Ricardian confusions squared

Hang on now, Paul, Mark, Progrowthliberal, and Antonio Fatas. Justin Yifu Lin (pdf) is basically right. He's making an important point. It's similar to one I made a couple of years back.

Sure, a temporary increase in useless government expenditure, even under Ricardian Equivalence, will give a bang for the buck. But it's a totally useless bang. No Keynesian economist would recommend it, even if there were no useful alternatives. And useful government expenditure will not only give a useful bang for the buck, it will give a bigger bang for the buck.

The intuition is simple. Paying unemployed resources to do something totally useless is equivalent to giving them a transfer payment. A transfer payment is equivalent to a tax cut. A tax cut has no effect under Ricardian Equivalence.

[Update: Paul Krugman responds.]

The Ricardian Equivalence Proposition says that a bond-financed increase in government expenditure is equivalent to a tax-financed increase in government expenditure. Because consumption depends only on permanent income and so on the present value of current and future tax liabilities. Assume Ricardian equivalence is true.

Assume, for simplicity, that the marginal propensity to consume out of permanent income is one. Ignore interest rates, exchange rates, and supply-constraints.

A permanent $100 increase in government expenditure will have zero effect on output, if it causes an equal $100 increase in permanent taxes, and so an equal $100 decline in consumption.

A temporary $100 increase in government expenditure will cause output to increase by less than $100, because it causes permanent taxes to increase by less than $100, and so consumption to decline by less than $100.

In the limit, as the $100 increase in government spending becomes purely transitory, we can ignore any offsetting effect of increased present value of taxes on current consumption. So output increases by $100. Just like the balanced budget multiplier in the old Keynesian Cross.

Now assume that the $100 increase in government expenditure is spent on something totally useless, like digging a hole and filling it in again. Which is what Justin Yifu Lin assumed. Does output really increase by $100?

Well, if you count that totally useless hole-digging-and-filling exercise as part of GDP, then yes, GDP does increase. And if you don't, it doesn't. And National Income Accountants can have great fun discussing whether the true value added of a filled hole is any different from the value added of a hole that doesn't exist because it was never dug in the first place. But it's still a totally useless activity. And no reasonable measure of output would include digging a hole and filling it in again for no purpose whatsoever. Its market value is precisely zero.

In terms of useful output, the multiplier of a useless project is precisely zero. And remember, there is no effect of increased consumption from the extra $100 earned by the people digging the hole, because it is exactly offset by the extra $100 present value of taxes. No Keynesian should recommend government expenditure on totally useless projects, under Ricardian Equivalence.

Now suppose the government spends the same $100 to dig a hole, plant a seed in the hole, and fill it in again. And the seed grows into a tree, and the tree produces exactly enough apples to repay the $100 the government borrowed including interest. So the government never needs to raise taxes. And people know this. What's the effect on output?

At a minimum, output increases by $100. And it really is a $100 increase in useful output. Unemployed resources, that would otherwise have produced $0, now produce something worth $100. And since that $100 increase in real wealth causes an additional increase in consumption expenditure, there will be a multiplier effect bigger than one.

Now suppose the government plants a tree at the cost of the same $100, but the tree now yields enough fruit to pay back double what it cost to plant. So the government can actually cut future taxes. You get the same direct effect as before, but real wealth has now increased by $200, which induces an even bigger additional increase in consumption expenditure, and an even bigger multiplier.

That's what Justin Lifu Lin is saying. (OK, that's a reasonable interpretation of what he's saying). He's right. If Ricardian Equivalence is true, then useless government expenditure, even if temporary, even if there are zero other sources of crowding out, will have a multiplier effect on useful output of precisely zero. And useful government expenditure, which requires no increase in present or future taxes, will have a bigger multiplier. And if it's even more useful, and allows a cut in future taxes, or an increase in future disposable income, it will have an even bigger multiplier.

Useful government expenditure doesn't just create a useful bang for the buck; it creates a bigger bang for the buck.

Comments

[Jacques René Giguère]

Yes. But Keynes was already aware that sometime we refuse useful spending. " We are so rationnal that we refuse to build a second railroad from London" (or Birmingham , I Kesey myself on the road to a political meeting so I quote from memory)to Manchester."
Anyway ,the thing is we must move money income to people who will use it differently if it looks as coming from work thanfrom tax cuts or lower interest rates. So the problem is often not about useful vs useless spending but about spending or not or even doing not very good vs doing detrimental things...

[James]

Fantastic post! However, you disagreed with Paul and Mark, so they will ignore you if you're right and criticize you if you're wrong.

[Mark Thoma]

James:

I see you're still angry because I didn't post the incorrect rebuttal to Krugman you emailed to me. Oh well.

Nick:

As you can tell from what everyone wrote, we all assumed he was talking about changes in government spending, or at least I did. For example, he talks about "traditional Keynesian investment" when describing the traditional policies, then asks "But how can the Ricardian trap be avoided" before ever mentioning digging holes and filling them in again (which I took as meaning spending that does not enhance productivity but does have value, i.e. that he was drawing the distinction between government consumption and government investment). So your example is correct, but it's debatable whether he meant, essentially, pure income transfers given what he says in other places -- it's not at all clear. It seems more likely he meant non-productivity enhancing government spending, i.e. government consumption, but, again, it's hard to tell.

In any case, contrary to your assertion in the comment you left, what I said wasn't wrong -- I said the empirical evidence does not support a full offset even when the changes satisfy the necessary theoretical conditions. That has nothing to do with whether what he was suggesting can really be interpreted as a change in taxes rather than a change in spending.

[James]

It seems to me that three scenarios are economically equivalent:

1) The government pays people $50 to dig holes in the morning and another $50 to fill them in during the afternoon.

2) It pays people $100 to guard random pieces of land to guarantee that they are in the same condition in the evening as they were in the morning.

3) It goes around handing people $100 bills.

In scenario #3, there are two ways to do it: Fiscal policy and monetary policy. If you use fiscal policy, the Treasury has to pay annual interest on the money it borrowed to hand out the $100 bills. If the Fed hands out $100 bills (via open market operations), it pays no interest.

[Adam P]

Nick: "Assume, for simplicity, that the marginal propensity to consume out of permanent income is one"

Seems to me that you're only correct here if the mpc out of transitory income is exctly zero. If it's positive then Krugman, Delong etc. are right.

The reason is simple, we agree that the dig-a-hole-and-fill-it-again type of stimulus is just transfer, fine. Suppose the $100 transfer is financed from a bond and assume the bond is a perpetuity, it's coupon would be somewhere around 5% I guess. Let's assume the coupon is exactly 5%.

Then, if the mpc out of transitory income is 0 then the multiplier is zero as you stated, taxes are permanently higher by $5 per year, causing permanent income to be $5 lower, causing consumption to fall by $5. This is offset by the $5 increase in permanent income from saving the $100 transfer (and investing it in the perpetuity that the government issued).

On the other hand, suppose the mpc of transitory income is greater than 5%, say because the government targets the transfer to low income, liquidity constrained, people. Suppose the mpc of transitory income is 20%.

In this case permanent income has fallen by $1. Since $80 were saved and earn the 5% return, adding $4 back to permanent income to offset the initial $5 decline, the total fall in permanent income is $1.

Further, $20 of the $100 transfer is spent directly out of transitory income. Thus, today we have a $1 reduction in consumption due to the $1 fall in permanent income and a $20 increase, thus today's AD expands by $19.

If mpc out of transitory income is anything greater than the interest rate on a perpetuity then the transfer expands AD today, even with full Ricardian Equivalence. You are only correct in this post if the mpc out of transitory income is zero.

[Frances Woolley]

James: "It seems to me that three scenarios are economically equivalent:" In this case, I'll choose option (3) please. Seriously, given the very different impacts they have on the well-being of the person receiving the $100, they *can't* be economically equivalent. It all comes down to whether or not one enjoys digging holes.

[Nick Rowe]

Frances' point is totally correct (damn micro!). I suspect James would agree. But since it only strengthens my case against useless government expenditure, I'm going to ignore it and assume the unemployed resources have zero opportunity cost and don't care what they spend their time doing.

Mark: OK. What I should have said was "I think you are all misinterpreting what Justin said and missing his point which is both right and important". It was late, so I wrote "you are wrong" and went to bed!

He *is* talking about government spending, not transfer payments. But I don't interpret him as talking about government consumption (paying for an enjoyable party) vs investment. I think he is talking about investments that are useless vs investments that are useful. Digging a hole and filling it in again sounds more like useless investment than enjoyable consumption. But spending on useless investments is *equivalent to* a transfer payment (except if the unemployed would prefer getting $100 for doing nothing than $100 for wasting their time doing something useless).

Adam: I'm pretty sure we agree on substance, but disagree on terminology.

Suppose you get a lump sum $100 (stock, not flow). The (real) interest rate on a perpetuity is 5% per year. For the infinitely-lived agent/dynasty assumed by REP, permanent income rises by $5 per year. And if the mpc out of permanent income is 1, consumption rises by $5 per year. Same as the interest rate.

Full agreement on above. Just, terminologically, "transitory" doesn't strictly mean "temporary". It means "current minus permanent". So you can either think of a "transitory" increase as the limiting case where the duration of the increase approaches zero, or as a current increase offset by a future decrease leaving the PV unchanged. So if there's a flow $100 per year increase in G, as the duration of that increase approaches zero, the effect on permanent G approaches zero, and it becomes purely transitory. If there's a one-time stock $100 increase in G, spread out over 1 year it will be $95 stock transitory, spread out over 6 months it will be $97.5 stock transitory, etc.

I think I've got that right. But you can see why I wanted to avoid getting into this in the post!

But the basic point is that, if consumption depends on permanent income only, and if transitory is defined as current minus permanent, then the mpc out of purely transitory income will be zero, while the mpc out of current income will be above zero, and will equal r times the duration of the increase in current income (approx).

[reason]

Nick,
so you are saying that starving people will save food stamps?

[Nick Rowe]

reason; no.

[reason]

P.S. I realise adam P. this is your point. I suppose Nick is arguing that are taking the (purely interpreted) permanent income hypothesis as given - and pure Ricardian equivalence. I just think it is a bit hard to take seriously. "Assume you have a can opener..."

[Nick Rowe]

reason: OK. But since the whole point of this post is arguing with 4 other bloggers about the *implications* of Ricardian Equivalence, it does make perfect sense for me to say "assume Ricardian Equivalence". Which is what I did.

And even if REP were only half true, so there were "Ricardian effects", the things I am talking about here would still matter.

[Jeremy Fox]

James: "However, you disagreed with Paul and Mark, so they will ignore you if you're right and criticize you if you're wrong."

Or, maybe Paul will agree with Nick while clarifying that he's trying to get at a different point:
http://krugman.blogs.nytimes.com/2011/03/11/ricardian-confusions-continued-seriously-wonkish/

[Sandwichman]

Am I the one really confused here? This seems to be a debate about the relationship between events at a discrete time "t" and events at a discrete time "t+1". Meanwhile, social production spins merrily along on its axis of a continuum.

How do we know that digging and filling holes or burying bottles filled with banknotes is less productive than, say, building a rail line from Birmingham to Manchester? Is the standard of value being used one of subjective utility or of capital accumulation?

