Friday, 3 June 2016

Experiments in political science and the Cartwright critique

Over the course of the last couple of years, the political science discipline has twice hit the headlines for scandals linked to "field experiments." Maybe this isn't surprising: such experiments have become incredibly fashionable. Success in an academically fashionable endeavour can bring large rewards, and it's certainly plausible this has created incentives making fraud or poor judgement more likely.

To the extent that bad behaviour reflects incentives, one can always try to to police against it more vigorously. But changing incentives may be more effective. In this spirit, I'd like to encourage political scientists to stop being so damned excited by experiments and offering such big reputational rewards for them!

As a reason to calm down, consider some arguments (or great lecture version) from the brilliant philosopher of science Nancy Cartwright. Experiments (of the presently fashionable sort) rely on the logic of randomly assigning groups to "treatment" and "no treatment," so that any difference in outcome between the two groups can confidently be ascribed to the treatment. Yay, science!

Cartwright's core insight is that what such experiments can establish is only the role of a particular link in what might be a complicated, and highly context-dependent, causal chain. To build on an example she uses: Suppose you had a set of toilets, and assigned each of them randomly to have the lever attached to its side pressed or not. On completion of the experiment, you could confidently assert the relationship "lever pressing leads to water release." This formulation, though, would entirely obscure the point that these levers only release water because they are part of a mechanism to open a chamber supplied with water by pipes, etc.

Thus, if you went off to deploy your exciting new experimental result to solve California's drought by having everyone push the levers attached to the sides of their toasters, you'd be disappointed by the results. As Cartwright says (p.102), "Once stated this is an obvious and familiar point," but nonetheless one too often overlooked. This she effectively demonstrates with empirical examples of the disappointing performance of 'experimentally validated' policy interventions in new contexts.

So why is it that experimentalists are overlooking this obvious and highly consequential point?  [I'm not going to defend in detail the claim that they are, but will assert that the discussions about 'external validity' from experiment evangelists are not nearly searching enough.]  Let's use a little notation to make the argument more compact: the causes of an outcome O of an experiment are the experimental intervention I (such as lever pushing) + the rest of the mechanism M.  

So the question becomes, why the emphasis on I rather than M?

  • A lot of the methodological backdrop for political science experiments is drawn from experimental medical trials. In these, the common features of human organisms are regarded as similar enough that M will function in the same way. This assumption can be criticised even in a medical context, but for social scientists the issue is orders of magnitude more significant. 
  • Unlike pressing a lever, field experiments in political science are difficult to organise and often quite expensive. After all that effort to demonstrate the role of I, it's hard to remember that the M is important too.
  • I will often have been chosen precisely because it's the aspect of a broader mechanism that is easiest to manipulate. If the effects of M cannot be assessed via randomised controlled trials, then experiment absolutists will deny the possibility of making any meaningful claims about those effects. They haven't faced up to the fact that this means that they will never have any basis sanctioned by their own methodological precepts to assert that the results of one experiment have any generalisable implications whatsoever.

Whatever its origins, the mania for measuring the effects of interventions, and the corresponding neglect of the causal import of the context of these interventions, strikes me as very bad thing for many reasons, on which I hope to expand on another occasion.

PS: Cartwright's is not the only impressive critique of experimentalism on offer; I especially recommend Dawn Teele's edited volume. But so far, the critiques don't seem to have made much of a dent in the popularity of field experiments. Political science as a discipline seems to have an almost congenital need to affirm its 'scientific status'. But we should be suspicious of anything we need so much. How much did that ring really help Gollum?

Thursday, 10 December 2015

Puffing the magic Draghi

Mario Draghi had a rough time last week.  The extension of QE he announced disappointed markets, who were apparently expecting him to exceed expectations.  (Sounds oxymoronic to me, but I'm just a political scientist, not clever like a bond trader.) The upshot was a sharp rise in the value of the euro, which is a problem for a Eurozone demand model heavily reliant on exports.

