By Henry Potrykus
The Federal Reserve Bank of Atlanta published an important paper a little while ago on “The Implications of a Graying Japan for Government Policy.” The title rather understates demographics’ fiscal gravity. The analysis within the paper does a better job. Here I relate the findings to those of us here in the West in a four-part development: ‘context,’ ‘strengths,’ ‘weaknesses,’ and ‘implications.’ You might want to skip down to “implications” if you just want the highlights. The ‘context’ to follow also makes for relatively easy reading. ‘Strengths’ and ‘weaknesses’ will not be for everyone. They contain economics, which is a practical necessity - the paper is on economics.
Japan is not the United States. Critically, Japan did not have a post-war baby boom. Instead, its realized family size was high just before the War. This is the main way in which Japan is a forerunning first-world economy. Germany and other Western European countries will follow Japan’s demographics, and then we’ll have our own run of it.
Japan had exceptionally strong trade surpluses (Current Account surpluses; like Germany does). Remember worrying about Japan owning the US? Worry about that anymore? Even more important than that penchant for over-seas saving (which is being cashed-in on), is that Japan controls its own debt: Japanese savers (future retirees) buy Japanese paper.
Thus, as a demographics “policy experiment,” Japan might be considered an ideal case: The fiscal picture - issuance and redemption of bonds; payout of pensions; and the whole fiscal balance sheet - is centrally manageable. I am sure there are benefits to the US having the reserve currency of the world. Control over federal bond auctions is not one of them. (A lot of swaps have to be written between a lot of interested parties to “control” that market.)
Watching Tsunami footage also shows one how unified Japan is in caring for its elderly. (In this case that means watching 60 year-olds take care of 80 year-olds.) So, again - Japan has a relatively manageable problem. By way of comparison, Western states see incipient riots over public pension changes.
Perhaps that’s not a perfectly fair comparison, but, in one way, Japan’s problem is more urgent; in another, it is less so. In the latter sense, Japan may just be able to smoothly reallocate and reduce its population’s consumption and so avert Argentina-like fiscal crises. There is more on this in the ‘implications’ section.
With this paper, finally, economists get demographics. The paper models populations moving through the life cycle. Cohorts are born, grow up, get to working, retire, and die.
It should be obvious that when one wants to understand the fiscal picture of a state, enumerating aging populations is required: The picture for entitlements (think Social Security and Medicare; Japan has its analogs) is determined by whom you tax (workers) and who receives remittances (retirees, who, except in special circumstances, have not yet died). The fiscal picture for welfare states (like our own) is driven by their entitlements. This holds especially for Japan. (Allow me to throw our generous poverty programs in here, and the statement becomes general, and air-tight.)
Another “strength” is the simplicity and constancy by which the modelers treat the Japanese credit markets. This will be controversial to some (in particular, the public behaves “non-Ricardian”). What I find important is that interest rate spreads are held constant. Rate spikes (“runs”) that might take place - perhaps at any time - are not considered. (See below for other credit events.) In this sense, the central bank helps the government manage its debt burden well.
Many economists will find the dynamic model, with its perfect foresight, etc., to be a strength of the paper’s analysis framework. You say strength, I say assumption. Certainly it is good to attempt to model how people will shift their behaviors (“dynamically”) to changes in government policy like pension generosity and tax increase. Without further comment, let’s just say, “how do you know, quantitatively?”
Two quantitative parts of the analysis that are real, powerful factors in determining future fiscal economics are the rate of productivity growth (“total factor” productivity of the macro economy, irrespective of capital or labor contribution) and the fertility rate (rate of growth of the population).
The analysts treat the first term optimistically. I am being generous here (as they are, to Japan): Japan had a “lost decade” in the 1990s, and the hoped-for productivity recovery didn’t materialize in the 2000s. I link this phenomenon to demographics in "Decline of Economic Growth: Human Capital and Population Change." (The term “total factor” means, precisely, that it is not so-linked in other analyses. The present analysis’ real departure is more subtle, however, again, because this work gets demographics “right.”) So, then, Japan has had two lost decades.
Productivity (irrespective of capital and labor) is especially important in systems where capital is crowded-out by an all-consuming public sector (trust for a little while that I am not being alarmist here; we’ll return to it as we go into ‘implications’ below), and labor declines (a demographic hole).
So, optimistic views of productivity can help us gloss-over public-sector rapacity and labor collapse. In other work I [and others] analyze on how labor collapse can be the harbinger of economic depression. Other think tanks have spilt plenty of ink on public-sector encroachment. Maybe I should weigh-in myself sometime.
