Monday, October 13, 2008

Death to Bond Measures

Okay. Just a parochial political rant today. And it's not even about the Presidency or anything the Network News drones on about 24-7.

I'm sick of bond measures. They're phenomenally stupid and an insult to Americans' intelligence (or an exploitation of their stupidity, take your pick). Oh, right. Some of you might not know what I'm talking about. I don't know how it works in other states. Here in California, invariably every election is choked with bond measures. Bond measures for everything. Boring, stupid bond measures.


It's the way it works here, that anyone--just some guy off the street, or not even a California citizen, could be some Afghan warlord, Gort from Planet 9, or even the California legislature (no kidding!)--can get an "initiative" on an election ballot with enough signatures of registered California voters. Dunno how many, but I seem to recall it's less than 100,000. I could look it up if I thought the five minutes would be worth the bother, but I don't, since it's not relevant to my point today, as I'm not against the idea--even Taliban insurgents in the Eureka state can put up whatever initiatives they want, ask for the institution of Sharia law, the beheading of Britney Spears, don't care. Since we all have to vote on it in the end anyway, so we get what we deserve, whatever the fiasco may be.

But I digress. "Initiatives" can be a change to the state constitution (there's one on this November's ballot to "save" hetero marriage from all those wicked homos daring to topple Western civilization under their perverted avalanche of wedding rings and successful careers) or a...yes, you know it...bond measure. I don't know the exact technical reason. But since our state is now constitutionally required to pass a balanced budget every year (imagine that...gee, when will those whiny Republicans in Congress start emulating this godless state of liberal perverts and actresses, I do wonder?...we even require our government to save money for future rainy days, such liberal freaks we are), the only way any money can be borrowed to do anything (outside an actual, serious emergency--which I mean in English, not Washingtonese, where "serious emergency" can mean something as lame as we can't afford stamps to mail our staff their checks this year unless we borrow a gagillion dollars to meet our tobacco price-fixing buyback obligations) is if the citizens directly vote to temporarily override the constitution (which they also voted for, by a massive landslide, as if they forgot) with a "bond measure," a special exception to the balanced budget rule whereby money can be borrowed and our state can be sunk back into debt, but hey, as long as the money is only spent in a very specific way (whatever stupid way the bond measure specifies).

Now every California ballot is littered in bond measures. There are at least three this November (not counting local municipal bond measures, of which I must endure at least one this year). Now, if 71% of voters already voted to compel our budgets to be balanced, how on earth could any bond measure pass? Wouldn't the same 71% say "Dude, we told you NO...how many f*cking times do we have to repeat ourselves?"? Or are Californians so stupid they don't even realize the inconsistency of passing a balanced budget amendment one year and then a bond measure the next? But never mind that. I'm sure at least a third of the human race is addicted to hypocrisy. It's a chemical in the water or something. That's not my complaint right now.

My complaint right now is stupidity. Either the abject stupidity of California voters who actually fall for this sh*t, or the stupidity of bond measure proponents who actually think their bond measure is a good idea. Because it has to be one or the other (or, dog forbid, both). Here's why...

A bond measure is nothing more than asking the California taxpayers for permission to go into debt, so something (something supposedly important, like giving candycanes to cancer kids or something) can be done. But if that "something" is so important, why can't we just move the existing budget around or raise taxes to get it done? "No one will vote for that," you might say. But voting for a bond measure already does that. Bond measures basically conceal the actual costs from the voter, so the voter thinks he's getting something for free, when in actual fact he's voting to pay twice as much. Yes, bond measures cost the taxpayer twice as much as just buying the damn thing outright. Because instead of just paying for it, we then have to pay the exact same amount for it in interest, which works out to a whole other bill for the same amount on top of the first one.

That money has to come from somewhere. It won't fall like manna from the sky. So where will it come from? To pay back the bond and the interest on the bond, we still either have to move the existing budget around or raise taxes. So why not just do that in the first place, and thus save half our money doing it? If you can't think of an answer, then you just proved you have an IQ anywhere above dumb.

For example (you know, in case anyone's still confused): California Proposition 1 (which I guess is now 1A, don't ask) asks voters to approve the issuance of $10 billion in bonds to build a spiffy new supertrain up and down half the state. Yes, they try to coax you with red herring bullsh*t like "It'll bring jobs!" (so would building a ten-mile high replica of the Moon from solid jack cheese...just think of the fantastic boost to the California dairy industry!). But what they're really asking you is to pay $20 billion for a $10 billion train. No kidding. The measure even says that, right in the pamphlet summary, you don't have to go digging at all: the cost of getting this $10 billion is $20 billion. Wait. What? Wouldn't just spending $10 billion be cheaper? Uhhh...10 billion dollars cheaper in fact?

"Well, it'll be spread out over thirty years." So? We still have to come up with the money. So are we raising taxes, or cutting police from the beat? For thirty straight years, no less. Remember, we have to come up with the dough every year. Is it supposed to come from finding pennies under the rug or something? As the measure says, the cost per annum will be $647 million. So, are we raising taxes by $647 million or closing $647 million worth of schools? Or what? That's not a facetious question. You'd better have an answer.

Now, think for, like, two seconds. A supertrain along the length of California linking a dozen cities. Hmmm. How long, honestly, do you think that will take to build? Fifteen years would not be an unreasonable estimate. It's supposed to connect at least four major metro areas in the state (LA, Silicon Valley, The Bay Area, and the Capitol, i.e. Sacramento, as if there was any important industry there), and eventually twice that (hitting almost every major city from San Diego to Fresno and in between and beyond). I seriously doubt they can complete even one of those linkages, from buying the land to having it operational, in less than three years. All four links would take almost ten years, and the whole line surely at least fifteen years. So why not just make it fifteen years and spend $647 million a year on each annual phase of the project (the same amount they already plan to spend on the bonds)? Then the total cost would be half (only ten billion instead of twenty). Yeah, the annual cost to the taxpayer would remain exactly the same, but that's actually the point: exactly the same annual cost as the bond will burden us with, yet
it would be paid off twice as fast (fifteen years instead of thirty).

It's not even a question of what's "more palatable" to the voter, since "tricking" the voter into spending $20 billion on a $10 billion train is always an outrage. Why not just be honest and tell the voter the train is still going to cost us $647 million a year either way, bond or not. And then pitch the fact that just voting that money right out of the budget (or from some new tax) will cost them half as much as the losses they'll eat if it's done with bonds instead? Really, who's the bigger idiot here? The bond measure proponents or the voters who fall for it?

Now I'll be honest, I'm against all government debt. It should only ever be resorted to in emergencies. And again, I mean actual emergencies. Like, when you have no choice, because there is nothing left to tax, or not enough time to collect it. You know, like if we suddenly find out an asteroid is going to vaporize the West coast in three months and we need to build a gigantic raygun or something before the fiscal tax year is in. But until that happens, they can suck it. So I always vote against bond measures. I don't care if they fund a time travelling ninja squad that will prevent the Holocaust (or the 2000 election of George Bush). I don't care if they buy shoes for orphans. I don't care if they give spontaneous orgasms to all law abiding citizens. No more debt. If it's really so important, f*cking pay for it. Otherwise, I have a simple rule: if they can't afford their pet projects, they don't get to have them. Period.

But even if you're not a fiscal hardliner like me, surely you at least aren't a drooling idiot. So why, why oh why, why in heaven and earth and all the lands of fey, would you approve a measure that will force us to pay twice as much for the same result? If you can't give any valid answer to that (and especially if you can't even conceive of an intelligible attempt at an answer), then please do not vote for any bond measure. Ever. No matter what it's for. Don't even read it. Because their expecting you to even so much as read such measures only insults your intelligence. And if you let them do that even one more time, it may be arguable that you don't have any to insult.

End rant.

36 comments:

Unknown said...

Spontaneous orgasms for everybody? How many? Is it just a one-off, or do we get regularly scheduled recurring spontaneous orgasms? That is the best f*cking bond measure ever! I'll sign that petition.

Ben said...

But what if there was a bond measure to fund the time traveling ninja squad to go back in time to prevent the idea of "bond measures" from ever coming about in the first place? You really don't think these things through, do you Rick? :p

Richard Carrier said...

Oh, but Keith, we can have all those orgasms for half the cost, without the bond measure! Hence surely we should just put it in the budget. :-)

And War_on_Error, since it's logically impossible for any society not to think of bond measures eventually, the ninja squad will spontaneously turn into a lemon maringue pie in order to prevent a paradox violation. That's just wasted money. I mean, why not just buy the pie outright? Surely that's cheaper.

Grégoire said...

Gee Richard, I couldn't disagree more with anything you've written.

Initiatives in American states like California and Washington are the envy of the rest of the bureaucracy. Most of the progressive, interesting, useful stuff is far too important to be left to the politicians.

macroman said...