The word "employment", which appeared several times in Paul Krugman's post doesn't appear at all in Nick's. So, is this debate about employment and policies to increase it or is it about abstract "bangs" that may be bigger or smaller? If the issue is really about employment, aren't there more urgent and realistic questions that need to be addressed than Ricardian Equivalence?

[RSJ]

Nick, you are assuming that government funded activities that produce no output cannot increase permanent income -- i.e. no spillover or network effects arising from the expenditure that can generate useful economic activity even if the original spending was not useful.

But that will depend on whether there is an output gap or not, not on whether Ricardian equivalence holds.

That's the point that Krugman was making -- even useless activities in a low IR environment have multiplier of 1 and therefore directly add to present period income, and so increase permanent income. "useless" isn't the same as "ineffective".

Btw, this is exactly the point that Keynes was making in Chapter 10, except that he also added the observation that

"When involuntary unemployment exists, the marginal disutility of labour is necessarily less than the utility of the marginal product. Indeed it may be much less. For a man who has been long unemployed some measure of labour, instead of involving disutility, may have a positive utility. If this is accepted, the above reasoning shows how “wasteful” loan expenditure may nevertheless enrich the community on balance."

"[...]If the Treasury were to fill old bottles with banknotes, bury them at suitable depths in disused coalmines which are then filled up to the surface with town rubbish, and leave it to private enterprise on well-tried principles of laissez-faire to dig the notes up again (the right to do so being obtained, of course, by tendering for leases of the note-bearing territory), there need be no more unemployment and, with the help of the repercussions [e.g. multipliers], the real income of the community, and its capital wealth also, would probably become a good deal greater than it actually is. It would, indeed, be more sensible to build houses and the like; but if there are political and practical difficulties in the way of this, the above would be better than nothing.

The analogy between this expedient and the goldmines of the real world is complete. At periods when gold is available at suitable depths experience shows that the real wealth of the world increases rapidly; and when but little of it is so available, our wealth suffers stagnation or decline. Thus gold-mines are of the greatest value and importance to civilisation. just as wars have been the only form of large-scale loan expenditure which statesmen have thought justifiable, so gold-mining is the only pretext for digging holes in the ground which has recommended itself to bankers as sound finance; and each of these activities has played its part in progress-failing something better." [comments mine]

[Nick Rowe]

Sandwichman:

1. We could do this in discrete time or continuous time. It's just a question of which is easier to work with.

2. "Bang for the buck" could be interpreted as either employment or output. That's not the issue in this debate.

RSJ: We are all assuming there's an output gap, with involuntary unemployment. So an increase in AD will increase output and employment. That's not the issue. The question is under what conditions, assuming REP, an increase in G will increase AD, and how much.

[Sandwichman]

Nick: "That's not the issue in this debate."

Right. That's what I'm saying. The debate is sterile because the meaning of "bang for buck" is not even at issue. But Peter Dorman says it better than me at EconoSpeak:

"The question of the day is, why should anyone give RE more than a moment’s attention? In particular, why would smart economists with state-of-the-art training be debating the fine points of what RE would mean if it were true?"

My guess is that RE acts as a kind of screen memory that blocks out other questions where smart economists really don't want to go.

[Nick Rowe]

Sandwichman: I read your blog post. You totally misunderstand this whole question. The debate here *is* about whether an increase in government spending will increase employment and output. Employment and output *are* the issue. It's the *distinction between* employment and output that is not the issue.

[Lord]

But if consumption depends on permanent income only then recessions don't occur.

[RSJ]

"The question is under what conditions, assuming REP, an increase in G will increase AD, and how much."

Of course. But I don't think you got my argument. Any government expenditure, even for a worthless investment, will have a spending multiplier that will tend to cause current period income to go up, and then a tax effect that will cause it to decrease, with some (net) change to income in the current period, but permanent income is the average income (weighted by NPV) of all periods, including the current period. Therefore an increase in current period income will cause permanent income to go up, just not by the full amount, and therefore if people have a propensity to spend of 1 with changes to permanent income, there will be net positive effects even from worthless expenditures if current period income increases at all.

Ricardian Equivalence + PIH doesn't give you enough information to determine what the net change to current period income will be, it only tells you that the tax effect will be enough to repay the original outlay, with interest.

But you do not know if the cost of repaying the outlay with interest is greater than the benefits of the spending multiplier -- even for a worthless expenditure.

Interestingly, this argument says that the deeper the liquidity trap, the more potent fiscal policy will be, which is the opposite situation with monetary policy.

[Nick Rowe]

Lord: No. Permanent income can change. Especially since it is an expectation about the future. And investment can change. And consumption can depend on interest rates, etc., as well as permanent income.

[Nick Rowe]

RSJ: "Any government expenditure, even for a worthless investment, will have a spending multiplier that will tend to cause current period income to go up, and then a tax effect that will cause it to decrease, with some (net) change to income in the current period,..."

No. Under REP, if the government expenditure is effectively a transfer payment, the net effect on current income will be zero. By definition of REP.

OK you guys: time for you all to read an intermediate macro text.

[Sandwichman]

Nick: "Employment and output *are* the issue. It's the *distinction between* employment and output that is not the issue."

I dunno, Nick. If you read my blog post, you're still not getting my point. You say the distinction between employment and output is not the issue. I say that's a problem right there. There IS a distinction between employment and output and that distinction becomes even more important under the circumstances of policy intervention. Surely you're familiar with the fading relevance of the rule of thumb known as Okun's "Law"?

It's not that I "totally misunderstand this whole question" it's that I fail to grasp the pertinence of a debate about a model in which the distinction between output and employment allegedly doesn't matter. Angels on pins on steroids.

Perhaps the difference in outlook between you and I is that you've never stared into the abyss of unemployment in spite of ample credentials and years of experience or worked at jobs several grades below your qualifications because the alternative was welfare. So, elegantly hypothetical discussions based on crapola abstractions and simplifications don't strike you as arrogantly meaningless displays of virtuosity.

O.K. Solving the problems people who are unemployed have is not "the problem". The problem is getting the symbols on two sides of some blackboard exercise to match up. Fine. I've got no objection to your intellectual priorities, then, Nick. It's your blog, man. Please understand, though, it's not that I'm "wrong". It's that I'm talking about something different.

[RSJ]

OK, a transfer from someone that has a lower propensity to consume to someone who has a greater propensity to consume or are liquidity constrained will have zero effect on current income?

And wouldn't unemployment insurance -- or employment of those who, if it weren't for the transfer, would otherwise have zero income -- fall into this category? Even if they were paid to dig ditches?

Moreover, government purchase of output will have a positive multiplier in the current period, and even a ditch digging program will require the purchase of a few shovels, right? Even if the government requires that workers bring their own shovels, still they will need to go out and buy them to join the program.

Which textbook do you recommend to set me straight? I have Blanchard and Fischer, and can't find your argument anywhere in there -- they are much more circumspect in their conclusions.

[Nick Rowe]

Sandwichman: Look at the way that Paul Krugman, Mark Thoma, and I argued. We (implicitly or explicitly) set aside a massive host of issues that we would *not* argue about. We assumed: Ricardian Equivalence; unemployed resources; a bond-financed increase in government spending; and loads of other things. That let us focus in on two questions: temporary vs permanent changes in government spending; useful vs useless government spending. And after 36 hours, with one back and forth, we are all now in basic agreement on these issues, as far as I can see. (Which is damned fast; Yay the blogosphere!).

We actually accomplished something, precisely because not everything was an issue.

One of the issues we set aside was Okun's Law. We all assumed that changes in output and changes in employment were essentially the same thing. That let us speak a common language, and actually get somewhere.

If you want to argue about Okun's Law, and labour hoarding, and the relation between changes in output and changes in employment when fiscal policy changes one or the other, that's fine. But it's a separate argument, and I can't see any obvious relation to the argument I was having.

So, from my perspective, your talking about the distinction between output and employment, in the context of this debate, makes about as much sense as a feminist (say) wading into one of your blog posts on the lump of labour fallacy and saying "What about the distinction between male and female employment???" "What do you mean it doesn't matter???". She would just come across as someone with an idee fixe, who can only see the world through one lens.

More studies in applied orthogonality.
http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/07/hst-and-jobs-more-studies-in-applied-orthogonality.html

Sure, there's a time and place for examining the distinction between output and employment. But this ain't it.

And, by the way, Keynes' GT was not about that distinction either.

Please try to consider my (implicit) advice carefully. Because this goes way beyond this particular blog topic.

RSJ: One thing I am really crap at is recommending things for people to read. I am so bad at this myself, I should not have been giving other people advice on this.

Somebody else please help me here. What is a good thing to read on the basics of Ricardian Equivalence?

OK. Wiki is a good place to start: http://en.wikipedia.org/wiki/Ricardian_equivalence

Ricardian Equivalence says that a bond-financed increase in G is equivalent to a tax-financed increase in G. And, as a direct corolorary, a bond-financed tax cut has zero effects. AD does not shift. IS does not shift. AE does not shift. Etc. People save the whole tax cut.

Yes, RE only works exactly under a whole host of assumptions. Permanent Income Hypothesis plus Rational Expectations are just two.

[Sandwichman]

Nick, I do get what you're saying. And, by the way, it happens quite frequently that the proverbial "feminist" will wade into one of my blog posts to argue about "the distinction between male and female employment". Usually, though, the idee fixes come from very conventional notions about income being assumed proportional to hours or quasi-fixed costs carved in stone.

Just to reassure you, I don't resent being told I'm monomaniacal. That's the story behind the Sandwichman persona, you know. It's a strategy; not a disorder. The idea is that there are certain topics of conversation that are systematically avoided not because they are irrelevant but precisely because they would shake the foundations of conventional belief. The common metaphor is "the elephant in the room".

Back in the '60s, when second-wave feminism started, the only way a feminist would ever get the overwhelmingly male economists to even acknowledge issues arising from "the distinction between male and female employment" would be to inappropriately bring it up in the course of a conversation about something less threatening to male economist privilege.

If I may explicate your implicit advice, you're telling me to f' off if I won't stay on topic. It's your blog and it's your call as to what is on or off topic.

[Nick Rowe]

Sandwichman: thanks for taking that the right way. Hmmm. Maybe you're not so monomanaical as I thought!

If you think it's really on-topic, then have another go at convincing me. But I just don't see it. Suppose, for example, that firms never laid off workers when sales and output declined. They just kept them on the payroll, doing nothing. Why would it affect the argument here?

[Sandwichman]

Thanks, Nick, you're too kind. But Sandwichman IS monomaniacal. It's Tom Walker who is less so.

Unless I missed something, Justin Yifu Lin didn't say anything about firms never laying off workers when sales declined.

What he was concerned with was 1. underutilization of capacity in high-income countries; 2. state debt problems in the US and sovereign debt problems in several other countries; and 3. low interest rates that combined with capacity underutilization result in capital outflows and 4. spikes in commodity prices that destabilize regimes in poor countries. In response to the dilemmas posed by those conditions, Yin proposed a Marshall Plan-style infrastructure investment in developing countries.