Maybe Mario will be cheered up after Politico published a puff piece about him today. I wasn't; the article veered from the uninformative to actively misleading, reporting inter alia:  
Draghi ... [took] bold steps that enabled him to save the euro. Now that the danger of a disintegration of the eurozone has abated, the ECB president is embarking on an even tougher political task: to convince Europe’s governments that they must do their part of the heavy lifting to take the continent out of the slump. ... 
He is prodding EU governments to boost spending to put the European recovery back on a path to growth. ... 
For the moment, Draghi is happy to let Coeuré [ECB board member] and Praet [ECB chief economist] push the message that Germany in particular needs to splash out more on public infrastructure, to address what one ECB executive board member called an “absurd situation” where spending is so subdued that the fiscal deficit of the eurozone is much lower than the 3 percent allowed by the Stability and Growth Pact.
This a reiteration the myth that Draghi is an active supporter of a demand-stimulus approach to resolving Europe's growth crisis. However, all of the arguments I made against this myth more than a year ago remain valid. Above all, Draghi's shown not the slightest inclination to use his ample sources of political leverage to push for increased spending stimulus.  Nor has he repudiated his key role (see pp.34-38) in pushing for an austerity-led reaction to the Eurozone bond crisis, continuing to imply that no other choice was possible. As he recently put it, "don’t blame the fire damage on the fire brigade."

Perhaps, though, Draghi is beating the drums for demand stimulus behind the scenes? He is, after all, somewhat constrained as the public face of the ECB leadership.  Consider this exchange from the latest ECB press conference
Question: Last year in Jackson Hole, you advocated for a policy mix with monetary policy reforms, investment and fiscal policy, and today you have emphasised the role of fiscal policy. Do you miss more fiscal stimulus in countries with margin, like Germany, for example, and do you consider that the neutral fiscal stance that the European Commission is advocating for the eurozone as a whole is adequate now, in a sort of liquidity trap?
Draghi: We had a brief exchange on this issue, and our conclusion now is that, first of all, the first answer should be given by the Commission. The second point is that we'll continue reflecting on this, and we will have a view on what is the degree of appropriateness of the fiscal stance; whether we have a view about the aggregate fiscal stance; what is the degree of compliance with existing rules; whether the flexibility which has been exercised before all the terrible happenings of this year – so before the recent terrorist attacks, but also before the refugees events – whether that flexibility would be justified. So there are lots of factors in play altogether. How do we assess the fiscal stance today given the presence of the previous flexibility, the refugees, the need for security of the euro area? It's a very complicated question, so we are going to reflect on that.
One might read this as a sign that political conflict at the top of the ECB is limiting what Draghi can do by way of advocating fiscal sanity.  However:

  • Draghi has more than once found ways to move beyond the consensus of the bank's leadership, and there's no evidence he's trying to do so on this issue.
  • There is likewise no evidence that he personally views austerity as a crucial component of the growth catastrophe.  Asked at a November Europarliament meeting about what was needed to promote growth, Draghi had literally not a single word to say about government spending (see p.11).

Monday, 9 February 2015

Can Greece escape the ultimatum game? (And no, Draghi's not trying to help them do it)

At the start of a crucial week for Syriza's effort to negotiate a revision to the disastrous current arrangements between Greece and the Troika, I thought it would be helpful to try to describe the bargaining situation in a systematic way using game theory a very simple diagram (yes, it is game theory, but it's very straightforward).  Here it is:

The European Commission (EC) needs to decide whether or not they will propose a compromise or insist on the Troika's current terms.  Greece then needs to accept or reject the offer.  The rectangles at the far right show possible outcomes--if Greece rejects whatever the EC proposes, it leaves the Euro (Grexit).  Let's make some assumptions about how the parties rate the outcomes: Greece prefers a compromise to the Troika's terms, and both to Grexit.  The EC prefers the Troika's terms to compromise, and both to Grexit.