I think you may have overlooked the issue of inter-generational fairness. I don’t know enough about the Californian super-train, but I can use the Sydney Harbor Bridge as an example. I think the interest may still be running on some bonds issued to build it (maybe the people who bought the bonds were ripped off but that is another matter). And it now costs $3 to use the bridge. If the taxpayers of 1923-32, (the period of construction) had to pay for the bridge up-front would that be practical or fair? Hardly practical because those taxpayers couldn’t afford it without massive privations – it was borrow or have no bridge. Businesses borrow to buy factories they expect to pay for out of future earnings. Is it automatic that taxpayers shouldn’t do something similar?

What about fair? Some of those taxpayers would have died before they ever got to use the bridge. They were poorer than we are now – should they suffer to leave a free gift to future generations, who took advantage of the bridge to create private wealth (access to more jobs, better division labor in the Sydney area)? Is it unreasonable that all those people who use the bridge pay for it, not just those who built it?

Thogek said...

Wow, educational *and* entertaining. Excuse me while I wait for my spleen to uncramp...

Very well put, all good and valid points. Responsible families try very hard to live within their means and budgets and avoid going much into debt, because they know what can happen to them living on too much overextended debt. Why our government (or its voters) seem completely immune to that simple practical realization escapes me every election cycle.

California has four bond measures this time around, all for arguably great causes. (Who wouldn't want to fund super-trains or children's hospitals or renewable energy or veterans?) But those causes aren't going to benefit from a state government that locks up and implodes under its own growing debt.

Stop treating debt like free money. It's not. Didn't anyone notice that that's how the mortgage industry melted down?

To the other points...

Yes, state initiatives give voters the ability to input directly into the lawmaking process, and that can be a great thing -- if the voters engage a little time and a few braincells and understand the implications of what they're voting on, which most of 'em don't, as they're quite happy with the blissful mindlessness of Kool-Aid-chugging instead.

And there's nothing fair, inter-generational or otherwise, about going into debt to pay double-value for anything. Do you take out huge long-term loans for big family projects with the intent of sticking your kids with paying off the latter half of it when you're gone? Either we have the budget to do it, or we don't. If we don't, then put it on the wish-list shelf (or rearrange the budget) until we do.

macroman said...

“Do you take out huge long-term loans for [a family house] with the intent of [leaving it to your kids even if there is small chance they may have to sell it to pay it out] when you're gone?”
Yes, many of us would do exactly that, wouldn’t we?

“Either we have the budget [for a house], or we don't. If we don't, then put [a house] on the wish-list shelf (or rearrange the budget).”
Many apparently reasonable people prefer to pay interest on a home loan rather than pay rent, even though the sum of the purchase price plus all the future interest payments may look like a huge sum, particularly when you confuse the future payments with their present value.

Thogek said...

Don't distort my comment. I didn't say "a family house", I said "a big family project". A family house -- i.e., covering the basic need of a place to live -- is one thing, but adding on other non-vital projects that extend an already-significant (and arguably towering) debt is quite another.

The issue is not only the added interest -- the fee for receiving the money up front while paying it back gradually over time -- but the regular monthly/yearly amounts that these repayments necessarily lock into your periodic budget for the next thirty years. If your current budget and spending do not have ready room for those payments (and you have any sense whatsoever), do not incur them. Otherwise, the result is going to be a serious tightening of the belt elsewhere, or somehow increasing your income -- which, in the government's case, generally means raising taxes.

California's continuing budget deficits and growing state debt should be a pretty strong signal that its budget (which is legally required to be balanced annually) does not have room to lock in more and more debt repayments without something seriously giving somewhere...

Richard Carrier said...

Grégoire: I can't tell if you're joking, but FYI I have no beef with initiatives, just bond measures. Hence, e.g., an initiative to raise taxes to build a supertrain would be fine by me. I might even vote for it.

Macroman said... If the taxpayers of 1923-32, (the period of construction) had to pay for the bridge up-front would that be practical or fair?

Yes, it would.

(1) As you say, people still pay to use the bridge. That money will always be needed to maintain the bridge. That's fair.

(2) We all buy things for future generations. We don't make them pay for it. Indeed, it's unjust to make people pay for things they didn't buy and might not even want. Maybe you have a weird ethic over there, but Americans would be ashamed if they left their children with their own debts, and for that very reason we all strive as a matter of moral common sense to pay off our cars and homes before we die. We don't think it unfair that our children then benefit from our work--to the contrary, we think it a blessing.

(3) Maybe this is also an American ethic (albeit one forgotten by our aristocracy), but ever since Ben Franklin said it, we've kept teaching it: If you can't afford it, you should find another way to get what you want. Or just give up your excessive ambitions in the first place.

(4) Eventually your bridge will be paid off, and since the number of potential people who will get to use that bridge for free is far larger than those who have already used it, your assumption of unfairness is unavoidable: most people in history will get a free bridge, so why does it matter how many generations pay for it? Is two generations fairer than one? Why? Especially since the second generation won't have even had a say in whether they want the thing.

(5) It makes no sense to say "it was borrow or have no bridge." If the bridge were genuinely that expensive, it should not have been built at all. If you can't afford it, you shouldn't get to have it. But I would check your facts. Just how much did the bridge cost, relative to regional corporate profits at the time or the potential sales and income tax increases over the previous five years? I'll bet a good analysis will show the bridge could have been built at no debt and half the cost without anything any reasonable person would call "massive privations."

(6) Businesses borrow against future earnings because they get future earnings--unlike taxation, they don't compel customers to buy their products. They actually have to make money by earning it. So if you are proposing all bond measures should be tied to a long-term net profit-making venture, then I wouldn't oppose such a measure. In fact, I think that would be a good idea. But that's not what we're talking about. For example, the proposed supertrain won't even pay for its own maintenance and operation with projected fares--it will be eternally dependent on tax subsidy. So it's not a for-profit venture. Why not? Because if anyone charged fares that would turn a profit (especially enough to pay down loans to build it), they calculated no one would use the train, so it would fail as a business. Which, incidentally, is why no private business will ever attempt it. They know better. So should we.

(7) In the same fashion, if any bridge can actually "turn a profit" (exceeding not just the cost of its construction but also its perpetual maintenance), then maybe such a bond effort would be justifiable, but in all the actual cases I've seen over here, they've never succeeded at creating such a measure. All our bridges, even the ones we pay to use, can't even be maintained on their own tolls. They eat taxes just to stay up. Forever. Which is fine, if we want the bridge that badly. But why would we want to double those taxes? If we just paid for the bridge, the taxes to keep it can be halved. Which is "fairer"? To saddle our children with twice the tax, or half the tax? (Indeed, you might want to check to see whether your Sydney Harbor Bridge is actually being paid for with the $3 tolls--that sounds low to me, given what it actually costs to maintain bridges--never mind also pay down loans to build them).

(8) Which is where government differs from business: if a business can raise capital directly to buy a factory, every business will prefer that to a loan, because they can then build the factory at half the cost, which entails double the profits. Businesses only take out loans when they can't find sufficient investors. But government has a different kind of investor base--everyone. They can tax all sales, all businesses, all people, whatever they want, and thus raise any capital they want. If they can convince voters to give them the go ahead. So if a bridge will benefit everyone (or most) then you will have the People as your investors. Convince them to invest, and you can pay for what you want (bridge, train, whatever) for half as much. And yes, asking those people to invest in benefits for future generations is perfectly valid, and often a winning argument.

All I'm saying is: which makes more sense, asking them (and their children) to pay for the bridge, or asking those same people to pay twice as much for the same bridge?

macroman said...

Richard,

Thanks for the long reply.

Before discussing it further, I need to clear up something. The crucial issue seems to be implied in your final paragraph, which tells us succinctly all you are saying, so we can concentrate on that:

“All I'm saying is: which makes more sense, asking them … to pay for the bridge, or asking those same people to pay twice as much for the same bridge”

(I assume we are talking here about a hypothetical bridge which is useful, not about the merits of a particular government project, nor the merits of any borrowing rules which may be forced on the government by its constitution).

It seems from your question that you don’t agree with the theory of interest and the “present value” of future money.

I assume you know that the theory of interest (“time preference”) says that the value of a promise to pay $1 in 10 years time is less than the value of paying $1 now (even assuming no inflation and negligible risk of default). Hence, to pay back an equal value for the $1 I get today I have to pay back $(1+x) in 10 years. In other words, the future payment of $(1+x) has a present value (discounted value) of $1 (now).

So assuming a rate of interest (which reflects the average “time preference” of market participants) and assuming standard accounting calculations, it can make exactly equal sense to ask citizens to pay $y for the bridge now or $(y+interest) spread over 30 years. Maybe that is where our disagreement lies – I am assuming that this theory is correct, while you reject it?

macroman said...