Yin briefly mentioned avoiding the "Ricardian trap" as the rationale for going "beyond conventional Keynesian stimulus of 'digging a hole and paving a hole' by investing in projects which increase future productivity." But it would seem to me that the conditions Yin was concerned with have more in common with the historical post-war European conditions confronted by the Marshall Plan than they do with more abstract theoretical issues of Ricardian Equivalence.

At this point a little narrative correction may be called for. The Marshall Plan has a nice ring to it, enhanced by the reputation of it having been a galloping success. Who remembers (or cares) that half-way through the plan, it was in crisis with no political prospect of being either completed by or renewed in 1952 (See Block, The Origins of International Economic Disorder, pp. 92-108).

That crisis of the Marshall Plan elicited responses from the U.S. Treasury, the European Co-operation Administration and the U.S. State Department. In the end, it was the State Department strategy of massive military rearmament, as proposed in National Security Council memorandum NSC-68, that won the day.

It would be nice if it had been the Marshall Plan and not Cold War rearmament that had resolved the economic conundrums of 1949-50. What we got, instead, was what General Eisenhower denounced during his 1952 presidential campaign as a "deceptive prosperity" based on a "calculated policy" of inflation and "the continued failure of our foreign policy [as] the only way to pay for the failure of our fiscal policy." In other words, Ricardian Equivalence or no Ricardian Equivalence, it was political expediency that prevailed. To then evoke false memories, 61 years later, of the Marshall Plan's ultimate success would be to motivate a proposal with a fantasy.

So, I guess that's my punch line. Yin offered avoiding the Ricardian trap as the theoretical rationale for his proposal, but his selling point and close was a mythologizing of the Marshall Plan.

"In the post WWII period, the United States helped herself and the world avoid a post war recession by an aggressive investment in the inter-state highway system and the Marshall Plan for European reconstruction. The government and financial sector in the United States can play a leadership role again in bringing the world to a 'new new normal'."

[RSJ]

OK, perhaps you are assuming that part of the definition of Ricardian equivalence is say, homothetic preferences and that the economy is not growing, etc. In that case, you are right!

But from my reading, Ricardian equivalence is an effect, not a model, and you should be able to apply it to a variety of models, even if those models contain growing economies or heterogenous preferences. In that case, the actual stipulation is that perfectly foresighted and perfectly altruistic dynasties will set aside funds necessary to pay the future tax burden created by present government deficit spending. That should be the bare-bones Ricardian Equivalence, and you need to work out what that means, in practice, in each model.

In particular, it does not specify whether preferences are homothetic or not, or what the target debt/GDP ratio is, or that this ratio be zero (e.g. that debt be "repaid"), or even that this ratio is constant over time. Blanchard and Fischer emphasize this point.

Whether or not the increase in G is stimulative is left open -- it depends on additional considerations. If preferences are heterogenous, then even tax-financed transfers can be stimulative, in which case tax revenues will go up as a result of the transfer, so that the equilibrium tax burden is lessened. It depends on who is getting the transfer and who is paying the transfer, none of which has anything to do with pre-cautionary savings per se.

Moreover, if the agents really are perfectly foresighted, then they should know that the interest rate on government debt is, on average, less than the growth rate of the economy, which would be the income growth rate of our dynasties. For any target Debt to GDP ratio, L, there is a target deficit to GDP ratio N = f(L), with N > 0 whenever L > 0, so that regardless of the current debt to GDP ratio, or what level of deficits were run in the past, the economy will eventually reach a ratio of L if all future deficits are exactly equal to N. You can say that infinitely lived agents are forward looking, ignoring any past deficits that occurred, as all that matters is the tail of the sequence.

The running of surpluses is never required if you are certain future deficits are capped. If you are not certain and want to run deficits exceeding exceeding N infinitely often (e.g. periodically during downturns), then you may need to run surpluses during booms, but the size of the surplus will always be less than the size of the deficit, provided that L > 0.

For example, from 1946 to 2010, the U.S. ran deficits 80% of the time, with the average deficit being 2.7% of GDP and the average surplus being 1.4% of of GDP, and yet the debt/GDP ratio declined from 110% to about 62% over that time period -- and I'm ignoring the huge deficits that were incurred during WW2. Foresighted agents would have saved only pennies for each dollar deficit spent, so that deficits would have been stimulative in every period.

Therefore for a variety of reasons, you can fully accept that households save sufficient income to pay for future tax liabilities while still believing that changes in G are generally stimulative to AD, and that they are not stimulative only in some weird corner cases -- e.g. useless spending *and* homothetic preferences *and* that you are in a boom period *and* the current debt to GDP ratio higher than the target ratio.

[Determinant]

What he was concerned with was 1. underutilization of capacity in high-income countries; 2. state debt problems in the US and sovereign debt problems in several other countries; and 3. low interest rates that combined with capacity underutilization result in capital outflows and 4. spikes in commodity prices that destabilize regimes in poor countries. In response to the dilemmas posed by those conditions, Yin proposed a Marshall Plan-style infrastructure investment in developing countries.

Careful what you use the Marshall Plan as evidence for. Marshall Plan funds were grants that were used exclusively for purchase of goods from the United States. They were recycled as much into the US Economy as the European economies the program helped. Buyers on the European end had to pay for their imports in local currency. The continuous account surplus this generated was used to finance additional spending. It still is in Germany in the form of the capital of the KfW Bank. Early purchases were staples like food, later purchases focused on capital equipment.

I also fail to see that mythologizing the US Interstate Highway system proves anything. The US already had an effective intercity and interstate transport infrastructure in place in the form of railroads and had had it since the 1860's. Other than allowing the development of interstate trucking and car travel between cities, I fail to see that the interstates improved point-to-point freight or passenger transportation much.

Though there is the historical point that everyone feared a return to Depression in the late 1940's. The popular thinking went that the Civil War was followed by the Long Depression of the 1870's - 1890's, the First World War was followed by the Great Depression and World War II was bound to be followed by hard times too. Stocks traded at extremely low price-to-earnings ratios like 3 - 5 (I'm serious) in expectation of this. That it never occurred is beside the point, popular opinion was wrong in this case.

[Patrick]

Sorta going OT but it's safer than swimming with the RE sharks...

In discussions about why the world didn't fall back into the trap of the 1930's post WWII I think we tend to neglect demographics and psychology. Say you where a young man (American, Canadian or Brit) in your early 20's when the war started. You'd grown-up in the depression. Your parents had come of age during WWI and lived through the nasty recession that followed, So you'd already had your fill of hardship. Then you spend the better part of your 20's with someone trying to kill you and watching your friends get killed. Then along comes Nuremberg as the icing on the cake, incase you hadn't really appreciated the depth of despair and savagery that were possible. They must have been so damn sick and tired of misery, death and destruction by that point. So what did they do? They went home and had sex. Lots of sex. And as a consequence they made babies. Lots of babies. And babies need stuff. And they grow-up into adults who need stuff. And the government has to build schools to educate them in, and libraries and swimming pools and baseball fields, etc ... Babies make parents need houses to put them in, and houses need to be filled with stuff. Somebody has to make all that stuff and build the schools etc ... So it seems plausible to me that it was babies saved the world from another Great Depression.

Similarly, I think it's more than coincidence that as the population ages we find ourselves with an increasingly moribund economy.

[Nick Rowe]

I have a very strong hunch that Patrick may be right on the demographics. Been wondering about this for some time. Though this is not something I have thought about or looked at enough. Is it a coincidence also that Japan turned Japanese at around the same time it started turning old? Now the whole world is aging, for the first time ever? Someone said recently that China will be the first country to get old before it got rich. This is a big global experiment. And yes, demographics must be the biggest thing affecting savings and investment. Think of all the stuff you buy when you get a job then start having kids. Then stop buying a decade or two later. Add demographic swings to the opening of global financial markets.

But again, this is something I have never sat down to think through theory and data rigourously.

RSJ: REP is about whether bond-financed tax cuts shift the IS and AD curves to the right. It does not assumed the economy is in a boom. It says nothing about AS. Yes, it requires a lot of assumptions to make the conclusions hold exactly. But not that one. The AS curve only decides whether a shift in AD raises P or Y. If the AD does not shift the AS is irrelevant.

[Nick Rowe]

Sandwichman: "Unless I missed something, Justin Yifu Lin didn't say anything about firms never laying off workers when sales declined."

I would have been very surprised if he had. It would have been totally irrelevant to his topic, as far as I can see.

"Yin briefly mentioned avoiding the "Ricardian trap" as the rationale for going "beyond conventional Keynesian stimulus of 'digging a hole and paving a hole' by investing in projects which increase future productivity." But it would seem to me that the conditions Yin was concerned with have more in common with the historical post-war European conditions confronted by the Marshall Plan than they do with more abstract theoretical issues of Ricardian Equivalence."

Yep. There is something very very weird about this minor blogosphere war. Yin makes a passing reference to REP, and how productive government investment spending will offset those Ricardian effects by making people richer in future. A perfectly correct point said briefly in ordinary language in what was *not* a theory paper, and everyone jumps all over him.

Maybe there's some agenda here. I had never heard of Yin before this blew up. I don't know who's who in economics. Maybe he's a U of Chicago graduate, or something. Dunno.

[Nick Rowe]

Bingo! I guessed right! Yin has a Chicago economics PhD, 1986. The first Chicago economics PhD from red China. http://en.wikipedia.org/wiki/Justin_Yifu_Lin

[Kevin Donoghue]

Nick,

Krugman is hardly being uncharitable when he interprets “an outcome where the government stimulus fails to boost aggregate demand because economic agents expect future tax increases to pay for larger deficits and thereby increase savings” as just another rehash of “the supposed story in which government spending is frustrated by a fall in private spending.”

Quite likely Yin’s critics never heard of him before either. They just saw a sadly typical example of Chicago thinking and quite naturally they objected. Of course it’s the likes of Fama and Cochrane who have sensitised them, otherwise they probably wouldn’t have noticed.

[Christopher Jillings]

@Nick and @Francis: I's like to echo Frances' comment in a different way. I dislike Krugman's comment "And maybe we should consider make-work jobs just a form of transfer payment, so that employment, properly measured, doesn’t rise either — although I would quarrel with that; a job feels like a job whether or not, in the judgment of an outside observer, it’s adding value."

Sorry, no. A menial job in which you know you are contributing enhances, dignity, well-being, and social cohesion. A job in which you are merely digging filling holes does not. Now, if between digging and filling the hole you place an acorn in the ground, the physical labour becomes meaningful.

A jobs program is great when needed. But build something useful - a park, a road, a new cross-country-ski house (to use a local stimulus example).

[Nick Rowe]

Kevin. OK. What Yin said is open to interpretation. Maybe I interpreted him too charitably; maybe they interpreted him too uncharitably. Yeh, I shouldn't assumed they went for him because he's from Chicago. But I really did surprise myself when I found my wild guess was right and he was from Chicago!

Christopher. I tend to agree with you. In a previous post, Greg Ransom said that his grandfather quit a job in disgust, in the 1930's, even though he needed the money, when he learned it was a make-work program doing nothing useful. We all (OK, at least some of us) really want to believe that our jobs and lives are worthwhile. Otherwise, just give the unemployed handouts, and let them try to find something meaningful to do with their time themselves. Some will succeed.