If this is an accurate depiction of the situation, the EC gets what it wants--it can face the Greece with the choice between taking the Troika's terms and Grexit, and Greece will choose the Troika.  It's an "ultimatum game;" Greece has to take whatever the EC offers, because otherwise it's faced with the catastrophe of Grexit.

So how can the Greek government change the game?

  1. Announce that it prefers Grexit to Troika terms. Then (bottom right in the diagram) Greece would be expected to choose reject, giving the Grexit outcome. Since the EC prefers compromise to Grexit, it would offer a compromise. So why doesn't Greece just do this?  
    • To announce this might set off an even worse banking panic than Greece is currently experiencing.
    • It only works if the EC reliably prefers compromise to Grexit.
    • It might not be believed
  2. Disable its capacity to accept Troika terms, or at least create uncertainty about whether it would be able to accept Troika terms (meaning that Troika intransigence would lead to Grexit)
    • This is one way to interpret Tsipras' "defiant" speech yesterday, in which he promised very publicly and very vigorously that Greece will not accept a continuation of the present Troika. (Political scientists just love talking about how "domestic audience costs" -- the costs of going back on a promise to a domestic constituency -- allow politicians to make threats on the international stage that would otherwise not be credible. Compare this discussion by Jacques Sapir.)
    • Again, this only works if the EC fears Grexit; Varoufakis is trying to make sure that they do
  3. Convince the EC leaders that compromise should be be preferred to Troika terms. So far this doesn't seem to be going very well.
  4. Find a way to create another possible outcome, with no compromise but also no Grexit
    • It's sometimes suggested that because Greece is presently running an obscenely large primary surplus (i.e., it's budget is heavily in the black before debt repayment is taken into account), it could just stop repayments, and abandon new borrowing.  However, since the Greek government can't print Euros, the sustainability of this path would depend on the cooperation of the ECB to keep the prospect of bank panics at bay.
So this week's negotiations will turn on whether the Greek government has managed credibly to cut off the possibility of accepting the Troika's terms and how much the rest of the Eurozone fears Grexit.  

PS: Was the ECB trying to help Greece by refusing to accept its bonds as collateral?

Last week, the ECB stopped accepting Greek government bonds as collateral for ECB loans. The immediate and obvious interpretation of this decision was that it continued the pattern (long version) of using the withholding of ECB emergency lending as a form of policy leverage. Paul Krugman pleaded unconvincingly that Draghi was too subtle for such a brutal display of strength, arguing that really this measure was meant to wake up Germany--a position the Greek government had little choice but to echo.  

One version of the claim that the ECB was not trying to intimidate Greece into accepting the Troika terms, proposed by Frances Coppola, rests on the idea was that weakening Greece strengthened its bargaining hand in a strategic context (as under 2, above).  But the argument doesn't work: A bargaining analysis offers no support for the idea that the ECB was trying to help Greece.  The ECB decision did nothing to make accepting the Troika conditions more difficult, or make Grexit relatively more attractive. Indeed, given that the ECB explicitly mentioned that it was the prospect of a failure in negotiations with the Troika over continued support for Greece that prompted the decision, it raised the benefits to Greece of a successful agreement. Moreover, to the extent that the decision signalled the likely attitude of the ECB toward supporting Greek banks in the absence of a successful agreement, it worked against option 4, as well. So to the extent that the ECB decision did reflect a bargaining logic (other things may have been at stake), it would make sense only as an effort to coerce Greece, not to help it. 

Tuesday, 16 December 2014

The Russian crisis and the Eurozone: some economic context

Pretty dramatic things are happening with the Russian rouble. Its dollar value is as of yesterday just about half of what it was on average in 2013.

What does this mean for the eurozone? Over the course of the first decade of this century, oil revenues and the real appreciation of the rouble made the Russian economy much larger in euro terms than it had been, and it correspondingly became a more important export market for eurozone countries. The share of Russia in the eurozone's exports tripled between 1999 and 2008; while it has declined slightly since, 4.7% of eurozone exports went to Russia in 2013. Given that even the eurozone's anaemic growth since the financial crisis is entirely attributable to export growth, this is not insignificant.

