February 12, 2009

Deep Impact

Quico says: Everyone loves 90s disaster movies, but before we get to that, a little Thought Experiment.

(Stop me if you've heard this one before...and if you've spent more than 20 minutes in a microeconomics course, you have heard this one before.)

Imagine your house is a mess: laundry hasn't been done in ages, the kitchen sink is overflowing with dirty dishes, mop-to-floor contact hasn't been achieved in like forever. You really need to clean. But you're lazy. You value having a neat house, but you also value laying around watching TV.

Now, what is the optimal amount of time you should spend cleaning your house today?

The scientifically rigorous answer is, it depends.

When you're cleaning your house, what you're doing is trading off enjoyment in the present against enjoyment in the future. You're saying that a little bit more enjoyment in the future (from having a clean house) is worth a little less enjoyment in the present (cleaning sucks.)

Your optimal trade off point will depend on how much you value enjoyment in the future relative to enjoyment in the present. If you're more future oriented, you'll invest more time cleaning. If you're more present oriented, you'll consume more TV.

Now, imagine that as you're laying around watching TV, a NewsFlash comes on saying NASA has just discovered an asteroid hurling towards the earth, big enough to kill everyone instantly, and due to hit in 20 minutes.

What is the optimal amount of time you'll spend doing dishes now?

The point of this slightly hackneyed little parable is that the solution to any intertemporal optimization problem hinges critically on one thing: your time horizon. Economists call it your discount rate, basically: how much consumption you're willing to forgo now in exchange for how much in the future. The shorter your time horizon, the more heavily you discount future consumption's value relative to current consuption. In fact, those are just two ways of saying the same thing.

Now, the thing to grasp is that the Consumption/Investment tradeoff facing the Venezuelan petrostate is, in principle, not that different from the dilemma facing our lazy housekeeper. The fundamental decision to be made about any given dollar that reaches PDVSA, Venezuela's public oil comglomerate, is whether they invest it to yield more production capacity for the future or they consume it right now. The balance between the two is at the heart of petrostate politics.

In fact, much of the struggle over control of PDVSA from 1999 to 2003 can be rendered in the language of intertemporal optimization, as a violent disagreement over the appropriate discount rate for the oil industry.

In 1975, when PDVSA was originally nationalized, everyone could see that elected governments were hard-wired to have relatively short time-horizons, and that if they could control PDVSA directly, they would tend to choose more consumption and less investment than is optimal over the long run. The experience of Pemex, Mexico's bloated, dysfunctional state-owned oil company, was widely seen as a cautionary tale of what could happen when politicians' hands are allowed to get too close to the honey pot.

That's why PDVSA was consciously designed to be functionally independent of the government, and overseen by a weak regulatory agency (MEM). If, before 2003, PDVSA invested more and generated a smaller revenue stream than the government wanted, that's because it was designed to do precisely that: optimize the revenue stream over a longer time horizon than any one government would prefer.

Theoretically, at least, chavismo had a point: it's possible to be too farsighted as well as not farsighted enough. In the housekeeping parallel, you might imagine someone with Obsessive Compulsive Disorder, cleaning house all day, every day and never stopping to enjoy their spotless house. In effect, chavismo's claim was that PDVSA had developed a case of institutional OCD, adopting a time horizon so long it had essentially become more concerned with endlessly expanding capacity than with producing a usable revenue stream for the government.

Of course, there's always something a bit artificial about translating Cháveztalk from the ranting of a lunatic into economically meaningful positions. You're forced to reverse engineer his train of thought, working backwards as you ask yourself: "if, instead of Chávez, it had been a sane person making this argument, how might the argument have come out?" I think what would've come out is something like this:

As of 2002, PDVSA's discount rate only made sense from the point of view of middle class people. Long-termism is a privilege that only folks who are reasonably certain they are going to get three meals a day can afford. But the "right" discount rate for the poor is higher than the "right" discount rate for the middle class: when you're hungry now you need to eat now. So there are good reasons for the poor to have a shorter time horizon than the middle class, and a responsive, democratic government needs to be attuned to the needs and preferences of the poor and reflect that shorter horizon in the management of the oil industry.

There's nothing wrong or contradictory about that argument. To retort something like, "but that only makes poor people poorer in the long run!" is to miss the point spectacularly: the whole point of the exercise was to refocus PDVSA on the short run, to make it more democratic in the sense of adopting a time horizon more congruent with the majority's.