[K]

[OT
Nick:"handouts"

Yup. But maybe find a different way to characterize it. Like obviously meaningless work, people are demeaned rather than elevated by charity. The truth is that the earth yields an enormous wealth *independent* of labour, and that this wealth is increasing with productive capacity of available technologies. Every new human deserves an equal claim on the total unimproved land value of the earth. Entitlement. Property. Dividend. Not a handout. ]

[Nick Rowe]

K: Wot! You gonna steal some of my Ricardian land rents? I already pay tax on them! (Same Ricardo, different theory, of course).

Maybe. Not sure it's quite that easy to re-frame. The English aristocracy lived off Ricardian land rents. Eventually they invented fox hunting, to make their lives meaningful again. It takes a long time to build a culture. Until then, they waste their lives sniffing fine brandies or glue. Paradoxically, the Labour Theory of Value and Distribution is so deeply ingrained within us all now, I'm not sure it could work.

[GabbyD]

i have a basic question:

" But it's still a totally useless activity. And no reasonable measure of output would include digging a hole and filling it in again for no purpose whatsoever. Its market value is precisely zero."

if market value is the determinant of whether or not an activity is included in GDP, then how is goverment spending generally considered to be part of GDP? if its really the provision of goods that arent provided by the market, does that mean these public goods are not counted?

if the public good is UNDER provided by the private market, then govt provision is OVER counting the provision of goods?

[anon]

GabbyD, government-provided goods are generally included in GDP on a cost basis. But GDP should not be expected to be a reasonable measure of "value", much less general well-being. It is best used as an indicator of short-term fluctuations in economic activity, which is obviously very useful in a macro context.

[Nick Rowe]

Gabby: anon said it very well.

[paine]

i love nick dearly
how can you not like him
he's almost perfectly cloud cuckoo

time after time
he's barking at the wrong bus
usually a parked bus

[GabbyD]

@nick, anon

thanks. the sentence following my quote says: "And no reasonable measure of output would include digging a hole and filling it in again for no purpose whatsoever. Its market value is precisely zero."

so, its clear that digging a hole and filling it up is part of govt expenditure, so increases GDP. does this mean that GDP is NOT a reasonable measure of output?

[K]

Nick: "You gonna steal some of my Ricardian land rents? I already pay tax on them!"

What ya gonna do about it, Atlas? Stop producing land??? But seriously, I would also get rid of consumption and income taxes so nobody is going to go Galt. You are definitely paying some land tax as a part of your municipal taxes, but not enough. Otherwise you wouldn't be speculating on land. The whole point is to reclaim the entire unimproved land value: i.e. the land tax has to be equal to the risk free rate on the market value of the land.

Could it work?  On the one hand it would increase the wealth of poor people so assuming concave utility, their marginal utility of income is decreased. On the other hand:

1) People would pay a lot less or no income tax, they would not lose welfare if they work (there's no welfare in the first place), and there would be no minimum wage. Many people would be moving from 100% or greater marginal tax rate to zero. I suspect it would increase the number of people engaged in paid employment.
2)  It has been argued that utility of consumption for poor people is convex.
3) the wealthy would have to invest in productive capital rather than just monopolizing idle land (yeah I'm looking at you Nick, and I'm guilty too). 

But hey, it's not all or nothing. The point is to start taxing externalities (unimproved land value is a positive externality created by economic activity in general but earned by the land owner) and paying them out as a citizen's dividend and eliminating or reducing distortionary taxes (income *and* consumption). These efficiencies can be phased in gradually so there is no cause for concern that it's incompatible with our culture. We can observe the effects and adapt the programs (and our culture) as we move ahead. 

[Jacques René Giguère]

Patrick's post on the baby-boom is the standard Richard Easterlin thesis. There have been disagreements ,some quite strong, but it still holds well.

Adding to Determinant
As for the effects of the Marshall Plan
1) in North America it replaced part of the wartime spending and sent money income to workers ( and money income from work is not a tax cut or a transfer in anyone's head)
2) was structural in Europe. It could not provide money income as the goods were transfered ( the payments was a scheme to balance the books and let the conservatives and republicans think that we were paid, even though in fact we donated the stuff...
3) started or restarted the european retail credit system. The goods were sold on credit to final users and the local currency "paid" to US and Canada so we could pretend we were paid even though these currency balances were worthless as there was nothing to buy ( otherwise there would have been no need for the Plan).Later we would use those balances to pay occupation and then NATO local expenses. We even encouraged veterans to go back visiting the battlefields. What real dollars we collected from them so they could get european currencies was the real payment for the Plan.( plus goods received in exchange for occupation expenses)
4) we provided both foreign exchange and money income to european workers plus useful things through the "Offshore" program. We would pay european factories
in hard currency to produce various things, mostly military goods.

5) the total effect in GDP% was rather small but the psychological and political was huge.

Nick:
Yes, a menial job could feel as bad as a handout,though my grand-parents had no trouble accepting the "perjury law" money once the priests explained that lying to the government under duress to feed your family was no sin ( in fact the civil servant extracting an oath that you didn't get any other meagre income was himself guity of "simony" which can send you to hell).
That why FDR, perhaps the politician in history who best understood and used psychology, called his hole-digging programs with lofty sounding names like the Civilian Conservation Corps.

[RSJ]

"The whole point is to reclaim the entire unimproved land value: i.e. the land tax has to be equal to the risk free rate on the market value of the land."

A business wants to open a factory, so it buys land and builds a structure, and this requires an investment of $100 for the land, and $100 for the structure. It will need to earn the risk free rate on the $200 investment (otherwise it will not invest).

But in addition to that, it will need to pay the risk-free rate on the $100 of land? That means that it needs to earn the risk-free rate on $300?

You can only charge the risk-free rate on land if land is free, otherwise no firm will be able to afford to buy land. But landlords are firms, with the same cost of capital as everyone else -- the risk-free rate.

Similarly, a household borrows to buy a house, paying the risk-free rate on the land twice? Once to the bank, and then again to the government?

That would be a subsidy to (drumroll) landlords!

[anon]

RSJ, in practice, you wouldn't tax *all* of the rent. You'd tax, say, 50% or 80% or 90% of it (depending on what works best) and the market value of land would decrease accordingly. Some of the rent would still be appropriated by the landlord, but it's better than nothing.

[Nick Rowe]

GabbyO "so, its clear that digging a hole and filling it up is part of govt expenditure, so increases GDP. does this mean that GDP is NOT a reasonable measure of output?"

Correct. But, it may still be a useful measure of output for other purposes, even if it does not measure useful output.

Bottom line: just because GDP goes up, doesn't mean we are better off. We've known that ever since GDP was invented. There are lots of counterexamples, this is just one of many.

[K]

RSJ: What anon said. In the theoretical case of a tax on 100% of the economic value, the land would have no market value. It should make no difference to your entrepreneur or home owner. Instead of paying interest on the $100 to the bank, they would pay it in tax. Nobody would ever have to finance land, since it would be free. You just have to pay the tax. In practice though, as anon says, you probably have to leave some of the land value to the owner.

[RSJ]

K,

There is a MRS between land and produced capital. Someone can build a fancier house in a less desirable area, and sell it for the same price as a lower quality structure in a more desirable area.

Only in the idealized case in which all locations are equally desirable, or there is no substitution between location and amenities can you argue that the tax is efficient.

But in reality, it is an inefficient tax.

Taxes on ownership of land succeed in lowering the price of land only because they succeed in lowering economic activity, so all produced goods become more scarce relative to the non-produced good. You are moving people off of their ideal location/amenities trade-off by making the amenities more scarce. That is the only way that a tax on land ownership can result in the price of land declining.

On the other hand, you will succeed in seizing rents by imposing a tax on the capital gain of the person selling the land, and this will not discourage economic activity.

But a tax on capital gains is not the same as a tax on ownership.

[Nick Rowe]

RSJ: the key assumption here is that the supply of land is perfectly inelastic. A tax on an good that has a perfectly inelastic supply (or demand) curve will not affect the quantity of that good exchanged. It will change the distribution of income, and any change in the distribution of income will (generally) cause a change in allocation. But it doesn't introduce a distortion from Pareto Optimality. Second theorem of welfare economics. Henry George.

[uire]

Let me modify your statement:

A tax on an good that has a perfectly inelastic supply (or demand) curve [in that period] will not affect the quantity of that good exchanged [in that period].

But there is no requirement that any land be sold in a period, nor is there a requirement that any land be bought in a period. The mere existence of land does not mean that the full stock must be exchanged every period.

[anon]

RSJ, not sure hat you're getting at. There is a cost to _occupying_ land by building something on it, but it's a cost that the developer pays regardless. Even if she owns the land, she faces an opportunity cost: she could have leased the land or sold it or built something else. A tax on rent simply transfer this cost and makes it more tangible.

[anon]

^ *what, **transfers

[uire]

By the way, this is a real problem with using the welfare theorems, which make no reference to time, and yet the supply and demand curves must represent flows of transactions, as that is what determines prices in each period.

These types of stock/flow issues are a serious problem. In particular, you wouldn't believe that capital income "is not income" if you didn't believe that capital is circulating (e.g. that it fully depreciates every period). It all falls apart as soon as you stop equating stocks with flows.

[Nick Rowe]

uire: Yep, but it's a tax on land rents, not a tax on land sales.

[RSJ]

For some reason, the posts by "uire" were by me.

Anon -- just assume the developer buys land, builds a structure, and then sells it, since this is what actually happens. Or, looking at the landlord, the landlord buys land, builds a structure, and then rents it out.

People do not get land for free that they can rent out. They need to buy the land first, and then they rent it out.

When determining what is an economic rent or not, you need to take into account the investment made by the landlord or homebuilder.

Only to the degree that they are earning a return in excess of the risk-free rate on their investment are they earning an economic rent.

But if land can be freely bought and sold, then as soon as an area becomes popular, the price of the land will climb so that there will be no excess return from being either a landlord or a homebuilder.

And that capital gain will accrue to the current (incumbent) land owner, so only that owner should have the rent taxed away in the same period in which the price increased.

But after that tax is paid, in the next period, no additional taxes should be paid unless the price of land increases again.

The tax is paid only on the gain that occurred in that period. It is not paid on the full market price each period.

Now given that land prices in most of the country are roughly constant most of the time (in real terms, and gains should be measured in real terms), and land prices only in the select areas tend to climb, that means that the government's tax receipts from economic land rents will be meager.

And then you have the issue of bubbles -- e.g. do you send rebates out for ad-valorem taxes when prices decline?

[RSJ]

"uire: Yep, but it's a tax on land rents, not a tax on land sales."

LOL. The goal is to fully tax away economic rents that accrue to those who own land. Those who own land have the option, in any period of

1) selling their land
2) renting their land

Whether they select 1) or 2) will depend on what is most profitable for them. Similarly, those who need to make use of land have the option, in any period, of

1) buying land
2) renting land

And they will select what is most profitable for them.

Therefore neither the rental of land, nor the sale of land, is inelastically supplied or demanded in each period.

Henry George was right "in spirit" but wrong in the details.