Source: Eurostat, OECD, EIA, own calculations. Brent prices and export/import figures nominal

In the chart, I've tried to give some indications of possible impacts of the rouble's fall by recalculating 2013 figures on Russian and eurozone GDP at yesterday's exchange rate, and giving a rough extrapolative estimate of how much exports to Russia might fall as a result of the declining purchasing power of Russian consumers in euro terms.

This estimate suggests a fall in sales to Russia of 0.5% of eurozone GDP, which would be highly significant, especially to an economy growing as slowly as the eurozone's is.  However, there are potential compensating factors (and potential further dangers):

Potential compensating factors

  • Russia's demand for imports may prove inelastic (consumers may not scale back purchases proportionately to the rouble's fall).
  • The eurozone will be spending less on importing oil, improving its trade balance. If consumers and businesses spend and invest the money no longer going to oil, this will promote growth. 
Potential further dangers
  • In liquidity trap conditions, consumers and businesses may not spend and invest the savings from cheaper energy prices. Then falling energy prices would simply contribute to the severe risk of deflation in the eurozone.
  • There could potentially be serious consequences to the international financial system of Russian companies being unable to pay back dollar-denominated debt (see here for a relatively sanguine discussion of their payment prospects). Financial fragility means that relatively small events can have a big impact. 

Thursday, 6 November 2014

Why Mario Draghi's ECB colleagues just kneecapped his credibility: the politics of market tripwires

Something is afoot at the European Central Bank.  People at the very top of the institution--leaders of the central banks of the member nations, and members of the small Governing Council responsible for key decisions--have apparently co-ordinated to anonymously tell Reuters just how very much they dislike Draghi's leadership style.  He even looks at his mobile phones--all three of them--when they're trying to say important things to him! 

I think it likely that these public complaints had a very specific goal--and it wasn't to get Mario to put his screens away.  Instead, they are intended to weaken Draghi's influence over ECB policy and strengthen the influence of other members of the ECB Governing Council. To do this, it was necessary to weaken the credence markets give to Draghi's statements, and that the ECB insiders shaped their remarks accordingly--or so I argue in this post.

The politics of market tripwires

To understand the conflict between Draghi and his colleagues, one needs to bear in mind that it is playing out on the backdrop of the intense attention financial markets pay to the pronouncements of central bankers.  

Financial markets, as is well-known (read Keynes on this) are characterised by self-fulfilling prophecies. If market participants believe an asset will fall in value, they will convert prediction into fact by selling it and driving down the price. Market participants thus have an understandable terror of being the last to sense a shift in the collective prophecy (more prosaically known as "market expectations"), unable to sell before the price has fallen or buy before it has risen. 

Thus, investors are especially sensitive to what I'll call "market tripwires" -- events expected to cause a general shift in market expectations. When one of these tripwires is triggered, investors rush to react as quickly as possible, in some circumstances creating a panic. An example of a market tripwire familiar from the financial pages is the earnings forecasts of stock market firms--whether these are met, exceeded, or undershot can set off large shifts in prices. 

Sometimes, political actors create market tripwires in order to impose costs or constraints as a tool of influence. For example, the IMF's Michael Mussa accused Argentina's Finance Minister Domingo Cavallo of doing just this to pressure the IMF into extending more help as Argentina fought to stave off devaluation of the peso in the summer of 2001 (I assume the story is true, but don't know for certain; for present purposes it's enough that it could be true):
Through leaks to the local press, the Argentine government circulated the story that the Fund … would augment [a planned] disbursement with an addition of about $8 billion. Financial market reacted positively to this news, and the bank runs slowed. ...The suggested augmentation of Fund support was announced without consultations with the rest of the Argentine government. ... There were no prior consultations with the Fund, nor any prior indication of support from the Fund for a substantial augmentation of its lending. Indeed, Cavallo's tactic was to force the Fund to augment its lending by creating a fait accompli. Financial markets and Argentine citizens reacted favorably to the announcement of augmented Fund support. If they were disappointed that this support was not forthcoming, the Fund (and the international community more broadly) would be responsible for the consequences. [Mussa, Michael. 2002. Argentina and the Fund: From Triumph to Tragedy. Washington, DC: Institute for International Economics, p. 41-42]
In our terms, Cavallo tried to create a market tripwire.  He hoped to turn the IMF's failure to provide additional support into a signal for market panic, betting that the prospect of the panic would cause the IMF to agree to his demands.  