If the first term is treated optimistically, the second term is treated fancifully. Japan will rediscover its lost desire to have families, we are told. Now, Japanese sociology deserves its own study, which I haven’t done yet! But, closer to home, there are plenty of sociologically compelling reasons to ignore fairy tales of demographic recovery. It is obvious that population replacement affects the fiscal picture, quantitatively. (Sensitivities vary. See the next paragraph and the ‘implications,’ below.) But I have another, seriously wonkish point to make here: The authors desire closed, solvable systems (in a very formal sense; they need so-called “transversality conditions”). I think it’s time we jettison silly assumptions that dictate family recovery and start accepting the solution that asymptotes to zero. The former is nowhere indicated by any of our real social policies anywhere. Zero is a mathematically serious number. It is fiscally serious too.
One will note that even without more reasonable demographics, the study still finds that there is nearly one pensioner for every working-age Japanese individual (by around 2090). Let’s get into the ‘implications’ of that.
If Japan doesn’t reform its “Social Security” and “Medicare” system (they don’t use our terms), nor its tax system, they would have to raise consumption taxes to never-witnessed levels; likely beyond 55 percent. This must happen before 2040, which turns out to be an annus mirabilis - see below.
That’s more than half of consumption, taxed away, if they just kick the can. If Japan raises taxes quickly to the required level (by 2018; remember this is “quick” by public standards), that level “only” needs to be 35 percent or so. The welfare-state paradises of Denmark and Finland have value-added taxes of about 25 percent. Higher levels (still below the needed 35 percent) are hardly seen. This, and the resistance to Prime Minister Abe’s comparatively small consumption tax increases, pretty much signals the infeasibility of this line of reckoning.
Obviously, what is going on here is that there are many retirees. Workers need to be taxed as much to ensure enough money is transferred to keep the pensioners at their current (and expected) standard of living. I have heard people say that this does not “necessarily” constitute an inter-generational inequity, but I personally find that to be a tough horse-pill to swallow. I’ll let you decide; but let’s look into those inequities.
First, the study goes through other ways to balance the budget. (The government is given the chance to do this in the long run; I’ll go into what this entails in two paragraphs.) Besides substantial “Medicare” copays and deeper cuts to pensions, the analysts also consider cuts to other government programs. For the US that would mean cuts to defense and poverty programs. Of these ways, it turns out that increased copays look the most fair. That is, the other balancing techniques have sometimes impressively large net income redistributions between the generations. These can be something like 10 percent or more of lifetime income. I would imagine most would consider these rather large inequities. Lowering pensions doesn’t level the playing field much from the (massive) consumption tax increases posited in the work.
In pretty much any case, big inter-generational transfers are afoot for Japan, and about any other first-world country with similar welfare-state programs.
Now back to why 2040 is an annus mirabilis. Elsewhere, Congressman Paul Ryan (referencing the CBO) mentioned an interesting phenomenon in fiscal modeling: When the going gets tough, sometimes you can’t find a budgetary solution! For the present study this happened at debt-to-GDP ratios below 4. What this means is, if you don’t change taxes or spending or both by enough and soon-enough (see above) there is no path forward for the government fiscal apparatus. Of course, this trusts, to a certain extent, the completeness of the computer’s algorithm in finding solutions. Let’s soft-peddle that wonkish issue and say these algorithms are “reasonably good” at doing their job (searching for and finding a budgetary solution over time, if one exists).
If that is the case, and the simulations reasonably mirror reality (they aren’t out-of-this world, I assure you; re-distributable GDP doesn’t fall from helicopters, after all), doing nothing up to 2040 means creating a set of real fiscal flows that cannot be sustained long term. ‘Fiscal flows’ here means pension outlays, medical payments, programs, taxation, and bond issuances and payments.
This failure (“epic fail,” the younger generations once said) is the simulation analog of the situation where compounding interest payments take up more and more of government income. That is, the government must float more debt just to pay interest on debt. At that point (interest compounding), things get out of hand fast, and it is safe to say a credit event will occur. For everyone apart from Dr. Krugman and maybe the Kirchners, this is a bad thing.
For this post, because we are happily dealing with the easier and insular Japanese case, it means the necessity of a radical restructuring of how pensions and medical care are financed. That is, the system doesn’t work. But remember what we’re interested in studying is (Japanese) systems that do function. This requires raising taxes, and also - see above - reducing program generosity. (I tacitly assume one cannot realistically jettison the welfare-state apparatus, which, given what political interests are, is a Libertarian fantasy.)
It turns out things are worse. Those earlier (high taxation) solutions already had gradual reductions in a reformed pension payments system baked-in. I’m sure Prime Minister Abe fought hard to make those reforms reality. Our system has these reductions baked-in too; it’s called the Trust Fund, and I’m sure it will prove a battle-ground in time as well (cf. Greece, Detroit).
So, even that reduction is not enough. The authors of this important paper find Japan also needs a “Medicare” copay of 30 percent, in addition to (touching) Denmark’s rate of taxation. Well, I guess we know what the future looks like.
Except, it will probably be worse.