The details are such that the economic rents can be fully taxed away every period by imposing a 100% tax on the ad-valorem capital gain in that period by whoever happens to own land in that period. You do not want to tax the full market value each period. You can tax whatever you want, but taxing the full value every period is not going to be an efficient tax.

In practice, there are lot of complications -- e.g. land price declines will require that rebates be sent out, credit constraints or zoning laws may prevent land from being freely bought or sold, etc.

[zosima]

What your argument amounts to is a claim that a useless activity should not be counted in our measures of economic activity. But this fundamentally undermines one of the fundamental premises of economics: we don't make subjective or normative judgments of the utility of various activities, but instead use the price paid for the activity as a proxy of its utility.

If we eliminate this premise economics loses any approximation of objectivity. Normally, we count junk-mailers, rent-seeking financiers, polluters, lobbyists, and government bureaucrats in GDP. With good reason. If we start throwing out one type of spending we have to ask what other useless activities we should exclude. GDP ceases to be a statistic reflecting something factual about the economy and simply becomes a reflection of ideology.

In other words, if you don't count paid activities that you view as useless towards economic activity, you're no longer doing economics you're doing politics.

[Nick Rowe]

zosima: but when we are arguing about whether fiscal policy would be a good thing we have already entered that normative world. There are two separate points here though: whether we economists see the re-filled hole as useful; and whether economic agents see the re-filled hole as useful. The latter is a positive question. And the answer to that second, positive question will affect the amount by which output (measured either way) will be affected by the policy. In this debate (to keep it as simple as possible) I have implicitly assumed that economists and economic agents share the same beliefs about the usefulness of the re-filled hole.

For example, if agents thought the hole would grow a very productive apple tree, even if we economists knew it wouldn't, we might recommend digging the hole in any case. Because (under certain assumptions) digging that useless hole might create some useful economic activity as a side-effect, because agents would increase their current demand for consumption anticipating a big crop of future apples. The confidence fairy can work!

[RSJ]

Suppose that our economy has two people, Tom and Jerry, that scratch each other's backs for a dollar. The economy contains $1 that is passed back and forth. There is also a government bank willing to lend anyone $1 as well -- no one is credit constrained.

In the morning, Tom scratches Jerry's back, and Jerry pays him $1. In the evening, Jerry scratches Tom's back, and Tom pays him $1.

Now suppose that one evening, Tom becomes afraid that Jerry wont want a backstratch tomorrow, so he decides to forego the backstratch today, keeping the dollar. Now Jerry sees that Tom has defected, and he defects.

Now supposes the government announces the following fiscal reaction function:

Whoever is unable to sell a backstratch can be hired to scratch a post for $1.

Scratching a post is less rewarding than scratching a real back.

Moreover, the post scratching program will be funded by levying a $1 tax on whoever has the dollar.

With such a reaction function, If Jerry gets worried that Tom will defect, then he knows that he can go scratch a post and rebuild his savings, independent of what Tom does, and moreover he can force Tom to pay for it.

And Tom knows the same.

Both sides are able to impose costs on the other, in the form of a threat of progressive taxation, which restores the economy to the good equilibrium. The relative unpleasantness of the work prevents each person from abusing the program and just imposing costs on each other instead of engaging in productive work.

With perfect foresight, the reaction function is all that is needed to prevent the recession in the first place. No one need actually scratch posts.

But with imperfect foresight, the reaction function will succeed in getting the economy to a good equilibrium, even though it corresponds to useless work -- post scratching -- and does not increase real wealth at all. Nevertheless the policy increases real output by preventing or limiting the duration of recessions.

[Nick Rowe]

RSJ: That's a monetary exchange economy. The government prevents recessions by taxing (threatening to tax) holdings of the medium of exchange (and redistributing the proceeds by helicopter). I like it. But it's more monetary than fiscal policy.

[K]

RSJ:  Under 100% land value tax (LVT), the land value (to the owner) *never goes up*.  It is zero, always zero.  No bubbles, no crashes. Why?  Because you pay the risk free rate on the site value of the land in tax.  Every year.  (It's not a 100% capital gains tax. It's interest on 100% of the site value.)

Anyways, we are not inventing this stuff.  Here's Adam Smith: "A tax upon ground-rents would not raise the rents of houses. It would fall altogether upon the owner of the ground-rent."  I.e. it has *no* impact on the value of houses or quantity of production. As pointed out by Nick above Ricardo elaborated this principle in great detail and Henry George made a whole political philosophy out of it in Progress and Poverty. If you are aware of any decent theoretical objection in the literature (I can't understand what you have written above) , I'd love to see it.

It makes sense to tax ownership of stuff that's in fixed supply because:

1) Being in fixed supply, you won't change the amount of it.  And profitability of marginal production is unaffected by the tax.  Your economy will produce exactly as much as it would without the tax. That beats any other form of taxation hands down.
2) Being in fixed supply there is no moral justification for some humans to have more of it than others. Nobody made the stuff. It's effectively the same thing as collective ownership with right of use rented to the highest bidder. Like the airwaves.

[RSJ]

K, it's not in fixed supply at all. What is missing from all these old-school analysis is the concept of credit markets that would allow the tenant to take out a loan at the risk free rate and trade places with the landlord whenever he wanted to.

At that point, whatever economic rents may exist become capitalized in the market price of the land. The old credit-constrained analysis assumes that land is too cheap, so that renting out earns you a profit in excess of the risk-free rate. But as soon as tenants can become landlords whenever they want, then rental properties are not in fixed supply at all, since they can be converted into user-owned properties and taken off the rental market.

[Nick Rowe]

RSJ: If a bit of land can be rented for $100 per year, the tax is set at $99 per year. The landowner pays the tax whether he rents it out or uses it himself.

[K]

And the existence of the $99 tax has *zero* impact on the amount of rent he can charge since the land will be used for exactly the same purpose as before. Zero deadweight loss.

[K]

And never mind the risk free rate. Think of it as a tax equal to most of the rent like Nick says. Makes no difference.

[K]

Nick: Can you explain why there is so much emphasis (on this blog and in general) on moving from income (especially corporate) to consumption tax when land value isn't near fully taxed? If all these debates are truly about efficiency, then land is the elephant in the room. Let's get rid of income *and* consumption taxes while we are at it. (And yes I understand the reasons for the rich and powerful not to be in favour. Personally I'd like a heads up myself before such a tax was announced. But it's not a legitimate reason to be opposed.)

[GabbyD]

yup, i agree with you.

thanks for ur responses!

[anon]

K, I second your question above, especially since I've got a hunch that some portion of corporate tax falls on firm-specific assets, including commercial real estate. Scott Sumner also supports the land tax (and taxing externalities), but if such taxes yield too little revenue, then a consumption tax is probably the second-best solution.

[Nick Rowe]

K: Put a one-time unanticipated lump sum 99% tax on all forms of wealth now. Land, labour, capital. Same thing. And promise never to do it again. Time-inconsistency.

[Determinant]

K, the reason is simple, and is demonstrated by a few bankrupt English aristocrats. We don't live in a land-dominated economy. Land by itself doesn't represent wealth. Land isn't inelastic, not in an realistic sense of the term. The Crown still owns 70% of Canada but I don't see that appearing on provincial balance sheets as a large item. Why? It's unimproved, remote and therefore nearly worthless.

Land isn't a wealth-creator in our economy. We live in a money/skills/production economy. Those three interrelated things are what represent real wealth. We don't want land for lands sake, we want stuff because we can use it, play with it or eat it.

Those English aristocrats had tons of land with lots of agricultural rents but they still went bust when agricultural prices collapsed as they did and do periodically. It was the new class of merchants, factory owners and skilled tradesmen who had secure incomes and therefore real wealth. The people who made stuff or owned the means to make stuff and even more importantly the ability to make more stuff.

The postage-stamp lots that you and Nick have are irrelevant by themselves. What really matters is the size of your house, it's relative size and features compared to your neighbours and most importantly the fact that it is surrounded by a more or less desirable group of people to live among. That's why Nick's house in Ottawa is more desirable and pricey than an unorganized township in Northern Ontario where.

We want to tax consumption because we want to tax the stream of revenue from stuff. We call it consumption to mean we won't tax investment, that is the ability to make more stuff, er, consumption.

My consumption tax will outprovide your Ricardian Land Tax any day.

[RSJ]

Land is of no value by itself. It must be combined with a structure, and the combination of land + structure must earn the risk-free rate.

By the way, even now you can get free land in several areas, if you promise to develop it. But it may not be economic to do so -- those are marginal areas, so the rents that people can pay are not enough to justify developing the property.

Now say that there is no tax on bond income, and no tax on structure-rent but 99% tax on land rent. Both owner-operators and landlords must pay the tax.

The landlord can sell the building and buy bonds instead. The market value of the building declines up until the after-tax rental yield, net of maintenance, is equal to the risk-free rate.

Landlords will take a *huge* capital loss -- but that is a one-time loss, not an on-going loss.

Other areas -- constrained areas -- have very expensive land, so in those areas, there would be many buildings with a negative present value, due to maintenance, that will be abandoned up until land becomes sufficiently cheap or rental yields sufficiently expensive, so that even in the constrained areas landlords obtain the risk-free rate on land + structure.

Perhaps the immediate effect of this would be migration out of the constrained areas and into the unconstrained areas.

Gradually buildings that have the right characteristics to be viable would remain.

But the results would not be efficient.

[Nick Rowe]

K: I'll leave the above 2 comments to you!

[K]

I'll get back to all of you tonight.

[anon]

"Other areas -- constrained areas -- have very expensive land, so in those areas, there would be many buildings with a negative present value, due to maintenance, that will be abandoned"

Is this really the case? If a building has negative present value after accounting for land rents, then a developer can buy the lot, tear the building down and sell the cleared lot or build something else on it. If demolition is very costly, then yes, the landlord might make negative profits on the building and seek to abandon the lot, which she wouldn't if there were no land tax. But this won't happen if the taxing authority takes demolition costs into account.

[zosima]

@Nick Rowe:

"when we are arguing about whether fiscal policy would be a good thing we have already entered that normative world"

Arguments don't work like that. They don't exist in a "normative world" or a "positive world", in general we base a larger normative case upon a collection of positive statements. They're always mixed. But if you want to see my claim as a positive statement, it is this: "Arguments that useless activities should not be counted towards economic activity are not part of the field of economics."

"There are two separate points here though: whether we economists see the re-filled hole as useful; and whether economic agents see the re-filled hole as useful."

If the government decides to do the policy, the Democratic system has already yielded its verdict on ditch digging: useful. It seems that you're trying to argue that government is a special case because its agency is not so clear as that of an individual or a corporation. But if an agent pays someone to do something, they think it is worth the price they paid for it. You may respond that you stipulated that the activity is useless, but then you've derived a contradiction, which implies that your question is ill-posed. Under the premises of economics, an agent cannot pay for a useless activity.

But you argued nearly opposite in your OP:

"But it's still a totally useless activity. And no reasonable measure of output would include digging a hole and filling it in again for no purpose whatsoever. Its market value is precisely zero."

