November 09, 2011
Some infrastructure spending is more stimulative than others.
Yoshiyasu Ono of Osaka University had an interesting article in the Journal of Money, Credit and Banking this June on The Keynesian Multiplier Effect Reconsidered. In it he analyzes the special case of government spending on projects that are literally useless, such as paying people to dig a hole in the ground and then fill it back up. Although that's not proposed as an accurate characterization of any actual government programs, the extreme case of literally useless spending helps shed light on some of the issues involved. According to traditional Keynesian models, even for the case of a completely useless government project, if we were to raise private-sector taxes by just the amount needed to pay the salaries of the hole-diggers, GDP would increase, with a balanced-budget multiplier of one. Yet, Professor Ono asks, how could paying the crew a salary to dig a useless hole possibly lead to an improvement in welfare relative to simply handing them a direct transfer and allowing them to spend more time safely and comfortably at home with the family? And, to make things very simple, if the source of funds for paying the workers was in fact a tax levied on those same individuals, how could we possibly conclude that the enterprise has increased total national income?
The answer is, we include government spending, even on useless projects, in the definition of GDP, and assume that the value of what is produced is the dollar sum that the government paid for it. The reason even useless government spending has a balanced-budget multiplier of one is that we now have a filled-in hole that we didn't have before. So we have more goods and services (in the form of a newly filled-in hole) than we used to, and impute the value of this new extra stuff as added income for the nation as a whole.
Although it's an extreme case, analyzing the consequences of a literally useless government project clarifies that, even in the midst of the deepest depression, the concern should not just be how much the government is spending, but also the direct value to society of the project itself.
And even in the midst of the deepest depression, the job market is still very dynamic with millions of Americans finding new jobs every quarter, even as others are unfortunately still losing theirs. There are lots of private-sector enterprises where someone thinks the value of what a worker can produce is more than the worker would be paid. The challenge is how to find even more of them.
I for this reason end up reaching a different conclusion from Free Exchange, which argues that our energy policy should be a completely separate question from how to get Americans back to work; John Whitehead made a related point. My counter is that there are some clearly identifiable projects that would put people to work constructing some extremely valuable infrastructure. Such projects ought to be on the top of anybody's list of the kind of stimulus spending we would like to see at the moment.
Exhibit A for some energy infrastructure that the nation clearly needs is a better way to transport oil from the increasingly productive fields in North Dakota to the markets where it is most needed. A producer in North Dakota today can sell their oil for $82 a barrel, whereas refiners on the Gulf Coast are paying over $110 to import similar crude from countries like Venezuela, Nigeria, and Iraq. That price differential is an astonishing signal from the market of just how inadequate our energy transportation infrastructure is for the task at hand. Rail is being used to transport the oil as much as possible. But what's really needed is considerably more pipeline capacity.
And TransCanada wants to spend $7 billion of its own money (no federal dollars asked for at all) to build exactly what we need in the form of the Keystone Gulf Coast Expansion Project. The pipeline would add capacity to transport another 500,000 barrels each day from Canada, North Dakota, and other regions in the U.S. to refiners on the Gulf Coast. At a price differential of more than $20/barrel, that would generate over ten million dollars in new wealth every day. Beneficiaries of that wealth creation include the estimated 20,000 Americans who would work on construction of the pipeline and $5 billion in estimated new property tax revenue for state and local governments over the pipeline's lifetime. Here are details of how producers in North Dakota would connect to the project.
And here is a bigger view of the full project.
Why isn't oil flowing through these pipelines right now? Because the White House has spent three years thinking about whether or not to grant TransCanada approval to go forward with the project.
Some shovels take longer to find the dirt than others.
Posted by James Hamilton at November 9, 2011 07:38 AMdigg this | reddit
What if the economy consisted of 2 people E and U. E is employed, U is unemployed.
E has hole in his yard. He could pay U $10 to fill in the hole, or he could let the government tax him $10, after which, the government will pay U $10 to fill in the hole.
If E pays $10 directly to U then a service has been rendered and GDP rises by $10.
If E pays $10 in tax to the government, and the government pays U $10 to fill the hole, then a service has been rendered and GDP rises by $10.
However, if the government is borrowing the $10 from E at 10% for 1 year (for easy math), then E gets $11 back on 1 year. However, to balance the budget, the government must raise E's taxes by $11 in one year to cover the principal + interest associated with providing a $10 government service. (U is poor and does not pay federal tax.)
