parts of the Tohoku region may never be rebuilt. One city was totally destroyed by a 1908 tsunami only to be rebuilt behind the supposed safety of a high floodwall; I personally hope people won't bet against another tsunami happening on the same once-a-century schedule. A handful of ghost towns might be a fitting memorial to the victims of the disaster, but are not substantial from the perspective of the economy as a whole.

Back on topic: manufacturing is more interconnected across geography than are services. Automotive News reports that Honda had been unable to reach anyone at several of its suppliers in the regions, sobering but hopefully a reflection of other priorities this past weekend. In any case, between shattered roads and the lack of utilities they aren't going to be producing this next week, so it's better for their staff to focus on cleaning up the random damage of their and their neighbors' homes. All too many workers will have relatives of whom nothing is yet known. Yet a number of factories in the area have already reported "only minor damage," while noting that debris-filled roads, downed power lines and empty gas stations make that moot. [Source: the Japanese-language mag2.com email magazine 自動車ニュース＆コラム that provides a daily summary of published automotive stories.] But surely there is damage at multiple factories, in the auto industry, in semiconductor manufacturing, and elsewhere. That will echo up and down the supply chain. Previous cases (an earthquake in Niigata in 2004 at M7.0, the 1995 earthquake in Kobe at M7.3) suggest most problems will be over the next month even within the region, while most firms have alternate suppliers. This quake has no precedent in size: it's 350 times larger than the Kobe quake, affecting multiple prefectures. So there will be more of an impact. But for goods production, factories were running below capacity in Japan (and in the US and in Europe). To some extent they'll be able to make up for lost output once back in operation – that is, in the interim domestic competitors will gladly pick up the slack. All of that mutes the impact on GDP.

One difference is that infrastructure may be harder to fix; there's too little information to know. Roads are one issue, and they're more important than rail. Ports may or may not need time to be cleared. Natural gas transmission lines, gasoline storage facilities, all that will need fixing. But cell phones are already operating in much of the region, albeit poorly., while in some places people apparently never lost internet access.

Outside the region directly affected by the quake, however, it's electric power that looms large, and its shadow will last longer. In principle utilities elsewhere in Japan ought to be able to pick up the slack; in practice that's not an option because the (unscathed) western half of Japan operates with alternating current at the US-standard 60Hz frequency, while Tokyo and areas further east use the 50Hz European frequency. The grids can't be interconnected. Tepco (Tokyo Electric Power Company) has no chance now of getting permission to restart the reactors that weren't operating at the time of quake, even if they replace the backup generators and other equipment washed away by the tsunami. I've heard nothing of the fate of coal-fired plants in eastern Japan, but if turbines were destroyed they too may remain offline for months. Japan has higher electricity prices than the US (well, almost every country does…) and so factories and offices and households tend to be more sparing in their use than are we. Conservation will be accordingly more costly: they already are relatively efficient in their usage, leaving for example buildings relatively cool in winter and warm in summer. Reliable electric power is the biggest hitch I see to speedy recovery for the economy as a whole.

Let me close by returning to one canard that crops up with some regularity: that Japan as a highly indebted nation can't afford to rebuild. Nonsense.

First, speaking as an economist Japan has a surplus of domestic savings; that's why they have a trade surplus and why Japanese institutional investors have been major purchases of US debt. No one inside Japan wanted to borrow. That will change, but not to the point of making Japan need to turn to the rest of the world for finance. The economy remains mired in deflation; nominal interest rates remain accordingly low, at 1.165% per annum for 10-year bonds – actually down since the earthquake. While at some point deficits will need to be pared, that is a chronic but not yet debilitating problem. With debt issued in yen, its own currency, and held domestically, Japan's case is simply not analogous to that of Greece last year or Thailand in 1997.

Furthermore, those making this claim are making a repugnant ethical judgment: that Japan as a rich county ought not use its resources to help the unfortunate. It is virtually impossible for people to buy insurance against natural disaster: there is too little information for insurance companies to price coverage, indeed they often find it hard to state what the risks are. Governments however can provide such insurance through the tax system.

Japan is not a poor, starving nation that literally cannot mobilize the resources it needs to cover basic needs. Rather it is a rich nation with hundreds of thousands of starving people. It can afford to provide relief, and it can readily raise taxes down the road by an amount commensurate to repay any short-term issuance of debt – it helps that it has extremely low tax rates by international standards.

Now most people making the argument about "too much debt" haven't thought through the implications of what they're saying – that seems to be the case of people approaching the issue from a finance background. But a well-trained economist doesn't have that excuse, we're taught to look behind the facade of jargon and theory to the underlying assumptions of models. If economists make this claim, it's because they don't have the courage to say what they really mean: "tough luck, northern Japan, we don't think the government should do anything to help."

That's not my stance, and I'm thankful that it's not the stance of Japanese politicians. Aid the region they will, and Japanese society will be the better for it – even if taxes have to be raised a tad down the road.

Earthquake and Economy

I've spent a couple hours watching the news from Japan, switching between TBS and NHK, both available via internet. Having watched footage, it's still unimaginable. The tsunami hit Rikuzentakata, population 23,000. It just … vanished. Hopefully most fled between the short interval between the initial quake and the tsunami, but aerial photos of occupied cars being washed off a highway in the initial wave suggest that interval was too short for some.