We don't make judgments about what activities are useless and useful in our economic measures for a foundational reason: We can't actually measure utility. For example, The market value of a steak that I've just eaten is precisely zero, you don't know what benefit I accrued from eating it and you might consider my eating steaks useless on the whole, but we still count my eating steak dinners as economic activity. This has troubled philosophers of utility theory(see Utility Monster), but economists sidestepped this issue very early on by deciding that they could assume that willingness to pay was a reasonable proxy of actual utility. To reject this premise now is changing the rules of the game in a fundamental way.

[Nick Rowe]

zosima: "If the government decides to do the policy, the Democratic system has already yielded its verdict on ditch digging: useful."

Not so fast. The democratic system might decide to dig the hole, not because the hole itself is useful, but because paying unemployed workers to dig the hole might have some other consequences it considers beneficial. And what we are arguing about is whether there would be any such consequences. The democratic system might be misinformed about those consequences.

I could make exactly the same argument for an individual.

[K]

Determinant: " Land isn't inelastic, not in an realistic sense of the term."

I don't think it makes any difference to the argument. Tax any random proportion of the site value of every piece of land in Canada. It will have absolutely no impact on what and how much is produced on each piece of land and what is the rental value of that land.

"The postage-stamp lots that you and Nick have are irrelevant by themselves. What really matters is the size of your house, it's relative size and features compared to your neighbours and most importantly the fact that it is surrounded by a more or less desirable group of people to live among."

I think structure may be worth a bit more than land in typical residential real estate. But Land, in the Ricardian sense, of course includes also commercial land and the natural resources both over and under the surface. I think 1) that's already quite a lot of value; 2) the rental value of real estate will go up as income, consumption and property taxes are cut; and 3) as I (and Nick) have argued here before, Land is going to be an increasing fraction of capital as human labour is increasingly marginalized from production. Your point that the value of unimproved land derives from its surroundings (i.e. intrinsic natural wealth as well as improvements made and technologies created by the rest of the community/world) is exactly right. This is the positive externality, the value of which deserves to be taxed back to the community.

"We want to tax consumption because we want to tax the stream of revenue from stuff."

Unless it's an externality, I don't *want* to tax anything. But if it is an externality it *should* be taxed regardless of revenue requirements.

"My consumption tax will outprovide your Ricardian Land Tax any day."

Maybe. But that doesn't recommend it either from an efficiency or a moral perspective.

Nick: "Put a one-time unanticipated lump sum 99% tax on all forms of wealth now. Land, labour, capital. Same thing."

I'm not following. Seems to me that 99% land rent tax might be compared to a one off 99% tax on site value. But I don't see how it's a tax on non-land capital. But maybe I misunderstood your point. Also I don't see the time inconsistency. Every tax on commerce is equivalent to its PV as a tax on capital in some theoretical way. That doesn't make those taxes time inconsistent.

RSJ: "Now say that there is no tax on bond income, and no tax on structure-rent but 99% tax on land rent. Both owner-operators and landlords must pay the tax.

The landlord can sell the building and buy bonds instead. The market value of the building declines up until the after-tax rental yield, net of maintenance, is equal to the risk-free rate.

Landlords will take a *huge* capital loss -- but that is a one-time loss, not an on-going loss."

Declines? Whatever the value of the structure before the LVT, will be the value of the structure after: The lesser of its replacement cost and the PV of future total rent minus maintenance minus the site value of the land. The owner of the land/structure will indeed take a huge loss: 99% of the site value of the land. I'd agree that land owners ought to be compensated during the transition.

RSJ: "Other areas -- constrained areas -- have very expensive land, so in those areas, there would be many buildings with a negative present value, due to maintenance, that will be abandoned"

anon: "But this won't happen if the taxing authority takes demolition costs into account."

Here I fully agree with anon in the analysis, but I don't think I agree with the solution. If I rent your house and trash it, then should you lower the rent to reflect the decreased value of my accommodations? Leaving a negative-value structure is pollution. The owner of the land gets to keep the value of any improvements they make; they ought also to bear the burden of any damage. So the LVT ought to reflect the value of unimproved and *undamaged* land. In practice, though, I would think that the burden of land tax would serve as a disincentive against underused land.

Here (while I'm at it) is another benefit of LVT: Any public infrastructure project is automatically paid for by those who receive the benefit. I.e. you pay exactly the amount by which that improvement increased your site value. Conversely, land owners are automatically compensated by the amount of loss caused by any pollution (e.g. a new airport runway pointing at their house). Fairness everywhere you look.

[K]

K:"replacement cost"

Oops: It's got nothing to do with replacement cost. If replacement cost is lower, its because land value is higher.

[RSJ]

K,

By "building", I meant land + structure.

You think that you can make the investment costs of acquiring land different from the investment costs of acquiring the structure.

But as far as the entrepeneur is concerned, he is borrowing $100 to buy some land and build a factory.

The land + factory is his "capital", in the cost of capital sense, regardless of whether you want to treat land as not being capital in some philosophical sense.

He does not get to pay one rate on land and another rate on the structure. He can't go back to his lender and say "I spent part of the $100 to buy land, and since land isn't really capital according to K, I wont pay you interest on that part. "

Therefore to reduce the amount of interest income earned from owning land, you need to decrease the cost of buying land. You need to make the market value of land close to zero, by convincing people to use less land than they are using now. A tax will do that, it will lead to taller buildings, or more housing in undesirable areas, and fewer housing in desirable areas, etc. That is distortionary.

But with all those distortions, you will never get land to be free, so even if, after building the narrow, tall factory when a flatter factory would have been more efficient, the investor still ends up spending $10 to buy land, and $90 to build the structure, he must now earn a return of 1.1*bond rate, so that after paying the land tax on 10% of his investment, he can still pay the bond rate to his lenders. And that means lower output and income, unemployment, etc, as the bond rate, and not 1.1*bond rate, is the market clearing rate.

Your entire analysis hinges on the assumption that the landlord got the land for free, and therefore does not need to earn a return on the land.

In terms of abandoning buildings, yes, it happens. In NYC rent control was not adjusted for inflation; during the 1970s landlords set fire to their buildings. Ideally, you would stop maintaing the investment and led it depreciate away to zero, to at least get some money out of it, but there are building codes, etc. In terms of disposal costs, that is the beauty of bankruptcy and using a corporate structure -- e.g. limited liability for owners. Also in the 1970s, in San Francisco, as all over the country, there was flight from the urban core. 20% vacancy rates, with many abandoned buildings. The SFPD used to set fire to abandoned Victorians as part of their training exercises. Recently we've had cities bulldozing newly built homes that were abandoned. Take a look at Detroit, which is actually shrinking -- e.g. changing the city boundaries so that they don't need to deal with all the abandoned land. Nature is reclaiming it. You do not get efficient outcomes when you try to make a factor of production uneconomical, and land is an important factor of production, because location is important.

[Determinant]

K:

You're trying to argue from the "sin tax" or "excise" view of taxation. That is you should tax some externality, such as imported goods, to normalize or even penalize them compared with domestic goods. Implicit in this argument is that taxation should be in some way voluntary, i.e. you pay the tax if you engage in the [sinful] behaviour, likewise you can avoid taxation through righteous behaviour. You're trying to make the tax base broad enough so that people can't escape it. That's the weakness with any excise-based system in a dynamic monetary market economy.

Your system has a fatal weakness in that it doesn't tax land-light and value-heavy activities like software programming or most services like banking and lawyering. It's blind to the revenue of those business. Your system creates a bias toward those activities for no apparent reason. As such it's a distortion and your system will fail for that very reason.

It is also blind to market dynamics. Those Victorian English Aristocrats owned huge amounts of land and had vast rental income, or so they thought. It wasn't the land itself that was the valuable item, it was the agricultural surplus from those lands which actually paid the rent. It's a crucial distinction that becomes apparent when agricultural prices decline. Due to the repeal of the Corn Laws and the large amount of new agricultural land brought into use in North America at the time food prices steadily declined in England in this period. The aristocrats had the same amount of land but their rents declined as prices fell. They just had lots of worthless space instead of a store of value and as a result they went bankrupt.

Any system of taxation has to produce the required revenue. In a dynamic monetary market economy with n goods and 2n markets, goods for money and money for goods, we can never be sure which market will have a high value for taxation in any given period. It all depends on fickle consumer preferences, demand, or just the way the wind blows. In fact a good taxation system should be indifferent to market dynamics. We just want to tax the value in a way nobody can avoid. In a monetary economy every transaction involves money, either buying or selling. So instead of taxing x, y and z goods out of the n basket, where x, y, and z may or may not be of value, tax the monetary component of the transaction. Now our tax system is goods and market indifferent.

That means consumption and income taxes. A high-value economic activity may or may not involve land, but in a monetary economy it by definition involves lots of money. Therefore taxing money is the only way to produce the required level of revenue for a modern government. In a monetary economy all transactions feature only one common element: money. It's the one common thread.

This result has been apparent since the 1790's and the First Income Tax under William Pitt, but we keep having the same argument. The corollary to a monetary transaction taxation system is that you can't avoid taxation through "righteousness". Yes, the taxman will come after you and there's nothing you can do about it, he will get his money. That is the reality of taxation. Taxes aren't tolls. Again, we've spent 200 years trying to avoid or downplay this reality, but we keep coming back to it. Because we need the money.

[anon]

RSJ, I agree that the land is "capital" from the developer's perspective, in the sense that she has to buy it and then sell the land + structure to the final user. In the long run (ignoring transient effects) taxing land is little more than an accounting change: the developer sells the structure to the final user, and the user pays a ground rent instead of buying the land in a lump sum.

The developer also has to pay a ground rent while building the structure, and this ground rent payment is equivalent (on average) to the "interest" she would pay on the land. Overall, it is not the case that the interest from land is reduced: rather, it (or a portion of it) is transferred to the government as tax. And yes, this means that existing landowners immediately bear the full impact of the tax as a loss of asset value, which makes it quite contentious--but this issue has been addressed in previous comments: it's less of a problem if the tax is introduced gradually and other taxes are lowered to offset the impact.

Determinant, I dispute the notion that software programming, banking and lawyering are "land-light" activities. These activities are very much dependent on location: they mostly take place in urban areas, and high-profile locations for them (such as NYC or Silicon Valley) have very high land costs. Some entrepreneurs are trying to outsource them to less expensive places, such as rural areas or foreign countries, with limited success.

[Determinant]

Does a lawyer need as much space as a factory to produce the same revenue? Why does the insurance office that has more revenue and fewer expenses than the grocery store down the street pay less tax? It occupies 20% of the space of the grocery store but has 50% higher profit, after expenses. Ditto for two factories I worked at, one had five times the profit of the other due to their different markets, but both occupied similar amounts of land.

Under the Land Value Tax activities that produce high revenue but occupy little space get an unfair break and a bias toward them through the tax system. How are we supposed to tax the financial firm which trades on the TSX through computer terminals yet has millions in revenue?

Land is not a straight proxy for wealth or wealth creation. It is a raw input akin to wood. It's monetary transactions that are true value in our economy and that's what needs to be taxed. Many economic writers who lived in the UK or mainland Europe can be excused for thinking that land is wealth. It's a highly populated area with concentrated land ownership. Canada provides a useful contrast where land is much more abundant and location, community and services is what really drives its value. In contrast to the UK we have lots of undesirable, remote areas of little value to anyone. In Canada it is entirely possible to have lots of land but still be poor.