Draghi's tripwires

More than once, Draghi has used his status as the ECB's main spokesperson to do something very similar--make public announcements about policy, shifting market expectations, and implicitly (or explicitly, for all I know) inviting his ECB colleagues to contemplate the consequences of failing to carry out the policy.

A highly consequential example was Draghi's famous "whatever it takes" statement in July 2012, inserted into a prepared speech at the last minute.
Just as shocked were the ECB boss's aides and his colleagues on the bank's policymaking Governing Council, none of whom knew Draghi would make such a sweeping promise. "Nobody knew this was going to happen. Nobody," one senior ECB official said of the speech.
The tactic worked. Draghi's statement had a huge immediate impact in alleviating panic on sovereign bond markets, creating a market tripwire: failure to agree on an official lender-of-last-resort programme would reignite the panic, almost certainly in worse form.  The well-reported narratives of the ensuing hard bargaining that concluded with the announcement of OMT demonstrate that Draghi's statement was made well before he could be sure that he could get support for the sort of programme he wanted. 

This week's anonymous attacks were motivated by an effort to prevent Draghi from doing it again. This time, the issue on the table is not bond market panic, but the threat of deflation, brought on by weak growth prospects and the contraction of ECB lending as banks pay back earlier ECB loans without taking on new ones. To deal with this contraction would necessarily involve the buying Eurozone sovereign bonds (imprecisely known as Quantitative Easing, or QE), which German members of the ECB leadership oppose.

In recent months, Draghi has clearly been trying to encourage the market to believe that this sovereign bond buying will happen, creating a market tripwire that he can use as leverage to make the program happen. Examples are:

  • Draghi's warning in his Jackson Hole speech--like "whatever it takes," inserted at the last minute beyond the control of the rest of the ECB leadership--that deflationary expectations were spreading and a promise that the ECB would "use all the available instruments" to try to fight them.
  • statement in early September that the ECB would try to expand its balance sheet to the levels of early 2012 (ie, reverse the contraction of its lending) 
Members of the ECB Council are perfectly aware of what this manipulation of market expectations is intended to accomplish.  According to one of Reuters' interviewees, the balance sheet statement,
"...created exactly the expectations we wanted to avoid," an ECB insider said. "Now everything we do is measured against the aim of increasing the balance sheet by a trillion (euros)... He created a rod for our own backs." 
You say that like it's a bad thing. Draghi wanted precisely that rod as a tool of policy influence. 

No, don't believe him

Of course, market tripwires only work as a tool of policy influence if markets believe the statements intended to set them. Cavallo would have gained no leverage over the IMF if the announcement of an impending expansion of support had not slowed bank runs. "Whatever it takes" would not have helped the passage of OMT without its calming effects on the markets.

So to disable the tripwire tactic, the German members of the ECB leadership Reuters' anonymous insiders set out to tell markets that Draghi is not to believed.  
"We specifically agreed at the meeting... not to put any numbers on the table," [said] one central banker. "Draghi's reference to the balance sheet of 2012 irritated a lot of colleagues. So he has had to backtrack a bit ... to compensate."
In other words, just because Draghi promises something, it's not necessarily going to happen. It is precisely Draghi's ECB opponents need to make this point, I believe, that explains why these complaints were made in public rather than in private. In July 2012, Draghi said:
Within our mandate, the ECB is ready to do whatever it takes to preserve the euro. And believe me, it will be enough.
What his ECB colleagues are saying is: no, don't believe him. Wait to hear from us.