Richard Carrier said “…please do not vote for any bond measure. Ever. No matter what it's for. Don't even read it.”
Let’s look at a case where the citizens of California ignored such advice.
“On November 4, 1930, voters within [San Francisco, Marin, Sonoma, Del Norte and parts of Medocino and Napa counties] went to the polls to support a $35 million bond issue to finance the building of the [Golden Gate] Bridge …145,697 people voted for the bond issue; 47,005 voted against it. The Bridge was completed … on May 27, 1937 … The last of the construction bonds were retired in 1971, with $35 million in principal and nearly $39 million in interest being financed entirely from Bridge tolls.Building the Golden Gate Bridge
The bridge cost $35m+$39m but the bridge may well have generated more than $35m+$39m in benefits (increased productivity in the Bay area through worker mobility and better division of labor, savings in fuel and travel time, the effect of drawing economic resources into the Bay area). That is why 75% of voters in 1930 probably thought they were doing future generations a favor. For later generations to complain that they (the later generations) could have been spared $39m (in tolls), if only their parents or grandparents had cut consumption in 1930-7, seems incredibly ungrateful, especially since consumption (wealth) in 1930 was about half the wealth of the 1960s generation that was stuck with paying for part of the bridge (Ratio of inflation adjusted US GDP figures for 1955 and 1965 is 1654/3214 = 51%. Data available from the Bureau of Economic Analysis). The point about interest payments, and the “present value” of money, is that the 1930s generation had other uses for the capital cost of $30m and those other uses also created wealth to be bequeathed to the next generation.

The rational response to a bond proposal is: read the cost-benefit analysis. It should take into account the estimated future wealth generated by the project. That wealth may include such things as an improvement in air quality. As part of a risk assessment, check the annual interest payments (on past bonds) by the state of California compared to the State’s annual revenue (it is about 4% in 2008). Allow for a certain portion of those payments to cease each year as the bonds expire. Consider how many other new bonds issues are proposed at the same time as this one. Allow for modest growth in tax revenue (at fixed tax rates). Allow for tolls or charges that might return to the state from this project. To test your personal “time preference” or “implied natural interest rate”, you might consider whether those who bought the $35m worth of Golden Gate Bridge bonds in 1930-7 were getting a good deal when they received $35m+$39m over 34 years, or whether it was the citizens of the Bay area who got a good deal. Then make an informed decision.

To forestall a possible side issue. The bridge was not conceived as a response to the depression. Planning began in 1921 and the design was selected on Aug 15, 1929. Of course, “for some the timing of the bond election was considered economically reckless as it would create bonded indebtedness during the Great Depression”. What are the chances that many of those who voted against the proposal would have thought creating bonded indebtedness during good times was also economically reckless?

To forestall another side issue: I am not talking about the merits of paying tolls rather than taxes, which is an argument about which members of the next generation should pay for the construction of the bridge. I find it convenient that the Interstate freeways in the USA are not toll roads but I understand the case for tolls as opposed to taxes; full cost recovery tolls on Interstate 5 might make the Supertrain look like a better proposition (I have no special information about the merits of the Supertrain). The point is that some of the next generation paid for some of the capital cost of the Golden Gate Bridge, and I am not convinced that was a bad, let alone un-American, state of affairs.

Richard Carrier said...

Macroman said... It seems from your question that you don’t agree with the theory of interest and the “present value” of future money.

No, I do, but only in real terms. That is, if there is an actual future return in money, and you can't get it any other way. This could be achievable with a bridge via a loan if somehow the bridge increases trade and thus tax revenue and thus makes more money than the loan--assuming the bridge could not have been built without the loan. But those are two big "ifs."

Like the Supertrain here in California: will that increase trade and thus revenue by more than we spend to build and maintain the train? Everyone agrees the answer is no. Do we need the loan to build the train? No. We can raise the revenue directly (in fact, the taxation expense to the public is halved if we skip the loan). Thus, no appeals to "present value of future money" can have any traction here. There just is no rational reason to build this thing on loans.

You need to look at the real world facts: on the loans, we pay $800 billion for 30 years, without them we pay $800 billion for 15 years, and yet on either option we end up with exactly the same train. You can't claim there is any sense in that, not even by appealing to "present value for future money."

In contrast, for example, you do make a good case for the Golden Gate bridge being profitable. However, that was the Depression. Ten years earlier and the revenue would have been available outright and no loans would have been needed. In fact, I'm not sure it wasn't available even during the Depression.

To finish the bridge, by your own numbers, it would only have cost $5 million a year, well within normal means outside a Depression, and even within Depression-era means: it would have cost each citizen of the city just $8 a year, equivalent in current spending power to $18 a month, perhaps steep during the Depression, but that's assuming you divided the cost equally rather than proportioned by wealth (as would make more obvious sense; even if you taxed only the employed, it would have been only $10 a year, the equivalent of $22 today), and also assuming you paid for it entirely by income tax rather than including a sales tax rate increase. In fact, if each citizen of the time spent just two more cents a day in sales tax (total, i.e. spread across all purchases made), they would have generated the required annual amount (indeed, how much were the people already spending per capita in those same years to retire the loans?).

If they had done that, we would have had the exact same bridge at half the cost and twice the profits.

The only possible justification, then, for doing it on the loans is that (for some reason, maybe sensible in the Depression or a World War, but unfathomable otherwise) the economy could only bear a certain expense limit per annum on the bridge (i.e. there just was no way to raise $5 million a year without doing harm), and that limit was below what it would cost per annum to build the bridge ($5 million per year), therefore the actual limit per annum (say, $2 million per year--just picking a number, you didn't say what the annual loan payments were for the bridge in the 1930's) had to be spent on loans so the cost could be spread over more years than it would take to build the bridge (typically in California bond measures are 10, 20, or 30 year obligations). In that extremely unusual circumstance, it makes sense to consider present value of future money and do the math before deciding. But in no other case is that ever so.

Now, of course, after I wrote, the economy tanked, and we may be looking at another Depression (although we'll hopefully be ending a very costly war, which might balance each other out). So maybe now we'll have some viable arguments for debt. But again, that's an extraordinary circumstance. Outside of that, the arguments just don't add up.

macroman said...

Richard Carrier said:


“You need to look at the real world facts: on the loans [of $12b] we pay $800[m] per year for 30 years, without them [the loans] we pay $800[m] for 15 years during construction, and yet on either option we end with exactly the same train. You can’t claim there is any sense in that, not even by appealing to ‘present value for future money’”


No one could favor the first option over the second, but that schedule of repayments presupposes that the entire $12b is borrowed on the first day of construction. In fact, the State will borrow $800m each year during the 15 years of construction, so the money is available when needed, but not before. This is standard practice. Pages 35-36 of the State of California, 2008 Debt Affordability Report , prepared by the State Treasurer, lists bonds authorized by 44 different Acts which have not yet been fully issued (sold to the public). Examples are: $2.2m un-issued out of $800m authorized, School Facilities Bond Act of 1988; $2.16b un-issued out of $3.44b, Coastal and Beach Protection Act of 2002; $6.98b un-issued out of $7.32b, Facilities Bond Act of 2006.

Thus, in year one of construction, the state will sell or "issue" the first $800m worth of bonds. In year two the State will pay the first installment of $53.3m on that first $800m (assuming a 30 year term and rate of interest of5.215%). In year two the State will also sell the second parcel of $800m worth of bonds, and so on. In year 16, after $12b worth of bonds have been sold, the ayments to bond-holders reach $800m each year, and stay that way for 16 years. The payments then decrease by $50.3m each year, until the last of the bonds is paid off in year 46.

The State will pay a total of $24b over 45 years, whereas by nor borrowing it could pay $12b over 15 years. If you think there is no difference between a dollar in the hand now and a dollar in the uture, you will think a total payout of $12b is obviously better than a total payout of $24b. But if you think that the value of having or paying a dollar now may be different from the value of having or paying a dollar in the future you have to consider the present/future value of money.

But first we have to clear away a red herring. Whether the train is a good or bad investment makes no difference to what is the best way to pay for it. Suppose the voters of California have made a colossal mistake and get no benefit of any kind from the train. Since each $800m of construction costs can be considered separately, you must decide which is worse; to throw away $800m now or throw away $53.3m each year for the next 30 years. Or if you were one of only 10 million taxpayers funding this absolutely worthless train: Would you be worse off if you burned $80 now, or if for each of the next 30 years you misplaced $5.33, never to be found again?

To answer this sort of question, you should consider the alternative uses for money in the hand. Assume Californians want to invest $800m this year. The State can levy an extra $800m in taxes this year and spend it on building railways. Or the State can borrow $800m, and leave $800m in the hands of its citizens. Assuming the citizens are reasonably wise they can use that $800m in various ways to generate a rate of return at least equal to 5.125% p.a. (this assumes that the “market rate of profit” is at least as great as the interest rate at which the State can borrow). Thus, with this return and the capital of $800m, Californians have the means to pay the extra taxes required for the next 30 years to pay off the debt. The options are equal because the immediate payment of $800m incurs an opportunity cost, the loss of its potential returns if it were used in other ways.

The theory of present value holds that rational economic decisions cannot be made unless payments made at different times are “discounted” to a common time, using the rate of interest. As an example of waste incurred by ignoring present value (or opportunity cost), suppose we could wait 7 years, and then, using some newly available equipment, spend $13b over 8 years to build the train by year 15. By ignoring the present/future value of money, we would conclude this new option is $1b worse than spending only $12b over years 1-15. However, we could invest $7.29b now to earn at least 5.125% p.a. and use the capital and earnings on it to pay for the construction of the train over years 8-15 or we could invest $8.18b now to pay for the construction of the train over years 1-15; in other words the present value of the first option is $7.29b while the present value of the second is $8.18b. By choosing to pay only $12b we have actually wasted $0.89b (and all its future earnings). Of course, I have ignored here the possible benefits of starting construction sooner to have some parts of the railway operating before the entire thing is completed.