Land purchases or rent payments aren't an externality, they are simply the price for the consumption of a particular location, e.g. the right to occupy and enjoy it. You can make an argument that all rent payments and land sales should be subject to a consumption tax, e.g. HST and I would agree with you, but to single out land in the extreme way K suggested is folly. It will result in an unstable tax base.

[anon]

"Does a lawyer need as much space as a factory to produce the same revenue?"

When you adjust it for value, it's probably in the same ballpark. The land beneath a factory is generally worth much less than the land beneath a retail store or office building. Here's where you might be confused: when LVT proponents mention land, they understand the term to include the value of location, community and even services.

These things are so diffuse that the only feasible way of compensating their creation is through local governance and service provision, either through local government or private analogues such as homeowner associations. Certainly, there's no case for "compensating", say, an absentee landlord who owns an apartment block in a poor neighborhood. She's not creating much locational value or improving the local community in that neighborhood.

As for incentives, the LVT tends to replicate the same situation as "no taxes at all". Yes, a "land-light" financial firm might pay little land rent, and a retail store might pay more. But the same situation obtains today wrt. private landowners, for the very reason that land rent is "the price to occupy and enjoy" a piece of land.

Whether LVT yields enough revenue for a modern government, or whether that revenue is stable enough in practice, is a different question. There's nothing wrong per se with taxing income or consumption if necessary, but LVT seems to be a better match for local governments.

[Determinant]

Then you've just marched right into the Property Tax system as understood here in Ontario. The Property Tax is calculated on Market Value Assessment, which is a close enough approximation of "true" value as a government agency can obtain without an actual house sale at hand to refute the number. It is close enough to fair (20% off from actual sale price usually) that it can be taken as fair. But it's still less fair than a direct tax on income and consumption.

In reality, Ontario's system has the very problem I mention, lots of people who are property-rich but cash-poor, frequently older citizens who have occupied their homes for decades. These people have seen their assessed value rise and their income doesn't rise to match it. In Ontario if your assessed value increases more than the average for your municipality, you will pay more tax.

Empirically, modern levels of government can't survive on just a property value tax alone. Income and consumption taxes are the real cash cows. Property taxes are OK for municipal services that are retail-like and constant like police, firefighters and building inspectors. In Ontario we also fund Welfare out of property taxes and it makes for a rather dysfunctional system. It is generally agreed that paying welfare out of general provincial revenues (income tax) was more fair to everyone.

We have the very problem I predicted, that land-rich income-poor citizens get the raw end of the property tax system. Taxing these people means asking them to produce a payment from meagre resources. Their land holdings are not a true proxy for what we really want, money. It's actual income and consumption that is the true measure of wealth. We're back to the modern version of the Improvident English Aristocrats I mentioned.

There is no reason to use a poor proxy for wealth in the form of land when we can tax the real thing, money, directly instead. Government doesn't spend land, it spends money. We don't want land in, we want money in.

[anon]

"In reality, Ontario's system has the very problem I mention, lots of people who are property-rich but cash-poor, frequently older citizens who have occupied their homes for decades."

We're going somewhat offtopic here, but this is an interesting subject. Yes, "house poor" people can run into problems when their property values rise dramatically, or when their property rates are suddenly hiked. But the cure that's usually promoted (capping property taxes) is far, far worse than the disease. Look at what happened to California (US) after the passage of Prop 13 in June 1978. Not only were state and local government starved of revenue, but the property tax cap promoted a sudden rise in property values, which in turn fueled a destructive housing bubble. In addition, liquidity and flexibility in the local real estate market were severely impaired, because the prospect of higher tax assessment was a huge disincentive to market transactions. The end result is that in California, _everyone_ became "house poor" to varying degrees, especially migrants from surrounding states. The only people who derived some benefits from the measure were established property owners (both residential and corporate), a comparative minority.

And what a simple solution: If someone is "house rich, cashflow poor", just defer the taxation until the house is sold or inherited. In fact, a similar result could be obtained through a reverse mortgage, with no need for state involvement.

[Determinant]

*shudder*

Reverse mortgages, don't get me started.

Anyway, property tax as a tool has its limits, and Prop 13 has never had a counterpart in Ontario. But Prop 13 does show the limits of land value taxation. It demonstrates that it is income that is really being taxed. It's income that being paid out.

The bulk of government is borne by income and consumption taxes in Ontario. Property taxes pay for most municipal services (minus areas policed by the Ontario Provincial Police, paid for 70% by the province) and a part of education. School boards now get a per-pupil allowance from the province. The education property tax goes to the province, but that doesn't cover the education budget and never has, so income and consumption taxes (HST) make up the rest. Even when school boards could set their own rates most relied on provincial grants available when they exceeded a given property tax threshold. Only Toronto and Ottawa ever funded education purely out of property taxes.

The Ontario solution has been to raise property taxes to an acceptable threshold and let income and sales taxes make up the rest.

[K]

RSJ: I'm not sure where to begin. It's bizarre to me that you engage in an extensive debate about *Land* Value Tax and discover in the middle that you have no idea what it actually is. The information is generally available on the web and I have extensively pointed you to it and the underlying theory. And then, when you suddenly realize it means a tax on *Land* not improvements, and that you have completely missed the point, it doesn't even cause you a moments hesitation. Not the slightest urge to reread the above comments or *any* of the provided references. What, after all, are the odds that Adam Smith, David Ricardo, Henry George, or for that matter Paul Samuelson, Milton Friedman, Franco Modigliani, Robert Solow, Herbert Simon, James Buchanan, William Vickrey, James Tobin or Joseph Stiglitz would have had any thoughts on the matter, that couldn't be better improvised by you in a few minutes? Take Vickrey: "Economists [with the notable exception of RSJ] are almost unanimous in conceding that the land tax has no adverse side effects. ...Landowners ought to look at both sides of the coin. Applying a tax to land values also means removing other taxes. This would so improve the efficiency of a city that land values would go up more than the increase in taxes on land."

Apart from that, everything you wrote is utterly wrong.

[K]

Determinant:

"We have the very problem I predicted, that land-rich income-poor citizens get the raw end of the property tax system" = "The idle rich get the raw end of the property tax system." Also true for land taxes. I weep for them. Here's the Center For the Study of Economics, an organization that has been at the forefront of practical research into the implementation of land tax, on the effects of moving to land tax from other taxes (principally property):

-Net tax reductions for the vast majority of residents
-inaccurate assessments are reduced
-By reducing or eliminating the tax on improvements, there is a greater incentive to build, to build with higher quality materials, to maintain, to avoid blight, and to redevelop economically depressed areas.
-The damage that taxes like sales and income taxes do to working families and local commerce can be lessened.
-There is no tax shifting to citizens and property owners who have already done their bit.
-A tax on land has been shown to result in better land use patterns and more in-fill development. This has the benefit of reducing sprawl.

"Their land holdings are not a true proxy for what we really want, money. It's actual income and consumption that is the true measure of wealth."

You object to land tax because it is not a good proxy for wealth and you take it as a given that taxes should be proportional to wealth. I disagree. *To the extent possible* taxes should be equal to what someone *takes* from society, not what they give. E.g. if I own a previously marginal piece of land that suddenly becomes valuable because of a road or a suddenly valuable natural resource, do I have have the same claim to my wealth as a producer of greatly desired works of art? What did I give to the world in exchange for my wealth?

As far as lawyers and bankers and IP workers go, there is no question that there is a long list of natural and unnatural sources of monopoly rents that ought to be eliminated or taxed. And in the meanwhile, income and consumption taxes will have to suffice. I think Stiglitz get's it right:

The question is: "Would it be better if we had more taxation of land and natural resource, and more revenue from natural resource management, and I would include atmosphere and spectrum. And less tax on income and savings." And I would say, "Yeah." And I think many economists would agree with that. So, if you want to sell it as a "Single Tax," then, no, you won't get anyone to agree that there's enough revenue there. If you look at it more as a "central" tax, then, yes, you will get most economists [with the notable exception of Determinant :-)] to agree with you.

[RSJ]

K,

"And then, when you suddenly realize it means a tax on *Land* not improvements, and that you have completely missed the point, it doesn't even cause you a moments hesitation."

You obviously haven't read my comments at all if you can think this.

Land is not free. People must borrow at the bond rate in order to buy land just as they must borrow at the bond rate in order to buy capital. They must pay a return equal to the bond rate on the full amount borrowed.

If you impose a tax on the entrepeneur's land holdings, then his capital holdings must earn a return in excess of the bond rate in order to allow the the debt to be serviced.

That means less usage of capital.

You haven't responded to this point. It's a simple point.

This is not an issue with Ricardo or Smith, because they assume that land cannot be bought or sold, but is only rented. That doesn't apply to our world.

[Determinant]

Very true, RSJ. Adam Smith was a Scot and Scotland had a feudal land tenure system until 2000. Land was "subinfeudiated" which means you made the person you were selling to your vassal, instead of transferring title outright. The decades after Ricardo's and Smith's passing saw the development of chartered banking and the rise of mortgage-lending. In their era most people could not buy land or real estate. Today we can.

You object to land tax because it is not a good proxy for wealth and you take it as a given that taxes should be proportional to wealth. I disagree. *To the extent possible* taxes should be equal to what someone *takes* from society, not what they give. E.g. if I own a previously marginal piece of land that suddenly becomes valuable because of a road or a suddenly valuable natural resource, do I have have the same claim to my wealth as a producer of greatly desired works of art? What did I give to the world in exchange for my wealth?

Again, the "sin tax" view. Taxes can be optimized for revenue, or they can be optimized to induce a behaviour (a fine in disguise really) but they can't be optimized for both at the same time.

You keep dodging what a tax is really for. A tax is a way to raise revenue for government purposes, period, end of. In order to produce revenue, it must be targeted at revenue, that is wealth. If a tax can't target enough revenue it won't work. Either you have to broaden the tax base and compromise on your non-revenue goals, either way you'll have to compromise.

Your system is also extremely punitive of uncertainty. Can a person be certain their land get that road? Why should they be punished for something they didn't expect and didn't necessarily ask for? They paid for their house, why are they being taxed excessively?

The land-rich cash-poor are not the idle rich. I fundamentally disagree with your view on that point. Furthermore, reverse mortgages are terrible financial products. I have very little time for them. They certainly make people rich, just not the landowner.

At root it comes down to static vs. dynamic. I expect the economy to be dynamic and don't want to punish people for living in one and being affected by actions beyond their control. You assume that the economy should be static and people have perfect foresight, therefore if people benefit by luck they should get the raw end of the deal.

[K]

RSJ: "Land is not free."

If you commit to taxing 100% of the rents from something then that something will be worthless. If you tax 99% of the rents, it will be worth 1%. The point is to *make* it free. In a zero tax world, if I want to buy a $100 piece of land with a $100 factory on it, I borrow $200 from the bank and pay e.g. $10/year in interest. Under 100% land tax I purchase the property for $100 and pay $5/year to the bank and $5/year to the state. Same thing, except government earns the tax instead of the bank. Makes not an iota of difference as to how to optimally use the land. *No impact on output!*.