In particular, as becomes clear in the final paragraphs of the Reuters piece, don't believe Draghi if he hints that QE is coming. You can believe in QE if and when it is announced as an official policy. The insiders claim that without a consensus for QE in the ECB--and it's currently opposed by "at least seven and possibly as many as 10 of the 24 council members"--it can't happen because it's too politically divisive. (On this matter, one can only hope that the ECB majority will listen to Paul de Grauwe. A long as the ECB has unaccountable power, it should be used for good and not only for evil.) 

Kneecapping Draghi's credibility will make it harder for him to use his public announcements as a tool of unilateral policy-making, but it has costs, too. If the anonymous insiders succeed in turning Draghi into "the boy who cried QE," he will be unable to reassure the markets at moments when they need it most. For them that may be a feature rather than a bug of their strategy, but the rest of us should fear the dismantling of whatever capacity there is to contain the dangerous fickleness intrinsic to financial markets, and the destruction of whatever limited hope there is that deflation can be avoided without reversing austerity.

Wednesday, 22 October 2014

Governing by panic

Readers of this blog may be interested in my new paper, "Governing by Panic: The Politics of the Eurozone Crisis."  Here is the abstract:

The Eurozone’s reaction to the economic crisis beginning in late 2008 involved both efforts to mitigate the arbitrarily destructive effects of markets and vigorous pursuit of policies aimed at austerity and deflation. To explain this paradoxical outcome, this paper builds on Karl Polanyi’s account of how politics reached a similar deadlock in the 1930s. Polanyi argued that democratic impulses pushed for the protective response to malfunctioning markets. However, under the gold standard the prospect of currency panic afforded great political influence to bankers, who used it to push for austerity, deflationary policies, and the political marginalization of labor. Only with the achievement of this last would bankers and their political allies countenance surrendering the gold standard. The paper reconstructs Polanyi’s theory of “governing by panic” and uses it to explain the course of the Eurozone policy over three key episodes in the course of 2010-2012. The prospect of panic on sovereign debt markets served as a political weapon capable of limiting a protective response, wielded in this case by the European Central Bank (ECB). Committed to the neoliberal “Brussels-Frankfurt consensus,” the ECB used the threat of staying idle during panic episodes to push policies and institutional changes promoting austerity and deflation. Germany’s Ordoliberalism, and its weight in European affairs, contributed to the credibility of this threat. While in September 2012 the ECB did accept a lender-of-last-resort role for sovereign debt, it did so only after successfully promoting institutional changes that severely complicated any deviation from its preferred policies. 

Monday, 20 October 2014

When is a social democrat not a social democrat?

When he's Sigmar Gabriel, head of Germany's SPD and Minister for the Economy.  Here's Mr Gabriel in an interview with the Bild newspaper, defending the so-called "black null," the plan for a balanced budget:
Bild: There's a discussion in the SPD about whether or not new [government] debts ought to be incurred.  Is budget discipline social-democratic?   
Gabriel: Yes. Workers [Arbeitnehmer] want their taxes to be spent on social security, schools, or policy and not on interest payments to big banks for government debt. Only big banks earn money from high government debts. Government borrowing is antisocial.
There's a major problem with this argument: as Ambrose Evans-Pritchard noted, last week German government bond interest rates were "touching levels never seen before in any major European country in recorded history." In fact, correcting for inflation, Germany can literally borrow money interest-free:  August's inflation rate was 0.8% per year, and its ten-year bonds as of today yield 0.85%.  

Anyone who cannot find a way for the state productively to invest interest-free loans cannot be characterised as a social democrat (there's certainly plenty of low-hanging fruit).  For that matter, anyone who deliberately misleads workers about the costs of borrowing doesn't deserve the title either.