The point is that government has increased GDP by $10 today, but reduced GDP by $11 in one year because E has $11 less to spend,(assuming E spends all his income.)
What am I missing? How is the mutliplier 1?
Posted by: tj at November 9, 2011 08:59 AM
Common sense/politics. Just another oxymoron like restricting production of a resource ubiquitous to nearly all economic activity.
Posted by: CoRev at November 9, 2011 09:23 AM
Unfortunately, I have to agree with your take on the TransCanada Pipeline debate. The big picture is, we have grown to become very reliant on oil, and like many addicts, its hard to quit.
Environmentally conscience people; start pushing alternative energy and electric vehicles if you don't want to see pipelines continue to pop up, its the only way.
Also, stop oil subsidies! Subsidies should be going to emerging green technologies to help build up multiple industries, not an industry that is already well established.
Posted by: Jake Lopata at November 9, 2011 10:11 AM
But not all values are accounted for in GDP, or in the money economy. What about the environmental degradation of the tar sands operations in Canada? Global warming? Danger to the Ogallala Aquifer? Security risk of thousands of miles of pipeline? Loss of wealth due to trade imbalances? Perpetuation of an energy economy predicated on global supply lines from and through some very hostile territory? Concentration of wealth in petroleum companies who have done no favors for democracy?
Posted by: VinceInSeattle at November 9, 2011 10:24 AM
Why not just build a shorter pipeline to Iowa with a new refinery there? Then you have more competition in the mid-west for gas that can be supplied to the St.Louis, KC, Minneapolis, Milwaukee, and Chicago metro areas. Beats sending it all the way to Texas and then shipping it all the way back.
Posted by: psychicwebbah at November 9, 2011 10:33 AM
Hi James, pls. provide a post about the current financial situation in europe.
I guess, many of us would appreciate it to read your opinion.
sincerely yours, Johannes
Posted by: Johannes at November 9, 2011 11:11 AM
Good post. I've always thought there is "good" stimulus and infrastructure spending and "bad" infrastructure spending. If the economy is well below potential and resources are not being used, then it makes sense to put those resources to use building useful infrastructure spending, such as repairing failing roads and bridges or building pipelines. The problem as I see it, though, is that politicians and bureacrats, for a variety of reasons, are often unable or unwilling to separate the useful projects from the not-as-useful projects or to admit that their pet projects aren't always the most socially valuable.
Posted by: Brian at November 9, 2011 11:52 AM
I don't know why you would settle on the claim of merely 20,000 jobs created when the other document you cited promises 250,000 to 550,000 jobs? Heck, why not a million jobs or 10 million jobs?
In all of the sunny projections, there is no mention of the cost of negative externalities. Oil sands are one of the dirtiest fuel sources. Carbon dioxide emissions are even greater than for coal.
If we need a pipeline from North Dakota to Texas, why doesn't it stand on its own merits? Because it doesn't. It is a phony argument to sneak in the real goal which is a connection to the Canada tar sands.
This is all familiar ground. We went through the same rosy forecasts by industry lobbyists related to the Alaska pipeline in the 70s. That also was the pipeline that was going to save the economy, reduce unemployment and free us from the tyranny of OPEC. Now most of it is gone forever, most of it piddled away at less than $20 a barrel.
Posted by: Joseph at November 9, 2011 12:32 PM
JDH I don't think this is just a case of the Obama Administration dilly-dallying just for the hell of it. There are some very serious environmental issues, particularly with respect to the Ogallala Aquifer, as VincentInSeattle notes. It is the main source of fresh water for irrigation, not to mention a whole lot of people. And it has a very shallow water table, making it quite vulnerable to even modest spills. And if you're a terrorist looking for no good, then a vulnerable pipeline running across the Aquifer might as well have a big red bullseye painted on the side. As a practical matter it is an indefensible target that, if hit, would create economic havoc orders of magnitude worse than 9/11. You can rebuild towers; you cannot refresh contaminated aquifers.
What I find revealing is that the promoters of the pipeline have been fanatically opposed to suggestions to reroute the pipeline around Nebraska. If you want a jobs program, then wouldn't a longer pipeline mean more construction jobs?
Posted by: 2slugbaits at November 9, 2011 02:31 PM
If we are talking about building pipelines and creating jobs, how about constructing a water pipeline from the state of Washington to the California Central Valley. Last time I heard unemployment was high and needed crops were not being grown. It seems to me this could be at least a “two-for” immediate jobs and supply of food. No worries about contaminating spills.