The payments on the bonds issued to build the train could also be funded by investing $8.18b now to earn 5.215% p.a. At the end of 45 years, $8.18b and all the return earned on it, would have exactly met the payments of $24b to bond holders. This option costs the same present value as paying up-front and at least has the advantage of making the true costs plain for everyone to see; by paying up-front many would think they have saved the large interest bill, and congratulate themselves on passing on that saving to future generations. What they have overlooked is that by depriving themselves of the earnings of $12b over the term of the loan, they, and therefore their dependants and beneficiaries, have borne a cost at least equal to the saving passed on. The future residents of the State, including new arrivals and tourists, are spared the interest payments; the present residents of the State, and their descedents and beneficaries, whether residents or not, bear the cost.

macroman said...

Richard Carrier wrote: “Now, if 71% of voters already voted to compel our budgets to be balanced, how on earth could any bond measure pass?”

You may have misunderstood Proposition 58: The California Balanced Budget Act, that was approved by 71% of voters. Proposition 58 established that the “General Fund” expenditures must not exceed the General Fund revenues. This basically means that taxes must cover State salaries, “consumerables”, maintenance, and any debt repayments for that year, but not new capital works.

The California constitution has always allowed borrowing for specified capital works, which must be approved by ballot if greater than a certain cost. Proposition 58 did not change that, which is why it is likely that Californians will continue to be asked to vote on road bonds, hospital bonds, school bonds, University bonds and anything else the State chooses to build.

Richard Carrier said...

Macroman:

1. Thanks for the accounting lesson. But I know all that. I never imagined a lump sum loan, for example, so that was a lot of wasted typing there. Especially as none of that has anything to do with my point: taxes still have to be raised (you have to pay the loans). Thus, there is no use in getting the loans if the exact same tax raise can simply pay for the project. And it demonstrably can, in almost every case (especially the train, as I showed). Deferring this cost only has use if you don't have the money now. But we almost always do.

2. Thanks for the correction on 58. However, I extremely doubt even a significant fraction of voters made any such distinction. They all voted for balanced budgets I'm sure. That's what they thought they were voting for. Or it was what they wanted to vote for even if they thought 58 was the best they'd get. But IMO they probably had no idea there was a loophole. At least, I'll bet a poll would prove the vast majority didn't know that, and/or if given the option, would vote in an amendment that ended all non-emergency borrowing. As national polls consistently show this is the view of a majority of Americans, even liberals.

macroman said...

Richard Carrier said:

“I never imagined a lump sum loan”

Your statement “on the loans we pay $800m per year for 30 years” is consistent with a lump sum issue of $12b of 30-year bonds, and an interest rate a little over 5%, so I assumed that is what you meant, but I didn’t analyze what I thought you had in mind.

Let me repeat where I agree with you. Option A (paying $800m p.a. over 15 years, your debt-free option) is obviously better than Option B (paying $800m p.a. over 30-years), since Option B includes an extra 15 years of paying the same taxes. A State Government which selected Option B, or presented Option B to the voters, would be deplorable. Its action would be inexplicable. Voters who accepted option B, knowing the facts, would be the idiots you say they are.

But Option B is not what was offered, and approved, and there was no need to analyze it further. I analyzed the cost of the actual schedule of bond-issues that will occur (Call it Option C): the State will issue 30-year bonds when needed over the (assumed) 15-year construction period. The payments (hence extra taxes) vary over the 45 years following the start of construction. According to the present value theory, those payments, taken from taxes, have the same value as taxes of $800m a year for 15 years during construction (the debt-free option A). In other words, the repayments on the bonds have the exactly the same present value as the money paid to buy the bonds (virtually a tautology under the assumptions of present value theory).

Of course, if these options A and C are equal in monetary value, other factors could swing the balance. One might question whether present value theory correctly estimates the opportunity cost of Option A, or one might say that long-term fiscal planning is easier with option C. Or one might say the State cannot afford or doesn’t need the project at all. But these other factors can hardly be discussed sensibly unless it is accepted that the costs of options A and C are at least roughly equal, because Option A incurs an opportunity cost roughly equal to the interest costs of Option C.

Note: I used your estimate of the cost and time required to build the train: 15 years at $800m per year = $12b which is bit larger than the figure you had in the original post. But the argument doesn’t depend on the actual figure.

macroman said...

Richard Carrier said:

“... most people in history will get a free bridge, so why does it matter how many generations pay for it? Is two generations fairer than one? Why?..."

Answer: Most people in history will not get a free bridge, because the bridge depreciates over the term of the loan. After that it will need constant repairs and maintenance. The depreciation costs/maintenance costs will be roughly equal to the loan repayments. With no borrowing, one generation would get a free bridge courtesy of taxpayers who were almost certainly poorer, and many of who died before getting much benefit from the bridge.

Richard Carrier said...

Macroman said... [Option C:] The payments (hence extra taxes) vary over the 45 years following the start of construction. According to the present value theory, those payments, taken from taxes, have the same value as taxes of $800m a year for 15 years during construction (the debt-free option A). In other words, the repayments on the bonds have the exactly the same present value as the money paid to buy the bonds (virtually a tautology under the assumptions of present value theory).

Maybe in some abstract sense of "value" but not in the actual state budget.

Stick with a concrete case: the California supertrain (which, incidentally, was voted a go). That initiative states that $647m will be paid out over 30 years to cover the bonds. Period. (And that only covers construction--estimated operation cost will be an additional $1b per year, indefinitely). This will provide the designated $9.5b for construction. But if construction were stretched out over 15 years (as it most likely will be anyway), it would cost $647m per year, exactly the same as the bonds, only the bonds we keep paying--for another additional 15 years. That means the state budget will have to cut $647m from schools, police, et al., every year for fifteen more years than would be the case if the train were built without bonds. No talk about "value theory" will change the fact that we will have $647m less for schools, police, et al. in those years, than if we'd just paid for the train. We will lose those services (fewer cops on the beat, fewer teachers, etc.). That's a tangible fact.

Now, suppose we stretched the bond payout to 45 years (not in fact what the initiative says) without raising the total interest accumulated (by staggering the bonds, so it still only costs a total of $19b divided by 45 years, even if some years its more and some less). That gives us an average drain on the state budget of $422m. So we can lose $422m of cops and teachers for 45 years (!) or lose $647m of cops and teachers for only 15 years. (If taxes are raised instead of cutting budget, or any combination of the two, comparable results occur.)

Now, in some complex calculus, maybe it could be shown that losing $647m in cops and teachers for 15 years is worse than losing $422m in cops and teachers for 45 years, but I doubt many voters actually know how to run that calculation, and certainly no one has run it (not the state, not the initiative's own proponents, nor even the voters who could run it if they thought to, much less the vast majority of voters who couldn't anyway). And it would involve highly subjective estimates of the relative value of trains, cops, teachers, short term vs. long term drains on budgets, and so on, so everyone will get a different outcome based on their different priorities and solutions.

For instance, if all the cops in California took a pay cut of just $2,700--their average salary is six figures, BTW, so such a cut is trivial--that would save $225m and thus neutralize any annual difference between the 15 year and 45 year plans, in which case the value of the 45 year bond plan is demonstrably worse, in fact three times worse, than the 15 year straight purchase plan plus the simple budget compromise.

macroman said... Most people in history will not get a free bridge, because the bridge depreciates over the term of the loan. After that it will need constant repairs and maintenance.

I already said that (you must have missed my remarks on the eternal maintenance cost, even though it was item 1 in the same numbered list you are responding to). That cost is paid by everyone, even the people who pay to build the bridge (bridge maintenance costs begin immediately and are always high, right from the start). Hence I'm not talking about a bridge that's free to maintain, but a bridge that comes into existence for free (i.e. when I move to a city with a bridge, technically I just got a free bridge--but that doesn't mean I don't pay for its upkeep, only that I didn't have to pay to build it).

[Depreciation, BTW, is not a real cost for property you will never sell--e.g. there is no depreciation faerie sending us a bill for actual dollar amounts as a bridge ages. In reality, for public property, depreciation only affects duration of utility, i.e. when the bridge has to be replaced.]

macroman said...

Richard Carrier said

“… maybe it could be shown that losing $647m in cops and teachers for 15 years [total $9.7b] is worse than losing [a total of $19.4b] in cops and teachers [over] 45 years, but I doubt many voters actually know how to run that calculation ..”

I did not say worse than – I said equal to. The fifteen years of borrowing and the varying repayments (these must be stated exactly) make it confusing, but it can be simplified with no loss of relevant information. The voters need to consider only one of the fifteen years because every other year requires exactly the same comparison.

The choices are (a) lose $647m in cops and teachers in one year only or (b) lose $43.3m in cops and teachers every year for the next thirty years. This can be scaled to amounts more familiar to the individual: (a) lose 15 teachers/cops for one year or (b) lose 1 teacher/cop for 30 years.