Determinant:

"Taxes can be optimized for revenue, or they can be optimized to induce a behaviour (a fine in disguise really) but they can't be optimized for both at the same time."

Optimizing revenue is most certainly not the goal. Optimizing output with the constraint of obtaining the required amount of revenues, maybe? And if the sinful behaviour that you are trying to curb is freeloading off the common good, then Pigovian taxes (including LVT), which also happen to optimize output, do in fact serve both purposes.

"You keep dodging what a tax is really for. A tax is a way to raise revenue for government purposes, period, end of."

First you want to optimize revenue. Now you say that as long as you get the required revenues, nothing else matters. Are you serious?

"In order to produce revenue, it must be targeted at revenue, that is wealth."

OK...? But that doesn't mean it has to be targeted at *all* wealth. It could be targeted at a particular asset, e.g. Land.

"If a tax can't target enough revenue it won't work. Either you have to broaden the tax base and compromise on your non-revenue goals, either way you'll have to compromise."

Yup. See Stiglitz quote above. We need to go for efficiency first. I.e. Land and other Pigovian, then consumption and income. As I said above, in the long run it should be a goal to eliminate or directly tax monopoly rents (including those that accrue to bankers, lawyers and IP workers). But in the meanwhile, we will have to settle for income taxes. Where did I claim that we have to make a total transition in one fell swoop?

"Can a person be certain their land get that road?"

No. But, if they don't get the road it wont yield any benefit to their land. Therefore neither the site value or the tax will go up.

"Why should they be punished for something they didn't expect and didn't necessarily ask for?"

Because the productivity of their land went up as a result of the actions of an external party. They received an economic benefit. In the case of land tax they are merely paying the exact economic value of an economic windfall. If they choose not to consume that windfall and fail to move aside in order to permit someone else to consume it, then, you're right, they lose. But with income taxes, people pay for services that aren't even intended for them: Didn't expect, didn't ask for, *and not even available to them*. Note that this is not intended as an argument against income tax; just pointing out that your objection to LVT applies even more so to any other tax.

"They paid for their house, why are they being taxed excessively"

Well, exactly! You convert your money into a house. Why should you suddenly pay a tax on that wealth? Improve the appearance of your house and land for the benefit of the whole community and they make you pay *even more* tax!!! You should just pay the market rent for the use of the land. No tax on houses.

Re reverse mortgages. I wasn't the one who made that point, and I've heard people get into trouble with them - though I'd tend to blame unscrupulous product design and sales. I'm all in favour of maximizing your consumption before you die, though. There ought be some kind of Pareto improving reverse mortgage design.

"You assume that the economy should be static and people have perfect foresight, therefore if people benefit by luck they should get the raw end of the deal."

I assume none of those things. I'm saying that if you get lucky and then it's all taken away, then you haven't been punished in life. You're just in the same situation as everyone else. And if you never expected to keep your windfall in the first place (due to e.g. land/inheritance tax), you wont even feel bad. And when it's a tax on blind luck, it's 100% efficient since you are not discouraging any activity.

[K]

"The land-rich cash-poor are not the idle rich."

Sure they are. They just chose to misallocate their wealth into a single asset. If they'd live somewhere they can afford, they'd have cash and the land they were previously monopolizing would likely be used more optimally by someone else.

[RSJ]

K,

There *is* literature on this, that I've known for about 10 years now. This is my bailey wick, actually.

1. No one knows the NPV of a house, so no one knows what the right LVT should be. But the wrong "V" will cause the tax to be massively distortionary.
2. Everyone is credit constrained, so you will have the house rich/land poor problem.
3. No one is currently making the best use of their property, so LVT will be distortionary
4. Land is not supplied inelastically. Most land is not supplied at all, and land is only brought online gradually.
5. The theory only works if there are no other obstacles, such as building codes, that prevent present owners from maximizing the value of the property.
6. The theory only works if there is an exogenously determined technique that is fixed for each type of land.

Which is not to say that there aren't land rents.

There *are* land rents, and we need to deal with them.

But levying a high tax on an illiquid asset whose value is indeterminate is a bad idea.

[anon]

For the record, I agree that reverse mortgages tend to be bad products in practice, since they are inherently hard to understand and often overpriced. My mention of them was only meant to argue for the economic soundness of tax deferment as a way of partially addressing the "house poor" problem: obviously, the amounts and interest rates involved would be vastly different.

Much of the recent discussion seems to touch on normative issues, although this is perhaps inevitable when arguing about tax policy. I for one have always maintained that the tax on land is an inherently contentious proposal, because the potential drawbacks are immediate and salient (fall in land values, "house poor" folks being forced to sell) while the benefits are less salient and extend into the long-run future. I'll let K reply to RSJ's points above, although to be frank, I am quite skeptical about them.

[vimothy]

Sorry to interject, but, does the supply of land really change with prices? Can someone point me at some evidence/data/whatvs? Cheers.

[Jacques René Giguère]

"The land-rich cash-poor are not the idle rich."
K:
"Sure they are. They just chose to misallocate their wealth into a single asset. If they'd live somewhere they can afford, they'd have cash and the land they were previously monopolizing would likely be used more optimally by someone else."

Choosing how to use your house is not a misallocation. It is an allocation.
A good old developper's trick is to buy a small piece of land at a high price. Then your cronies on the city council raise the taxes on the other owners. The price per sq. footage may be high but the total not enough for you to move to a really better place. So your ancestral land is bought from under you, you relocate to some miserable place and the "capital is better allocated" to another use, such as a golf resort. That's how the Gullah

http://en.wikipedia.org/wiki/Gullah

are currently being cleared from the Coastal Carolinas islands...

[RSJ]

"does the supply of land really change with prices?"

Land is differentiated. Marginal land always has a price near zero -- not really zero, as it has option value in the expectation that it will be developed in the future. In a static economy with no development, it would have a value of zero.

But the margins themselves grow as more people move into an area, or as they are able to spend more on housing.

It's better to say that when the price of land changes, the supply of *other* land increases. Obviously an increase in the price of city-center land is not going to create more city center land. Neither will a tax on city center land reduce the amount of city center land supplied. But arbitrage by developers may cause the city center land to be used less intensively and for land on the outskirts to be brought into use, if the proportion of the land value is lower the farther you go out.

The empirical data is the growth of urban centers.

Imagine fisherman living near the coast. They all work on the coast. Our city is a half-line, with the coast at the origin. The utility from renting a house at position x = log(housing consumed) - x. The marginal disutility of commuting does not lessen as you commute more :)

Everyone wants to be at the origin, so developers can build tall buildings there that house many people. But there are diminishing returns (increasing construction inputs per unit of housing service as the building gets taller).

Say, at each location x, the housing services delivered, as a function of structural inputs = sqrt(c). So it would be inefficient for everyone to be stacked up in a 1 mile tall building at the origin, and nothing farther out.

Instead, they will spread themselves out in a curve, with more density next to the coast, and less density as you go out. At the last house, land is free, and all the value is in the structure. Beyond that, no one is willing to live that far out.

This is a fun calculus of variations problem -- how to maximize (total utility - total construction inputs).

As you move farther out, you get less of the benefit of living near the coast -- or equivalently you face a higher cost of commuting -- and to compensate you will pay less in rent per unit of housing service delivered. This is what incentivizes the landlord to build taller buildings near the oasis, even though those buildings are less efficient, just in terms of housing service delivered/structural input.

Now if you levy a tax on the land value, then landlords can lessen the burden of this tax by investing more in land farther out and less in land closer to the oasis. This is because the proportion of land value declines as you move away from the oasis -- finally being zero at the last house, which pays no LVT at all.

Now the distributed market solution to the problem is no longer the social planner's solution, and you are not maximizing total benefit - total cost. The land value tax, in this model, has a dead weight loss.

Paradoxically, this skewed investment pattern is more efficient, from a physical transformational perspective. Taller buildings are inefficient, but they are built in order to access a location-specific benefit. That benefit is captured in land prices -- you are paying for the locational benefit when you buy the land. As the value of extracting that location benefit declines, landlords are encouraged to not attempt to extract the benefit, but to build more efficiently, from a structural input perspective. But less efficiently from a utility maximization perspective -- or social welfare perspective.

Anyways, this is a dead thread, so if the above is not clear, then I can write out the equations. I think the argument -- as well as the assumptions on which it relies -- should be clear. It's important to take into account geometry and spatial relationships when trying to model real estate.

[POWinCA]

Most people would rather have a one-time transfer of $100 than a job digging holes and filling them up for $100. But some will consider the longer-term.

At the end of period, the person who dug a hole can tell a prospective employer he "worked" for the period while the person receiving the transfer only consumed leisure. If the prospective employer is a construction contractor, that "useless" work digging holes might be worthwhile experience and signaling, EVEN IF the hole digging project had a negative return. The prospective employer cares only about the productivity of the worker, not the poor decision-making skills of the previous employer.

However, if the transfer or "work" (useless or not) is only temporary, it will only temporarily fill the output gap. We're assuming that the decline in output was due to some exogenous shock that will correct itself over the temporary period of work/transfers. But recessions hardly have exogenous causes and hardly have definitive durations.

If the length of the recession was fixed in time, I don't think there is any argument that we would borrow from the future to bridge the gap. But the uncertainty of the time to return to full employment matters greatly because of crowding out effects and structural changes. Whenever we "employ" resources to do something useful, useless, or nothing at all, we deprive the economy (and the worker) of the opportunity to adapt to the changing dynamics of the economy. There are parts of the economy where we underinvested for many years.

During the housing bubble, we commanded real and financial resources to pour into production of durable goods: house and their complements. We built up an enormous "resource debt". Because they are durable, we literally stole resources from the future for present consumption. Now the resource debt is being paid back with unemployment. Furthermore, investment in residential and commercial real estate produced little innovation to contribute to sustained economic growth; we became marginally better at building houses and buildings, but learned little else.

People who call themselves "carpenters" and "plumbers" "electricians" and "iron workers" are no longer needed for a very long time. We drew too many people into these job classifications and devoted too many resources into building non-transferable human capital. We deprived other industries of these investments for a long time, and that's where the unemployed resources and available capital should flow!

Public works projects might put these people into jobs as carpenters, plumbers, electricians, and iron workers, but that only compounds the resource debt and hinders the market-desired accumulation of human capital. We will still have TOO MANY of these laborers, pay them too high wages, and eliminate their opportunity and incentive to build new human capital!

So temporary expenditures (useful, useless, or transfers) are completely counterproductive in every respect except for maintaining minimally adequate basic needs. But provisions for those basic needs should have been made by THE WORKERS THEMSELVES during the boom. They should have had fully-funded private unemployment insurance with premiums commensurate to the replacement wages they desired.

Government should not be funding this because revenues fall at precisely the time when social services are needed the most - revenues are volatile, but expenditures are sticky downward. If debt were zero going into the recession, then government could afford to smooth the cycle. But it wasn't, so it cannot. With volatile revenues, public expenditures must grow at a SLOWER rate than revenue growth to maintain a near-balanced budget during downturns. This is more important for state and local government than the federal government.