Posted by: AS at November 9, 2011 03:46 PM
While there are arguments for the pipeline, there are arguments against it as well. That means it isn't really "shovel-ready". It certainly isn't at the state level because each state is not ready and some object.
A true shovel-ready project is something like a bridge or road repair. Of course, I read an objection to that too, to the effect that needing to repair the New Jersey Turnpike so often is proof the government can't do anything. I consider that kind of objection blatant stupidity but I can't so dismiss the objections to the pipeline.
Posted by: jonathan at November 9, 2011 03:51 PM
tj Rethink your example. The multiplier argument assumes that some fraction of E's income is saved and not spent. So in your example E would not employ U; E would save the $10. But if no one borrowed E's $10, then E might as well bury that $10 in that hole because it has leaked from the income flow. It is precisely because the government borrows E's $10 and uses it to employ U that the initial $10 is retained within the income flow. The multiplier does not create something out of nothing; it works by soaking up excess saving that no one else wants to borrow and would otherwise be permanently lost as a flow variable.
I think part of the problem is that people tend to frame the multiplier issue in terms of stock variables. The right way to understand the multiplier is how it works in the context of an income flow...and GDP is a flow variable.
Posted by: 2slugbaits at November 9, 2011 04:48 PM
Does Ono distinguish between helicopter drops of money, which would stimulate but not affect incentives, versus disincentives like paying people not to work? Clearly unemployment insurance has a negative multiplier almost by design, and certainly in reality. That is unless we add in the value of leisure to GNP...in which case it might just be 0, since you tax a dollar to pay someone to consume leisure.
I recall that in my macro classes government spending was considered to be "buying airplanes and throwing them in the ocean".
Posted by: pete at November 9, 2011 05:39 PM
The issue is that this is seen to be a key battleground issue on the one hand and an election year issue on the other hand.
Posted by: RB at November 9, 2011 06:40 PM
@tj: I think you answered your own question except that you inverted the timing. The government borrows $10 from E's savings and gives it to U, then a year later it levies $11 from E and then pays back E $11. E is down $10 (plus one filled hole), the government has a balanced budget, and U got to work, earned $10.
It does have the flavor of a forced assistance program from E to U, the difference is that by exchanging the $10 through useless work instead of a direct transfer it gets measured as GDP (even if valueless). The hope is that the work is used in something useful, not just filling holes, and I would add that U still needs to spend his $10.
On a related point: if your concern is that the interest makes the whole borrowing too expensive (through crowding out of E's consumption in your example through taxation), the zero-lower-bound people will remind you that current borrowing costs are essentially zero.
Posted by: gg at November 9, 2011 07:48 PM
Interesting point on the multiplier. On the pipeline, what am I missing? Canada produces lots of oil in the middle of nowhere. Because they can't get their oil to world markets, they are forced to sell it cheaply to nearby Americans. Instead, the Canadians would like to build a pipeline through America so that they (and incidentally some oil producers in North Dakota, but that is very much incidental and not the reason the pipeline is being built) can sell their oil to international buyers at higher prices. This will require the exercise of rights of eminent domain in order to take private property for the public good, and will also have various negative environmental consequences. Once this pipeline is in operation, the many Americans who now get cheap oil from Canada will have to pay higher world prices--and the costs of that will far outweigh any benefits to oil producers in North Dakota. Of course, economists often claim that distribution issues are beneath them, but the issue is not being decided by the Republic of Economists but rather by the State Department. They are also obliged to take our relations with Canada into account, but in the absence of that consideration it is hard to see how the proposed pipeline benefits the public interest from an American perspective.
Posted by: Jim at November 9, 2011 07:58 PM
You forgot two risks:
1) Security risk due to swaping our cash for oil from people who fund fanatics that fly planes into our buidlings (I'm guessing you don't live in NYC)
2) Concentration of wealth in petroleum *countries* who have done no favors for democracy (you know, like letting women vote or drive cars)
Here's a hint, neither of the above apply to Canada. But if we don't let Keystone be built, more of our money will be going to coutries to which these two risks do apply.
If we don't buy the tar sands oil, the Chinese will. This will increase the price of oil in North America and worsen our trade deficit. That oil will be carried to China in ships that burn wickedly polluting bunker fuel. Rather than coming by very energy efficient pipelines, our oil will be shipped halfway around the world in ships that burn wickedly polluting bunker fuel and which sometimes run aground and spill millions and millions of gallons of oil.