I think voters do have some idea of for the costs of option (a) (bad policing or bad education for one year). I think they can estimate the negative effects on State residents of one fewer teacher/cop during the next 30 years under (b). Is it really so difficult to make some sort of choice? Economists seem to think that people do this sort of “calculation” naturally all the time.

If the voters have a natural time-preference of about 5% p.a. for immediate as opposed to future costs (or benefits), then it is a close decision and they can think further. They might then think that sometime in the next thirty years the State will be wealthy enough to employ that one extra cop/teacher. They might think about the effects of a sudden large change in State employment under option (a).

macroman said...

Richard Carrier said: “bridge maintenance costs begin immediately and are always high, right from the start”

I was thinking of maintenance costs (I should not have said depreciation). I assumed maintenance costs would not be high right from the start but would increase with time to some constant level. But I agree now that the possible variations of maintenance costs would have a small to negligible effect on the distribution of the costs amongst those who get the benefits.

macroman said...

Richard Carrier said: "the California supertrain ... initiative states that $647m will be paid out over 30 years to cover the bonds. Period."

In fact, Proposition 1A asked voters whether they approved of a proposed Law. The Law (
Text of Proposed Law, Proposition 1A
) was available. It says

2704.13 ... Successive issue of bonds may be issued and sold to carry out these actions progressively, and it is not necessary that all the bonds authorized be issued and sold at any one time.

Past practice has been that bonds are issued as required (see the Annual California Debt Affordability Reports).
Commonsense says that no State Treasurer has any motive for doing otherwise and in any case, any money borrowed before it can be spent can be invested
to earn enough to cover the bond repayments without new taxes.


Except for the silence of your readers on your assertions about the repayment schedule, we have no evidence that any other voters
misunderstood the proposition.
The “calculation” the voters need to make is just like that they make when they apply for a mortgage.
The difference is that the State is applying for a “new mortgage” for every year of construction.

macroman said...

Richard Carrier said:

“For instance, if all the cops in California took a pay cut of just $2,700 … that would save $225m and thus neutralize any annual difference between the 15 year and 45 year plans, in which case the value of the 45 year bond plan is demonstrably worse … than the 15 year straight purchase plan plus the simple budget compromise.”

You are making the same mistake as you did in your original imagined schedule of bond repayments; you have assumed (implicitly) that the repayments will be made some years before necessary. So of course this imagined schedule of repayments has a greater present value than “straight purchase”.

The actual repayments are zero for the first year of construction, $43m for the second, $86m for the third year and so on. The “annual difference between the 15 year and the 45 year plans” starts at $647m (a pay cut of $7,764 for cops). The annual difference does not fall to your assumed value of $225m until the eleventh year.

Richard Carrier said...

Macroman said... The choices are (a) lose $647m in cops and teachers in one year only or (b) lose $43.3m in cops and teachers every year for the next thirty years.

But since we are doing this fifteen times (assuming a 15 year project duration as we have been), (b) still averages $647m every year for the next thirty years, instead of $647m for just those fifteen years. The only difference is that the hurt is delayed (with the bond plan the worst payment years will come at and after the end of construction), but it's still twice the hurt. I've just been doing the math with averages. But let's do the math in terms of the actual physical reality...

In reality--the reality we actually have to live in--we lose $647m in cops and teachers every year for fifteen years on the straight build-out plan. That's a material fact. But the alternative is not $43 per year for thirty years, but more than $647 for the same fifteen years postponed. Assuming project costs are the same every year (and of course assuming all is kept within budget), a $9.7b train will cost the state $647m each year for 15 years. If instead of just paying this, the state issues nothing but 30-year bonds (just to keep the math easy), then this is what happens:

year 1: state borrows $647m
year 30: state pays $1.3b to those bondholders
year 2: state borrows $647m
year 31: state pays $1.3b to those bondholders
etc. (for fifteen years, at which point the state will have paid out the twenty billion projected, for a ten billion train)

Now, in this scenario, yes we can say we lose zero teachers/cops for thirty tears, then lose $1.3b of them every year for 15 years thereafter. But the choice is still between losing $647m for 15 years now or losing twice as many teachers/cops for the same 15 year period, only later. This is a no brainer. Debt is clearly the worse idea of the two, if you can afford either.

Now, in reality different bond issues will be involved, some 1-year short-term, some 5-year, and some 10-year, and so on. But that doesn't change the fact that paying back those bonds will cost the state more overall (in fact double, according to the state's own projections) what it would have paid just to build the train. The only difference is that the hurt comes later rather than sooner. But eventually the state has to pay. Even if the state gets lucky and can fully finance everything with short-term bonds, though it will pay less than double, it will still pay more what the train actually cost. So for as many years as it took to build the train, the state will be paying more per year, only later, what it would have paid per year earlier.

For example, if the first year of construction (= $647m in bonds issued) was fully financed by 5-year bonds, and those cost the state (at the 5-year maturity mark) only $80m in interest to pay back (i.e. the bondholders that year take $727m from the state), then five years later we will lose $727m in teachers/cops for that year, rather than having lost $647m in teachers/cops the first year. And so on, for any bond sale or combination thereof you imagine. It always comes out worse in real-world cost to the taxpayer (as in actual physical losses in a single year, the only difference being which year that will be).

I think you are still failing to understand the difference between this and for-profit ventures, where the "costs more in year five" reality can be offset by "profits over the intervening five years." There are no profits here. Zero. This is a non-profit venture funded by the taxpayer.

Indeed, the taxpayer doesn't end up with a private train in his garage. He doesn't even get to use the train for free. He is eating the cost merely so the train can exist. Which is why it's ultimately a decision between relative values of different taxpayers: do we want this train to exist, or to increase the number of teachers and cops for the next fifteen years (since we could just as well take out the same bonds to do the latter)? But once you've made that decision, you have another: do I want this train to cost more per year for fifteen years later, or less per year for fifteen years now?

This can be scaled to amounts more familiar to the individual: (a) lose 15 teachers/cops for one year or (b) lose 1 teacher/cop for 30 years.

Not in reality. In reality a straight build will cost us $647 each year for 15 years. That's over 3000 police officers (counting all expenses, e.g. medical and other insurance, pensions, cars, academies, etc.). If we use 30-year bonds instead of just paying the cost, then in year 30 we pay back $1.3b in bonds issued in year 1, equal to over 6000 cops, and we have to keep doing that for 15 years until the train is paid off (since we will have issued as many bonds each subsequent year to sustain the project). So we lose twice as many cops for the same number of years, just later.

Though the economic product of California might grow in 30 years, so will the cost of services (so that the state might be making twice as much income, but cops might cost twice as much to hire), and the dynamic between them is quite hard to predict, and certainly not within the means of the average voter to predict. Indeed, if a 30-year bond bought at $100 and paying out at $200 will actually be worth only $100 in buying power in 30 years, you would have to be a fool to buy that bond in the first place, yet only if that happens will the state break even in actual costs (i.e. only then will $1.3b equal the loss of only 3000 cops rather than 6000 and thus be the same as having lost those cops 30 years earlier).

I think voters do have some idea of for the costs of option (a) (bad policing or bad education for one year).

Let's be honest. We (and hence they) have no idea what the actual differential effects will be of selecting between (a) or (b) on the future of their children or the state economy or crime rate (etc.), even if (b) did involve fewer lost per year (though in reality it doesn't, even if in a fictional mathematical realm it does).

That is, they may know what losing cops and teachers means in general, but they won't know what the difference of losing 1 for 45 years will be from losing (say) 2 for 15 years. Whether that difference is worth it or not is beyond most people's ability to calculate--and depends on relative values anyway (and thus not everyone will come to the same result). And of course, we are not in fact talking about 1 or 2 teachers/cops, but thousands (or, of course, the equivalent, since the hurt will be spread around). There is a huge difference between calculating what effect losing 1 cop will have on the entire state of California, and calculating what effect losing 6000 cops will have on the entire state of California. I can't tell you what the latter will be, so surely the average voter can't. Other than to predict it's unlikely to be good. Likewise for the inverse (raising taxes--especially if the tax hike is progressive or targeted, hence the still-unresolved debate between trickle-downers and consumer-stimulators, etc.).

If the voters have a natural time-preference of about 5% p.a. for immediate as opposed to future costs (or benefits), then it is a close decision and they can think further.

Most voters have no idea what that sentence even means.

Actual intuitions here have been demonstrated to be irrational anyway, so relying on uninformed human intuition is probably a bad idea all around (the psychological study of delegated value shows that people are very bad at assessing the differential merits of immediate vs. delayed rewards, etc.). But that's a whole other debate.

They might then think that sometime in the next thirty years the State will be wealthy enough to employ that one extra cop/teacher.

But they won't be able to know reliably how much wealthier, if at all, the state will be vis-a-vis the actual cost of that cop (etc.) in thirty years. You have an unusually unrealistic idea of what average people actually know.

They might think about the effects of a sudden large change in State employment under option (a).