Another way to get at this dynamic is to borrow the language of the "principal-agent problem". They may not have known the academic terminology, but when it came to the Old PDVSA, chavismo had no trouble understanding that when a "principal" - the Venezuelan people - delegates authority over a common resource to an "agent" - the old PDVSA - getting the agent to act the way the principal wants will always be a problem. How can the Venezuelan people exercise their authority to make sure their agent really is acting on their behalf and doesn't gradually start to focus on its own interests, instead?

This, when you overlook all the scatology, is the "sane chavista" rap against the old PDVSA: the agent forgot all about the principal. The Old PDVSA was all about imposing a time horizon on the oil industry that didn't match the principal's own, much shorter, time horizon.

Whether it makes sense to force the agent to be just as short-sighted as the principal is another question altogether. Personally, I think that's crazy and counterproductive, but for now, all we have to do is concede that this re-constructed, "sane chavista" argument is not internally inconsistent or self-evidently wrong.

But of course, there's a catch: good as chavismo may have been at noticing the principal-agent problem with regard to the old PDVSA, it catastrophically fails to apply the same insights to itself. Now that the "agent" that effectively controls the oil industry is Chávez himself rather than a technocratic elite, the possibility that he might strike off and start working in his own interest, rather than the people's interest, never seems to occur to them.

Partly, the inability to see this is ideological: years of "Chávez es el pueblo" propaganda have totally blinded some to the possibility that Chávez could have one set of interests and the people another.

But the agent-control problem is, if anything, far worse now than it was before. The whole strategic stance of PDVSA, in terms of the investment/consumption trade off, has been molded to the needs, indeed the whims, of a single human being. If the old PDVSA's time horizons were, arguably, longer than the principal's, the Chávez controlled oil industry's time horizons are insanely, suicidally shorter.

In effect, Chávez is treating the referendum just like our lazy housekeeper would treat the Deep Impact asteroid! Just as there'd be no sense in cleaning house 20 minutes before an asteroid destroys the planet, Chávez seems to have decided there'd be no point in keeping the oil industry able to function if he can't have indefinite re-election. And so, he's focusing all of the oil industry's resources on the economic policy equivalent of the next 20 minutes. That his time horizon doesn't match the pueblo's is neither here nor there because the principal has lost all control over its agent. Chávez se olvidó del pueblo!

It bears noting that, in acting this way, Chávez is showing with some precision why the 1975 nationalizers were wise to keep PDVSA at some distance from its political masters. The thinking back then was that, if allowed, politicians were sure to try to manipulate the oil industry's revenue stream to try to swing elections. The principal-agent problem was very much front and center in policy-makers' minds back then; everyone understood that there's no easier way to wreck the oil industry than injecting it into the electoral game. Their warnings, our reality.

Just as the most outrageous kinds of delinquent housekeeping - say, tossing old chicken bones on the carpet instead of taking the trouble to walk them to the garbage bin - become "rational" if an asteroid is about to vaporize the planet, the absolutely craziest ways of managing the oil industry - failing to pay the people who run the drilling rigs, for instance - become "rational" if you decide all you care about is winning the next election.

Now, this is a conceptual post, so it's not really the place to go into a long litany of complaints about what's been happening inside the oil industry. "Tossing chicken bones on the carpet" is about the size of it, though: a form of extreme, kamikaze short-termism that completely boggles the mind.

My sense is that PDVSA is now approaching a series of critical thresholds where the sticking plaster that's been holding it together is bound to give way. The company is just too far behind on too many payments to too many critical service providers. We're not talking catering or la Fuller here, we're talking core payments to core contractors without whose help PDVSA cannot lift oil out of the ground. It's reached such an extreme that what's news - man-bites-dog-wise - is no longer when PDVSA doesn't pay its bills, but when it does!

how far to the left we are on that Investment/Consumption trade-off chart.

Distracted by yet another pointless vote, nobody's paying attention to the Tsunami of Piss brewing in the oil fields - truly the leading edge of Hurricane Feces. Because, lets be clear here, a petrostate can just about muddle through when oil prices are merely low, but it absolutely can't function if it can't lift it to begin with.

My sense is that this, and not the referendum result, is going to be the story we'll remember about 2009-2010. In the big scheme of things, the referendum will be remembered as an asterisk, not an asteroid: "that crazy campaign when Chávez spent all of PDVSA's money trying to amend the constitution right before the Great Oil Collapse of 2009."