I'll take my oil from the pipeline thank you.
Posted by: Equityval at November 9, 2011 08:42 PM
Cheap WTI-priced oil has made its way around the economy, with refiners in some parts of the country acquiring oil at $10 or more below Brent. However, all this margin is being captured by the refiners. It is not showing up in gasoline prices in any region of the country (at the PADD level). Gasoline prices have continued to track Brent.
Thus, the Keystone pipeline would not be anticipated to affect gasoline prices. Instead, it will increase WTI and related prices, with higher payments to oil producers from Alberta to Oklahoma, on the one hand, and lower margins to US refiners, on the other. This gap already appears to be closing, with WTI closing the gap on Brent in recent weeks.
I would note WTI is at $98 this morning, outside our estimated comfort zone for the US economy.
Posted by: Steven Kopits at November 10, 2011 04:57 AM
All: Please forgive my delay in responding as I was tied up all day yesterday.
tj: The "balanced-budget multiplier" refers to the case when taxes are raised contemporaneously to pay for the current spending so as to keep the budget balanced. As 2slugbaits and others point out in response to your question, for the case you pose (deficit-financed spending), the question is how does E respond to the fact that his taxes next year will be $11 higher? Someone who believes in Ricardian equivalence would answer, E saves exactly $10, buying the newly issued government bond, whose principal and interest he then uses next year to meet his tax bill. In this case, the multiplier would be zero, because the increase $10 government spending is exactly offset by a decrease of $10 in consumption spending. A more Keynesian formulation, by contrast, would posit that E reduces spending by only c times $10, where c is the Keynesian marginal propensity to consume. We then have: (1) $10 of increased government spending on a nice, new filled-in hole; (2) c times $10 of new consumption spending (and possible multiplier additions thereto) by the newly employed U who is now happily spending a fraction c of his new income, and c times $10 of less consumption spending (and identical multiplier subtractions therefrom) by the tax-worried E. The net sum of these is 10 + 10c - 10c = 10. So that's a way one could get a Keynesian multiplier of exactly 1 for your scenario.
Johannes: Perhaps I will have a chance to discuss Europe this weekend. In the mean time, check out Tim Duy.
2slugbaits: The U.S. has over a quarter million miles of oil and gas pipelines already. Should we shut those down so terrorists aren't tempted? And how many millions of dollars of aquifer damage can you identify those as having caused? And as for a jobs program, my whole point (and I realize this is different from how Keynesians make the argument) is that it's a better jobs program the more real value is being created per worker. Digging the trenches with spoons would of course mean more "jobs", but my position is that a digging-with-spoons jobs program would be a way to impoverish rather than enrich America.
jim: Sorry, it makes no sense to simultaneously ship Canadian oil from Port Arthur to international buyers and at the same time import oil to Port Arthur from Saudi Arabia, Venezuela, Nigeria, and Iraq. The U.S. is currently importing 3.4 million barrels per day from those 4 countries. Getting an extra 500,000/day from Canada and North Dakota via the Keystone is not going to turn the U.S. into a net oil exporter.
Posted by: JDH at November 10, 2011 09:06 AM
Page 11 of this link has an image with all the refined petroleum pipelines crossing the U.S. http://www.pipeline101.com/reports/Notes.pdf
Regarding the multiplier, thanks JDH and 2lsugs for clarifying. I don't want to divert the thread away from the main point, so I will drop it.
Posted by: tj at November 10, 2011 09:32 AM
This is the fundamental problem...
“The world is locking itself into an unsustainable energy future which would have far-reaching consequences, IEA warns in its latest World Energy Outlook” (link)
"In the New Policies Scenario, cumulative CO2 emissions over the next 25 years amount to three-quarters of the total from the past 110 years, leading to a long-term average temperature rise of 3.5°C. China's per-capita emissions match the OECD average in 2035. Were the new policies not implemented, we are on an even more dangerous track, to an increase of 6°C."
Posted by: Andrew Henry at November 10, 2011 10:33 AM
Equityval, you don't address global warming-perhaps you think it's all some liberal plot? But here's what you're not appreciating: by enhancing the infrastructure of the oil economy, we delay the switch to a renewable economy. This will just perpetuate world reliance on Middle East oil, even if America gets most of its imports from Canada, Mexico, Venezuela, etc. We should be working on mass transit, electric cars and rooftop solar to power them (every house an energy producer!) instead of making new concessions for the oil economy.