The magnitude of the damage is unknown; the focus remains basic rescue. A quick check of the Japan Meteorological Agency web site shows over 150 aftershocks big enough to be listed, many over Magnitude 6 – under normal circumstances strong enough to have made the news. It's not clear yet whether cooling can be restored at the Fukushima nuclear power plant before a partial meltdown – the initial reports suggest "no". Elsewhere fires burn out of control, with no ability to do more than watch from a distance. Amazingly, many elevated roads seemed to have survived both the quake and the tsunami passing underneath; wooden structures hit with water however just … disappeared into debris.

Indeed, writing 12 hours after the above -- 10 am US East Coast on Sat 12 March, midnight of 12 March in Japan -- power is already back on in parts of Sendai, with traffic lights operating amid light traffic. Japan's earthquake codes seemed to have done their job: no scenes of the pancaked buildings familiar from Haiti or New Zealand. Indeed, at least some ferro-concrete structures appear to have remained standing in areas otherwise swept clean by the tsunami. The earthquake itself seems to have caused damage far less than the size of the quake might suggest; it was the tsunami that was inexorable in force. Disruptions there will be, but there is nothing to suggest wholesale destruction of physical productive capacity. Factories will likely stay shut until aftershocks die down – one friend was kept awake all night yesterday by the constant shaking 150 miles away in Chiba. The worst damaged nuclear power plant has managed to start pumping in sea water; it will never operate again, but a catastrophic meltdown now appears less likely.

For now the productive capacity that matters are the hospitals: they're accepting patients on a triage basis in all but the worst-hit places along the coast. And the information is getting out, it seems the internet remains operable in areas away from the coast, even while electricity remained cut.

So what to make of it as an economist? Here are quick thoughts.

First and foremost, this wasn't the big earthquake that everyone feared: it was not a repeat of the Great Kanto earthquake that hit at 12:01 pm on September 1, 1923. The city erupted in flames, as charcoal was the primary fuel for cooking lunch; tens of thousands were asphyxiated in the following firestorm. On average the Tokyo area has been hit by a major quake once every 50 years -- it's now been almost 90. So while there were casualties in Tokyo proper, and not all the subways are operating, Japan's major population center effectively escaped damage. The other two major centers, Nagoya and the Osaka-Kyoto area, were unaffected -- with Tokyo, those three metropolitan areas account for over half of Japan's population. Nevertheless, Sendai is a city of 1.0 million, so the earthquake may have produced tens of thousands of casualties.

This will obviously have a short-term effect on the economy; production in the region will cease for some time to come. Initial very partial reports suggest damage at factories in the area, including several Toyota subsidiaries that specialize in small cars. Ports facilities are damaged (or just gone) while roads are impassible. Given the complex overlay of supply chains in many areas of manufacturing, this will hit production elsewhere in (and outside of) Japan in a manner that is yet impossible to delineate. Now many plants were away from the coast and designed to withstand moderate earthquakes. This one was not moderate. So between clearing roads and making basic repairs, it may be a couple weeks before anything can restart. It's not possible to do more than note the issue – it could be much worse.

This is in a context of recovery from a recession. The most recent data, for Oct-Dec 2010, showed negative growth; projections (e.g., Morgan Stanley MUFG) were of 1.5% growth this quarter. The short-term impact will be small this quarter, because it's largely over. Nevertheless the disruption to manufacturing throughout the country, and the short-run loss of most output for the remainder of March in the northeast, could push the economy back to nearly zero. The short-run impact will appear mainly in next quarter's data.

Japan however is not a manufacturing power – or not primarily one – but a service economy, as is true of the US and the European economies. Hospitals in most of the country will continue to treat patients as always, restaurants will continue to serve lunches and dinners, schools will do their thing. There is no particular reason that the Sendai earthquake and tsunami will change those in the (economic) majority of Japan. Indeed, refugees may boost demand elsewhere, partially offsetting the collapse of activity in the areas directly affected by the temblor.

Then the rebuilding will begin. Since the Japanese economy operates below capacity, and youth in particular are underemployed, this won't be at the cost of other production, or at least not on a one-for-one basis. In other words, on a net basis the boost in employment would offset most if not all of the jobs lost. Of course Japan is running a large deficit, but unlike the US there is no unreasoned panic about debt and deficits. Interest rates remain close to 1% on government and high-quality corporate bonds. If the government needs to 10% of GDP, then it can probably do so without much problem. Yes, in the long run the deficit needs to be turned into a surplus. Short-run needs will dominate for another year or so. And since the Kan government was close to collapse, unable to pass legislation, trimming the deficit wasn't going to happen anytime soon.

So a year from now … the economy will be back on track. There are uncertainties. What will be the impact on markets of insurance companies liquidating portfolios to pay claims? Or (my at present uninformed guess) will they be partially bailed out? Beside China, the economies of Japan's major trading partners are growing slowly; next year will surely be better... Will the disaster speed political realignment? My guess is it will – after all, Japan's politics can't get worse, can they?

This isn't good news. It's just not as bad as the news could have been.

Mike Smitka, Lexington VA