There is no way for an average voter to calculate that (unless they can expect to be among the employed, but that would be a calculation of self-interest rather than community benefit--not that there is anything wrong with that, only that it's a completely different consideration). As far as the link between increased employment and state revenue, the state cannot possibly get more money back from laborers it pays, than the money it pays them. And it can only get that much if it taxes incomes (of workers) and businesses (where the workers spend their money) at an absurd rate of 100%, which it won't.

Actual wealth generation through tax spending requires (a) trade with parties outside the state (which this example involves nothing of) or (b) new industries or inventions that increase materials or productivity--which a supertrain is very unlikely to have much impact as. Though it will have some, directly and indirectly, it's unlikely to be so much as to exceed what was spent on building it--and in any case, hardly any voter can calculate what this impact will be, much less whether it will be enough that, when taxed, increases state revenue significantly enough to offset the losses incurred in building it. And even then, those losses will be less if the train is just paid for now. Will the train increase revenue so much that 6000 cops in thirty years will cost the same as 3000 cops now? I find that unlikely, and in any case have no idea how I would answer that question, and neither does the average voter.

Except for the silence of your readers on your assertions about the repayment schedule, we have no evidence that any other voters misunderstood the proposition.

I'm a voter, and I did (I assumed the guide correctly stated what the law said). I seriously doubt a poll of voters would find even a tiny fraction who thought the bond expenses could be stretched out over 45 years instead of 30 (up to forty years maximum maturity by law, hence maximum period of 45 years as you said, if we assume 15 years to project completion as we have been). But the amounts in the voter guide are projected-actual, in other words most probably it will in fact be (an average of) $647 million over 30 years, and not (an average of) $422m over 45 years, although I've already addressed the latter possibility anyway.

In any case, when we look at the actual annual bond payouts (when the state will actually, physically have to buy the bonds back, and what that will actually, physically cost the state at that time), this becomes moot.

The “calculation” the voters need to make is just like that they make when they apply for a mortgage.

Except that (a) most voters can't do that (most require the bank to explain the options to them, as IMO the recent collapse of the mortgage market proves--indeed, evidently, even the experts couldn't do this math) and (b) you can't live in the supertrain (so the analogy is completely inapposite) and (c) people aren't governments. For example, everyone pays a monthly to live somewhere already (rent), so a mortgage payment for them is compared with rent. There is nothing straightforwardly comparable for the supertrain. Likewise, you get to live in a house. There is no straightforwardly comparable benefit to a state supertrain. And most people aren't budgeting for private school, private security, road maintenance, private fire-suppression squad, management staff, etc., which they would have to cut from to divert funds to a mortgage--yet there is no easy or obvious or even conclusive way to evaluate that. And people are paying a mortgage from their own income, whereas the state is paying for the train with taxes, which are distributed unevenly (sales taxes target specific groups, income taxes are progressive, and both affect money available for investment in the state economy, etc.) and can be set arbitrarily (we can't just choose to make more money, but the state can just choose to take more taxes).

But most importantly of all: if we could afford to buy land and build our own house without a mortgage, then none of us would ever take a mortgage (except those few aiming to make money selling their house). That's my point. It's always cheaper to just do it than to borrow to do it, if in fact you can afford to. The only reason to borrow is if you can't afford to buy it outright, which is the only reason mortgages exist at all. If people had the money, and didn't need that money for something else, they would just buy a house. Hardly anyone would take out a mortgage and pay twice as much, even though it would be stretched out over many years. Though mathematically we could show them that at a certain balance of monthly and total cost, the mortgage would cost them the same overall (or even less), we would have to take great pains to convince them of that.

So living in reality, as we do, we need to consider what real human beings actually are capable of and actually value and actually do.

Richard Carrier said...

P.S. To get us back to the original point: (a) none of the bond advocates make any of the arguments you do, but instead (b) foster the illusion that buying this train with bonds won't raise taxes, when in fact it necessarily will (or else we must lose services). We can only quibble about how much.

Hence I would rather these measures all be voted down until people honestly say where the money is going to come from, and why taxing us longer is better than taxing us more for less time (and then actually show that their specific plan will do this). Indeed, they should explain how people receiving the bonds expect to make any money, if in paying down those bonds we are losing none.

Be that as it may, we voted for a balanced budget. And that ought to mean we don't get to buy things unless we pay for them--regardless of whether "paying" means actually buying what we want or paying down bonds for it. Until bond measures honestly address that reality, they should never have our support.

...except in times of economic distress. Though even then need should trump luxury--I would rather have seen that $20b spent on a network of solar-thermal plants (in our ample low-rent, high-sun deserts) to finally replace our gas-and-coal electrical supply. Instead we bought a lousy train. All that does is go fast, at a cost of actually increasing our dependency on fossil fuels to meet its vast power requirements. But that's a whole other rant...

macroman said...

Richard Carrier said:

“P. S. To get us back to the original point: (a) none of the bond advocates make any of the arguments you do, but instead (b) foster the illusion that buying this train with bonds won't raise taxes, when in fact it necessarily will (or else we must lose services). We can only quibble about how much.”

I am not sure exactly who the bond advocates are, but you should distinguish between bond advocates and train advocates. As I indicated, someone may think the train is the worst option to solve various problems, or does actual harm, and yet think, now that the disastrous project has to be paid for, that the best way to throw away $10b is via installments over many years.

You may be referring to the Voter Information - Argument and rebuttals which says “without raising taxes”. That looks like a commitment to not raise the State sales tax or income tax rates. Since the Voter Information mentions the $647m annual cost of the bonds, there is no suggestion that the bonds are a “free” option. Clearly some or all state services would have to be cut, as you say, or not increase as fast as they might otherwise. There is nothing unusual about Governments deciding spending priorities for a limited amount of revenue.

Dig a little deeper in the arguments and rebuttals and you find the claim that “Building high-speed rail is cheaper than expanding highways and airports to meet California’s population growth.” That might be interpreted as a specific claim of what can be cut: airport and highway spending. That would go some way to getting more people to pay for their transportation (by using the train), rather than rely on the “free” highways, with all the “spill-over” costs toll-free roads impose on everybody.

macroman said...

In answer your question of why anyone would buy bonds if the repayments have the same (present) value as the capital sum:

Risk adverse people buy Government bonds because they value the repayments more than the capital (and they don’t want to put the capital into more risky investments). Less risk adverse people (risk-takers) amongst taxpayers value the future taxes (repayments) less than the capital sum. They'd rather not pay the lump sum tax now because they think that money left in their own hands can earn them more than enough to cover their future taxes.

You will probably argue (or have been arguing) that the risk-takers are wrong; they undervalue future repayments or they dishonorably hope someone else will pay the future taxes. The risk-takers could say that you neglect or underestimate the other uses they have for their money now, and they will pay future taxes when due out of their own future earnings.

macroman said...

Richard Carrier said:

“if a 30-year bond bought at $100 and paying out at $200 will actually be worth only $100 in buying power in 30 years, you would have to be a fool to buy that bond in the first place”

I only a fool would buy the bonds, then the voters are not “drooling idiots” to sell them.

I was wondering when you would realize that inflation makes the bonds an even better option (because inflation lowers the effective interest rate, possibly to zero). I tried to indicate this when I wrote “you might consider whether those who bought the $35m worth of Golden Gate Bridge bonds in 1930-7 were getting a good deal when they received $35m+$39m over 34 years, or whether it was the citizens of the Bay area who got a good deal”.

The point about the State getting richer is that the historical trend is for per capita wealth, adjusted for inflation, to increase with time. If that continues, it will become even easier for the State to make the repayments, since they represent a decreasing fraction of the State’s inflation-adjusted GDP. It is irrelevant, now that you are going to have it, whether the train is a dead loss – it is still a dead loss if you pay up-front, but then you pay a greater portion of the State’s GDP.

macroman said...

Richard,

This may be another lot of “wasted typing”. However

(1) It seems you have misunderstood how the State repays money to bond-holders. For one $10m bond (say) the State does not pay $20m to the bond holder thirty years later. It sells some form of “coupon bond” - a legal document which has 30 coupons attached to it. Each coupon says something like “The State of California promises to pay the bearer/owner $667,000 on (a given date, one for each of thirty years)”. The bond bearer/owner sends the coupon to the State on the due date and gets the $667,000.

(2) The California HSRA released its business plan in November 2009 (after the ballot, which is disgraceful I know). Scaling from Fig. 27 in the plan we get the following planned spending of Prop 1A funds over 7 years starting in 2009 (there may be some delay in starting).

By year: $100m, $150m, $400m, $1300m, $2300m, $2850m, $2850m.

For whatever reason, the State does not want to find room in its budget for up to $2850m in one year. Hence it sells 30-year bonds (I think you accept that the bonds are sold when needed, not before) : i.e. the State will receive from bond buyers the above amounts over the years 2009 to 2015.