As for Canada selling oil to China (the boogeyman used to be Japan), Canada will do what it will do, although there's opposition to it in Canada too. Regardless, as Steven Kopits points out, oil prices are set in international markets, and marginal supply from the tar sands won't make much difference in energy prices in the US. Reliance on Canadian oil is still reliance on oil, and reliance on oil means world reliance on Arabian, Iraqi, Iranian, and Russian oil, among others.
Posted by: VinceInSeattle at November 10, 2011 11:36 AM
JDH I don't have a problem with most oil pipelines, so no, I am not in favor of shutting down a quarter million miles of existing pipelines. But I do have a problem...actually, two problems with the Keystone pipeline. First, te aquifer that it transverses has an unusually high water table and is the major source of underground water for people living between the Mississippi River and the Rocky Mountains. This would be an especially tempting target for terrorists because the economic costs would be substantial and the operational risk would be low. So my objection here is really with the path of the pipeline. Rerouting the pipeline would address my first concern. Yes, it would cost more to reroute, but doing so would reduce the cost of negative externalities...and of course it would also increase employment even more. What the pipeline's promoters want to do is (surprise!!!) privatize profits and socialize risks. And the industry's track record on cleaning up its messes isn't too good. Just today we heard about two major spills in Nigeria that damaged fresh water supplies, and not surprisingly the oil industry is refusing to man-up and accept responsibility. And if the spill was due to a terrorist act, then the taxpayer will get the bill because that would be an act of God...or perhaps Anti-God. You get the picture. My second objection is that building the pipeline increases the chances that we'll be stuck with a white elephant that ends up distorting future policies. Twenty years from now the pipeline's owners and petroleum industry lobbyists will create all kinds of excuses as to why we need to expand oil exploration in order to keep the pipeline profitable. It's the old sunk cost argument to protect an asset. President Sarah Palin won't be able to say no. In fact, not too long ago Gov. Sarah Palin was making similar arguments about the Alaska pipeline...we needed to expand Alaskan oil exploration or else the pipeline would lose value. It's a bad idea to let a sunk cost guide public policy. We already have enough Republican elephants in Congress, we don't need another white elephant to further confound policy issues.
Posted by: 2slugbaits at November 10, 2011 01:01 PM
You said: "...the estimated 20,000 Americans who would work on construction of the pipeline...."
The study said: "...[the] project is expected to directly create more than 20,000 high-wage manufacturing jobs and construction jobs in 2011-2012 across the U.S. ...."
Why give the greenies a red herring to obscure your point?
Posted by: Neely at December 16, 2011 09:42 AM
According to T. Boone Pickens, there are 50 pipelines crossing Nebraska (many of them decades old) and 25,000 miles of pipeline that cross the Ogalalla acquifer. Now we want to build a brand-new pipeline with the latest technology and is the safest industry can offer (forget the red herring of what goes on in Nigeria) and the greens are all up in arms! Those 25,000 miles of pipeline have not caused any problems for decades, but we're supposed to import more oil via tankers instead. This is childish.
Btw, the comment about just go ahead and build a new refinery in Iowa is even more naive. The last refinery was built in this country 40 years ago. A new one (if it could get permitted) would cost billions. Why can't people think these things through?
Posted by: Bunny Slippers at December 16, 2011 10:39 AM
Obama is wasting trillions of useless stimulus that does nothing for jobs, but saddles us with debt. But here we can get thousands of jobs, plus future economic output increases, plus jobs created because energy is cheaper, and energy independence. But better yet we dont even have to pay for it, since Canada is willing to pay the bill! And all Obama has to do is get out of the way, and let the Canadians pay us to create jobs, and give us cheaper oil. So what does he do??? This decision illustrates more than any other, maybe even more than Solyndra, the economic idiocy, and environmentalist cronyism, of the Obama administration.
Posted by: richard40 at December 16, 2011 06:28 PM
Republicans should pound Obama on this.
It's a political loser for him. And we sure as heck do need the jobs.
Posted by: John at December 16, 2011 07:21 PM
"The big picture is, we have grown to become very reliant on oil, and like many addicts, its hard to quit."
We've grown to be very reliant on food too. I'm trying to give that up as well, but not having much luck. The windmill on my tractor isn't getting the job done and the crops aren't what they used to be.
Posted by: willis at December 17, 2011 05:47 AM