How to Build Independent Regulators (and keep them that way)
January 2, 2013 Leave a comment
From the Autumn 2012 Research Bulletin, by Chris Hanretty, Lecturer in Politics
Most markets require some form of regulation, and most countries keep that regulation away from government departments and within independent regulatory agencies. Designing those regulators so that they are genuinely independent from government is an important task. Chris Hanretty shows how some environmental and design features matter more than we think.
As a political scientist, I have a natural sympathy for politicians. When people make negative claims about them, I feel pressed to respond. Many people claim, and much of my work is based on the premise, that politicians make consistently bad decisions when it comes to regulated markets, particularly regulated markets which were previously state monopolies.
In particular, many argue that politicians are temporally inconsistent. In the bright new dawn of an incoming government, they seek out fresh market entrants, and encourage bold new infrastructure investments. But in the jaded gallop to the polls five years later, they put the squeeze on consumer prices, holding them to levels that won’t cover any form of capital maintenance, let alone fresh investment.
Such inconsistent preferences are a problem for everyone because, if left unchecked, they create uncertainty, and uncertainty is inimical to investment. The fault, of course, lies not in the stars, but in ourselves. We voters are the ones who get sniffy about higher prices, and tend to punish governments that let prices rise.
If inconsistent preferences are such a problem, what can politicians to do about them? The usual answer is that they must make some kind of credible commitment. Commitment devices are wonderful things. The most famous example is when Achilles bound himself to the mast of his ship in order to listen to the Sirens without becoming so enchanted by their music that he dashed the ship on to rocks. Less romantic commitment devices are found in many aspects of every-day life. I regularly make a credible commitment to waking up early by placing my alarm clock at some considerable distance from my bed.
For politicians, one way of ensuring a credible commitment to non-interference is to create an independent regulatory agency: an Ofcom, or an Ofgem, or the like, entrusted with promoting certain objectives that politicians (and the public) desire, and independent of politicians and thus capable of resisting subsequent attempts by politicians to go back on their word. The creation of such agencies represents a credible commitment to certain kinds of market outcome or process. The continuance of such agencies represents an on-going bind on new governments, one which they may secretly be glad of.
My work enters this happy story at two points. First, I look at the ways in which politicians can and have structured these regulatory agencies, and how they can alter their formal, or de jure independence. Second, I look at whether levels of de jure independence do really translate into actual, or de facto, independence. Is it possible for politicians to suborn the mechanisms they create to bind themselves, in the same way that many of us repeatedly hit the snooze button on our morning alarms?
The issue of designing for independence might seem straightforward. Make sure that the regulator has regular access to the funds it needs without budgets being subject to approval; make sure that ministers can’t dismiss the heads of these regulators, or that if they can they can only do so on grounds that would be assented to by all (ill health, malfeasance, non-performance of duties); and make sure that term lengths and framework agreements last long enough that regulators don’t pull their punches now in anticipation of the next set of appointments. There are, however, a number of awkward questions concerning institutional design. I’d like to take two issues in particular: appointments, and legislative scrutiny.
On appointments, many have assumed that, the more distant (in some sense) the appointments process is from the executive, the more independent the appointed body will be. On some indices of independence, regulators get scored higher if they are appointed by the legislature and the executive acting together, than if they are appointed by the legislature alone, and higher if they are appointed by the legislature alone than by cabinet, and higher if by cabinet than by a sole minister. A model law for an independent regulatory authority would therefore have all board members nominated by the executive and confirmed by parliament.
On legislative scrutiny, it seems obvious that certain kinds of legislative scrutiny can limit independent action by regulators. It’s not pleasant to go before a parliamentary select committee and be harangued; if regulators were regularly subject to such harangues they might decide to conform more to politicians’ temporally inconsistent preferences, vitiating their reason for being. But if independence requires an absence of legislative scrutiny, how are we to hold regulators accountable? How are we able to ensure that regulators are, say, taking social objectives seriously?
I wanted to check whether some of the assumptions that people had made about appointments provisions and legislative scrutiny were really sound. Together with Christel Koop (King’s College London), I went out and collected information on the governance of a large number of regulators worldwide, asking not just about appointments but a wide range of governance provisions. Our idea was to ask a large number of questions, and boil down the answers so that we were left with a core of provisions which separated high de jure independence regulators from low de jure independence regulators.
In the process of boiling down some of those answers, we found out that some provisions just didn’t fit with the rest – including provisions on appointment and legislative scrutiny (in technical terms, they didn’t map on to a latent factor of de jure independence). In other words: I can look at how a regulator is set up in terms of the tenure of its board members, the security of its funding, the difficulty of dismissal — but statistically this tells me nothing about the method by which board members are appointed or the frequency of legislative scrutiny.
Strictly speaking, the fact that these provisions don’t map on to a latent trait doesn’t mean that they are unrelated to independence. One could still insist that these provisions matter for independence, and it’s just happenstance that they don’t fit other criteria. To give a sporting example: it’s a bit like finding that managerial salaries are unrelated to other things which make clubs good (number of passes completed, shots on target), and insisting that an expensive manager should still be brought in to improve the club’s performance.
We didn’t opt for that interpretation, and we think that this finding has happy consequences. It is good news: good because it shows that we can have our cake and eat it. We can have independent regulators that still answer to Parliament, and we in the UK can keep our current systems of appointment. Whilst there might be reasons for moving away from ministerial appointment, a need for greater independence is not one of them.
Of course, sometimes commitments can be revoked or circumvented. A second part of my research looks at what happens after the creation of regulatory agencies with a given level of de jure independence. I wanted to understand whether greater de jure independence translated into greater de facto independence. In a second paper with Christel Koop, I used a proxy for de facto independence, based on two questions: did the chief executives of these regulatory agencies leave office very frequently? And was there a link between changes in government and changes of chief executive, such that the latter frequently followed the former? The more often chief executives changed, and the more often changes of government were followed by changes of chief executive, the lower the de facto independence of the regulator.
We gathered information on chief executive turnover for a large number of European regulators across several sectors. The country-by-country and sectoral patterns were unsurprising. Regulators of network industries had the highest levels of de facto independence. Regulators of softer concerns — environment, pharmaceuticals came out worse. Regulators of financial markets occupied an intermediate position. In terms of country patterns, there was a general north-south split, with Greece, Spain and Italy coming out at the bottom.
What we really wanted to know was whether these levels of de facto independence were related to de jure independence. If they weren’t, then it would mean that all the effort put into designing these institutions was in vain. Fortunately, there was a link. Higher de jure independence did really lead to higher de facto independence. However, this was moderated by the level of the rule of law in a country. The reason Greek, Italian and Spanish regulators scored so badly was because generally the law is obeyed less in these countries than in Sweden or the UK. Within these countries, regulators with higher levels of de jure independence generally do have higher levels of de facto independence — but they start from a much lower baseline. An Italian regulator would have to be designed in an exceptionally water-tight fashion to compensate for the built-in advantage (in terms of the rule of law) that any British regulator starts with.
 Hanretty, C., and Koop, C. (2012) “Measuring the formal independence of regulatory agencies”. Journal of European Public Policy 19(2), 198—216.
 Hanretty, C., and Koop, C. (2013) “Shall the law set them free? The formal and actual independence of regulatory agencies”. Regulation and Governance, forthcoming.