Assume the State can sell the coupon bonds described above at face value (assume an interest rate of a little over 5.2%). Then the schedule of repayments that the State will have to meet is (millions):

2010: $6.7
2011: $16.7
2012: $43.3
2013: $130.0
2014: $283.3
2015: $473.3
2016-2039: $663.3
2040: $656.7
2041: $646.7
2042: $620.0
2043: $533.3
2044: $380.0
2045: $190.3


Was the two-point summary in the voter information deceptive? The first point said ” … costs of about $19.4b, assuming thirty years to pay off both principal ($9.95 billion) and interest ($9.5 billion) costs of the bonds. Payments of about $647 million a year”.

The summary says “about” and “assuming”. The summary divided the total cost by 30 to get $647m a year, and got close to what will probably pertain, because nearly all of the money will be borrowed (and spent) in just two years.

I think that is a fair summary: it implies that each $1b borrowed will require $2b in repayments over thirty years. You have all the information you need, whatever the timing of the borrowing/spending on the project is.

macroman said...

Richard

You imply voters are too stupid or somehow cannot make decisions about future payments/costs. I still doubt it.

Suppose the Government said to each taxpayer: "You, personally, can pay an extra tax of $995 this year, or you (and your estate if you die) can sign-up for an extra tax of $64.7 a year for the next 30 years".

I think you have implied you go for the $995 and anything else is irrational. I don’t think you have thought it through.

Imagine someone who could use $995 he had saved to publish his next book. Just like those people in the 1930s who could have paid for the Golden Gate Bridge, he won’t starve if he pays the $995 tax now. Can you really think of no sane reason why he might chose to pay $1941 over 30 years?

Other people with $995 to spare may have what they think are equally good reasons to prefer the time-payment option.

Have I mentioned "opportunity cost" before? It is the key to present value. You have to upset present value theory (otherwise your arguments make no sense) which means you need to discredit "opportunity cost". The lack of an "opportunity cost faerie" is not going to be enough, in my view.

You could try lack of opportunity - the voters are too stupid to find anything useful to do with the $995. You could try bad business conditions: the economy is so bad that no one, however smart, can find anything useful to do with $995. When there are no opportunities we may as well throw away $995 now, as throw away $1941 later. (I am adoptiung your jaundiced assumption that State capital works funding, of any kind, is money thrown away. The reality is that the State, which doesn't agree with you, will nevertheless want the $995 now or the $1941 later).

A more reasonable form of the first approach is that the voters are generally risk adverse, and would rather take the sure saving (of interest not paid) over the nebulous benefits of investing the $995. This partly explains why I generally paid extra cash, as it became available, into my mortgage to shorten the term: it was a sure benefit as opposed to an uncretain benefit from investing the extra cash.

(The tax situation, which is probably different from California's, also made the saved interest equivalent to a tax-free return at the mortage interest rate, whereas investment earnings are subject to income tax: but that's another story).

macroman said...

I see that the State can buy back old, and issue new, Prop1A bonds in order “to effect a saving in debt service cost … as measured by the present value …” (Proposition 1A law, Section 2704.19 and California Government Code Section 16780). If present value is a misleading estimate of cost, your State may get into more financial trouble than you thought.

macroman said...

Richard Carrier said

`I think you are failing to understand the difference between this and for-profit ventures, where the “cost more in year five” reality can be offset by “profits over the intervening five years.” There are no profits here. Zero. This is a non-profit venture funded by the taxpayer.’

I think you are failing to understand why I have said that my argument does not depend on any profits from the project. It depends on the profits of everything else the taxpayer could spend his or her own money on. It is an argument about how to pay; not about whether you should buy.

The State can take (a) X dollars from me now, or (b) X + Y dollars from me later. If (b) I now have X dollars to invest elsewhere (it could be in “intangibles” like my own education or my own health). To break even I need my wealth to grow at an annual percentage rate, over the long term, equal to the rate of interest the State pays on bonds (about 5.2% in good times). The value of X dollars to me now is equal to the value of X + Y dollars to me later because X dollars now can be transformed into X + Y later, by my own decisions.

macroman said...

Richard Carrier said `Though the economic product of California might grow in 30 years, so will the cost of services (so that the state might be making twice as much income, but cops might cost twice as much to hire), and the dynamic between them is quite hard to predict ...'

For every 30-year period from 1928-1958 up to 1978-2008, US GDP (in dollars) has increased by more than a factor of five (often by much more). In 1978-2008, the factor was six. It seems conservative to assume that California's revenue, with no increase in tax rates will be five times larger (in dollars) in 2039. In that case the fixed annual bond repayments of $650m, which start at 0.6% of revenue in 2009, will be 0.27% of revenue in 2024 and 0.12% of revenue in 2039. Thus by 2024 the Legislature will find itself with an extra 0.3% or more of State revenue not committed to bond repayments. That 0.3% of revenue can be allocated amongst the various State services (including cops) according to the cost and benefits of those services at the time. If economic growth continues as usual, borrowing can easily be a better option than sacrificing a large portion of this year's State revenue now.

Richard Carrier said...

Okay, you've convinced me on the main points. I get what you're saying now.

Just a few points left...

Macroman said... you should distinguish between bond advocates and train advocates. As I indicated, someone may think the train is the worst option to solve various problems, or does actual harm, and yet think, now that the disastrous project has to be paid for, that the best way to throw away $10b is via installments over many years.

In the present exchange I'm only speaking about bond advocates, a category that by logical necessity includes the people you just describe, since what they think of the train is irrelevant to the only point currently at issue: whether this particular bond plan is the best way to build it.

So what I should have distinguished are actual bond advocates over here in California (not just related to the train--there are dozens of these bond projects), and ideal bond advocates, i.e. people who are honest and correct about the actual costs of their project (and who advocate bond plans that actually are better than direct-buy plans, since in the galaxy of all logically possible bond plans, a third will be worse than direct-buy plans, a third the same in cost, and a third better, in some manner or other, once we define the measures we are comparing as indicators of "better" and "worse").

You may be referring to the Voter Information - Argument and rebuttals which says “without raising taxes”. That looks like a commitment to not raise the State sales tax or income tax rates.

The bill does not say that, i.e. the law does not prohibit raising taxes. Hence the point is: the money has to come from somewhere. Dodging that fact is dishonest. So either our legislature will have to raise taxes to cover it, or cut services. Claiming the bill itself doesn't raise taxes is a rhetorical way to dodge this reality. And yet in reality, if voters expect no taxes to pay for the train, they can only expect cuts in services, and yet voters would not likely vote for a bond plan that said "removes $647 million from the budget of state schools and police to build a train," for example. I think it is thus playing on the stupidity or ignorance or manipulability of the voters, rather than being honest with them about where the money for this train is actually going to come from.

For example, "'Building high-speed rail is cheaper than expanding highways and airports to meet California’s population growth'...might be interpreted as a specific claim of what can be cut: airport and highway spending." Indeed, if the bill said "will remove $647 million from current airport and highway budgets to build a train," that would be the sort of honesty I'm talking about. And yet watch how fast that bill gets voted down, and you'll see what I mean.

Risk adverse people buy Government bonds because they value the repayments more than the capital

But they still get that capital back. Watch what happens when the government says, "Oh, you never get the capital back. You're just trading it for the payments." Then see who buys bonds. Remember, the government is spending capital, too. The government is just another buyer of investments like everyone else. The question is what the government gets for its investment, and whether it's worth it, or whether it could have gotten the same thing for half the price.

But you've persuaded me here that it might in many cases be worthwhile to pay double over twice the time.

You will probably argue (or have been arguing) that the risk-takers are wrong; they undervalue future repayments or they dishonorably hope someone else will pay the future taxes.

I wouldn't argue the first, and as to the second it wouldn't be dishonorable, because it's not a hope: it's a known fact to all parties that someone else will pay the future taxes (since there isn't anywhere else the money can come from, unless the government starts selling things for profit). After all, if the bond holder himself had to accept a tax increase exactly equal to the bond payout, he would never buy bonds.

It seems you have misunderstood how the State repays money to bond-holders. For one $10m bond (say) the State does not pay $20m to the bond holder thirty years later. It sells some form of “coupon bond” - a legal document which has 30 coupons attached to it.

I don't know if that's how these California bonds work, but it's not how Federal bonds work. I know, as I own some. It's a single bond, with a one-time total payout at term end (with penalties for early liquidation). No coupons. The law as written seems to suggest this is how the state bonds work, too (at least in the train case, but arguably it doesn't specify, so I can't say).

Suppose the Government said to each taxpayer: "You, personally, can pay an extra tax of $995 this year, or you (and your estate if you die) can sign-up for an extra tax of $64.7 a year for the next 30 years".

"...or neither."

Yes, that's exactly what the government should do (and require of special interest groups as well, when they get their own bond measures up). That's what I've been saying.

I think you have implied you go for the $995 and anything else is irrational. I don’t think you have thought it through.

I'm not saying it's irrational. I'm saying it's not that simple. The analogy isn't entirely apposite, because the cost of the train is not divided out this way. First, it's not each citizen ponying up his equal share (the rich will pay more, not only by linear percentage, but progressively, or the taxes will be divided among sales of a commodity, either way not hitting each citizen the same way--for example, there are easily five people in California who are paying the state $650 million a year in taxes already, just by themselves--hence a single coffee-clatch of five taxpayers could earmark their taxes to cover the entire expense of the train). Second, it's not just the one year. The bill comes every year for either fifteen years or forty-five, not one year or thirty.

Third, and most relevantly, I'm poor. I'd choose the lower tax simply because I had to (assuming I wanted the thing at all). The state doesn't have to. It already earns a hundred billion a year. So increasing taxes to cover $650 million a year entails raising its taxes by a mere 0.65%. So assuming we shared the burden equally across all citizens, the question isn't "$995 only this year or $64.7 every year for 30 years" but "$13 only this year or 86 cents every year for 30 years." And then, per the second point, it's actually "$13 every year for fifteen years or an ascending-then-descending amount in increments of 86 cents every year for 30-45 years, a curve that will probably top out at a $13 charge in the fifteenth year, ascending every year up to that, then descending every year thereafter." And finally, this is then sitting next to fifteen other bond measures and every other state budget item, e.g. "either that or 86 cents every year to permanently expand education and law enforcement."

For example, why not just do nothing but pay the entire state budget with bond measures? All police salaries: bond measure. All regular highway maintenance: bond measure. Etc. The answer would normally be (and could only be) bond measure projects are of a limited duration (otherwise, if it was always cheaper, we should pay for everything with bonds). But in reality they aren't, because another bond measure will immediately replace it. After the train is paid off, it will be a space port, or a solar thermal grid, or a nuclear power plant, or a desalination project, or a save the whales project, etc. Not that the government shouldn't do these things (even I would vote for some of them)--rather, it's that the government is always going to be doing these things. So why not just build it into the budget and be done with it?

The only available answer appears to be: "because then no one would vote for these things." If that's the case, then good riddens: if you can't sell the projects on their merits, then they shouldn't get paid for to begin with. But if you can sell the projects on their merits, then you should be honest about their costs, and where that money is going to come from. And if it then becomes an honest request, "Hey, with this clever bond measure we can give you an awesome train for a tax of just a couple dollars every year for the next forty-five years (i.e. effectively the rest of your life), which we think is a better deal than a tax of ten or so dollars every year for the next fifteen years," that would be great. If only that's what they did. It isn't.

Moreover, I'd want the opportunity to put a competing measure on the ballot, "Hey, we can build that same train straight up for that ten or so dollars a year for the same exact same pay period, but also build you a solar thermal grid and a desalination system, one of the three in each fifteen year period until the forty-five years are up, then we can do another sequence of fifteen-year projects, and so on forever, simply as a part of our new public works budget, why don't we do that?"

Here I'm not objecting to what you say about opportunity cost, as if I can't find anything to do with $995. Rather, what happens when I'm stuck with fifteen $65/year projects for 30 years? I'm still out my $995. And though I might get a bunch of cool stuff now (fifteen cool things even), rather than waiting for it to be built in sequence over thirty years, am I really supposed to believe that once all that stuff is built, we'll stop borrowing money to build stuff and just kick back and relax so I can keep my $995 every year for the next ten or so years I've got to live? The question has to be: why don't we just build this stuff into our budget? In other words, what's the point of the bond? Why do we need to do an end-run around the budget and procurement process we already voted for?

I think you are failing to understand why I have said that my argument does not depend on any profits from the project. It depends on the profits of everything else the taxpayer could spend his or her own money on. It is an argument about how to pay; not about whether you should buy.

This much you've convinced me of. I understand what you mean about that. The question thus is not whether paying twice as much over twice the time is always a bad idea, but when it's a good idea. Surely it can't always be a good idea, or else we should pay our whole state budget with bonds.

For every 30-year period from 1928-1958 up to 1978-2008, US GDP (in dollars) has increased by more than a factor of five (often by much more). In 1978-2008, the factor was six. It seems conservative to assume that California's revenue, with no increase in tax rates will be five times larger (in dollars) in 2039. In that case the fixed annual bond repayments of $650m, which start at 0.6% of revenue in 2009, will be 0.27% of revenue in 2024 and 0.12% of revenue in 2039. Thus by 2024 the Legislature will find itself with an extra 0.3% or more of State revenue not committed to bond repayments. That 0.3% of revenue can be allocated amongst the various State services (including cops) according to the cost and benefits of those services at the time. If economic growth continues as usual, borrowing can easily be a better option than sacrificing a large portion of this year's State revenue now.

But over those same periods money became worth less. So you have to adjust for inflation. If, for example, the cost of cops (etc.) increased by a factor of five from 1958 to 1978, that wipes out any advantage of quintupling revenue in that same period (even assuming increase in state revenue matches every increase in GDP). Although the bond payout will become less and less of the state budget, the state will be struggling with higher and higher costs for its regular services. They could cancel each other out, or render the advantage rather small. Hence my point that the truth is not simple, and is generally beyond any average taxpayer's ability to calculate (least of all when considering the interference of occasional but unpredictable crises like the current one, where bond obligations entered into ten years ago are now a more enormous burden on the state than would have been predicted at their issuance).

Better to have an adjustable budget and just build what you can afford when you have means, tightening the belt when you need to, and expanding it when you can. As in the example earlier, you can use bonds to build fifteen things now and keep paying for 45 years (as if bond measures will somehow cease being voted for in the interim), or you can just budget to build fifteen things over 45 years and pay exactly the same annual amount. The advantage of the latter is that you can slow or speed up expenditures as the economic conditions require, and aren't obligated by old debt.

macroman said...

Thank you, Richard, for your long reply. There are a few issues I would like to comment upon, and here I will consider one. You say “bond obligations entered into ten years ago are now a more enormous burden on the state than would have been predicted at their issuance” (because of the current bad economic conditions).
Yes, when revenue falls, the bonds are a greater burden than might have been expected. However, bond repayments consume less than 5% of State revenue and are not an “enormous” burden to begin with. To investigate the “more” onerous aspect, consider the effect of the current bad economic conditions.
California’s revenue in 2008-09 is expected to be $102b and to fall to $86.3b in 2009-10 (Governor’s budget for 2009-10, Fig REV01. Debt repayment obligation for the year 2008-09 was $4.4b and will rise to $4.9b in 2009-10, the net effect of new bonds issued and old bonds which were finally paid off in 2008(State of California, 2008 Debt Affordability Report, p14). Thus, because of the recession, debt service payments will be 5.6% of revenue, rather than 5%. I cannot see how this can be fairly described as a “more enormous burden”. The economic pain due to a 15.4% decrease in revenue should not be attributed to repayments on past debt, though I won’t be surprised if some demagogue tries exactly that.

macroman said...

Richard Carrier said
“But over those same periods money became worth less. So you have to adjust for inflation. … Although the bond payout will become less and less of the state budget, the state will be struggling with higher and higher costs for its regular services.”
I disagree and stand by my original analysis; expressing all costs as a fraction of State revenue automatically adjusts for inflation (and GDP growth). At fixed tax rates, inflation increases State revenue (in dollars) as much as it increases the cost of services (in dollars). The state could be struggling with higher and higher costs (relative to its revenue) for its regular services if the voters have opted to lower the tax collected (as a fraction of GDP) while demanding regular services worth the same or more (as a fraction of GDP). That is bad and irrational but it could happen with or without debt. It is certainly not the fault of repayments on past debt which fall as a fraction of revenue.
And if voters keep approving new bond projects so that repayments on past debt stay at (say) 5% of current revenue, the State will always have 95% of its revenue to devote to its regular services; it cannot be struggling unless voters opt for something like lower taxes (as a fraction of GDP) without a corresponding reduction in services (as a fraction of GDP). This is supposed to be impossible, since according to Proposition 58: revenue minus debt repayment must equal spending on services. (I am ignoring the “rainy day fund” established as a smoothing device).

Richard Carrier said...

I want to thank you for taking so much time with me here. I've learned a lot, and though I doubt anyone else will endure this exhaustingly long exchange, I did, and gladly. You have a convert. :-)

Macroman said... bond repayments consume less than 5% of State revenue and are not an “enormous” burden to begin with.

I agree, insofar as that remains the case. The Federal government is in quite the reverse situation, and thus a bad example to aim for, which is why I am keen to prevent my state from going down that path by being stuck with ever-increasing voter-initiated bond obligations.

expressing all costs as a fraction of State revenue automatically adjusts for inflation (and GDP growth).

That's fair up to a point. I still fear this plays into the illusion that inflation is a monolith, when in fact different elements of the economy inflate and deflate in value (e.g. averaging it all out gives us a figure called "inflation" but the reality may be quite different when you start considering specific expenses, e.g. teacher health care plans, or buying gas for the state's cars, trucks, and machinery, two occasions where average inflation doesn't track the actual inflation in these instances).

At fixed tax rates, inflation increases State revenue (in dollars) as much as it increases the cost of services (in dollars).

Similarly, this assumes a monolithic relationship that might not exist (e.g. it depends on which taxes are raised, on whom, and by how much, especially considering the complication of tax deductions and credits, and even if we say it's all a straight tax on income affecting all equally, this does not entail police or teacher salaries, for example, will only rise by the same amount as average state incomes), but you are right that this is still a reasonable assumption without evidence of any particular rising disparity.