I recently read a surprising statement by top officials in the Bank of Israel (BOI), the country’s central bank. They made public their disappointment from the commercial banks’ unwillingness to act as the the central bankers expected them to. More precisely, the BOI’s economists were disappointed that the outcome of their new policy regulation was far from the one they were aiming for.
Policy failures are not rare. Policy-makers try to navigate a complicated institutional set-up in a highly uncertain, multi-faceted and heavily political-laden environment, and failures are an unavoidable part of policy-making. However, their astonishment insinuated that the policy’s outcome had questioned some fundamental beliefs they held regarding the mortgage market or the banking system at large. Since my research deals with the philosophical scrutiny of macroeconomic policy, I could not help myself reflect on the decision-making practice that led to their astonishment. Once again, the reflection process affirmed the distinction between the reasoning employed in economic research and economic policymaking. I will try to formulate one aspect of the conceptual framework that philosophically distinguishes economic research from economic policymaking. Public or private comments that can assist in developing and improving the analysis are welcome.
Looking for an Explanation
First, a few words on the BOI’s policy decision can help with some initial sorting out. It started when the BOI’s economists observed that despite a steady decline in the banks’ costs on the financial capital they need for lending (such as interest the banks pay on deposits), a similar decline was not followed in the interest rate the banks receive from borrowers. The BOI decided to narrow this interest gap to benefit the mortgage borrowers without endangering the banks’ financial stability. I will not get into the technical details of the regulation they issued eventually (if interested, it is found here in English). It suffices to say that the policy decision was to increase the portion of the mortgage interest rate determined by the market instead of by the central planner. The idea was that more competition between banks would surely end up with lower interest rates.
I haven’t talked to the BOI’s economists involved in the policy decision. Still, from my background knowledge on the reasoning process of economic policy-making, the first thing they did is explain the phenomenon of interest, i.e., the interest rate gap. They wanted to know why this fact has happened because only once they grasp the reason (or reasons) for the gap could they set policy accordingly to reach the desired outcome. In other words, they were looking for an explanation.
The point is that unlike in scientific research, in the context of policymaking an explanation (or a cluster of explanations) to a social phenomenon is not an epistemic goal in itself, nor does the somewhat evasive notion of ‘understanding’ that follows a true explanation. Given the desired outcome, policy-makers’ goal is to pinpoint the policy action needed to achieve it. Therefore, it seems more accurate to formulate policy-makers’ epistemic goal as achieving a certain ‘grasp’ of the phenomenon’s policy-relevant aspects. The bottom line is that this grasp stems from a policy-relevant explanatory hypothesis (or a cluster of hypotheses). The question, then, is whether the prefix ‘policy-relevant’ comes with a different type of considerations for justifying or selecting an explanation, other than the ones already common in scientific research, such as the explanatory virtues of simplicity and unification.
If the answer is yes, selecting an explanatory hypothesis is done according to standard explanatory virtues, followed by policy-makers’ deliberation as to the right policy measure for achieving the policy goal. If the answer is no, then the explanation itself is also selected according to its policy usefulness, and therefore there must be some explanatory virtues – not reducible to the standard ones mentioned above – that denote policy usefulness. Of course, asserting for the latter is not self-evident. Still, it seems to be even less plausible to think of policy-makers as scientists in a lab, where the same epistemic considerations that apply in pure research is also at work in the selection of policy-oriented explanations. For instance, it is reasonable to assume that explanations which pinpoint the required policy more clearly have cognitive advantage over explanations which demand more cognitive effort to extract a policy from them.
In short, an explanation can excel in all of the standard virtues, but if it does not support the required level of policy-usefulness, then the hypothesis will probably be dismissed. On the contrary, if an explanation has less explanatory power but facilitates a more useful policy-decision, it can have a cognitive advantage for policy-makers and, for that reason, be preferred over less policy-useful explanations. Although further scrutiny of this assumption is due, it can wait for a more suitable venue than the abridged version a blog post can accommodate.
Efficacy as an Explanatory Virtue
Suppose the search of policy-makers is after a policy-useful explanation. In that case, there must be some epistemic virtue that classifies the usefulness of the relevant explanatory hypotheses accordingly (It should be noted that the usefulness refers to the policy decision, not the active setting of policy, to account for situations where the policy decision is not to act). This explanatory virtue is what I wish to call ‘efficacy’.
At this point, I prefer to think of efficacy as a title for a set of virtues, instead of a single virtue, because it allows further examination of the various virtues applied in the construction of policy-relevant explanation. Whether this cognitive character is correct – and whether it matters – will have to wait for another time. For now, it is interesting to contemplate the implications that efficacy entails. For example, explanantia are also judged according to each explanans’ policy derivation, and accordingly an explanans which effective policy can be derived from relatively easily than alternative explanantia has epistemic advantage despite not necessarily being the ‘best’ explanation, whatever that means exactly. The gap between policy goodness and policy usefulness is also where policy mistakes might arise from.
With this in mind, let us go back to the BOI. The explanation the BOI’s economists gave for the interest rates gap was not that the mortgage market was un-competitive, but rather that there is some bug that prevents the competition from taking place. The un-competitive-market explanation is perhaps much simpler than the “some bug” explanation. But, if the explanation is un-competitiveness, it requires a tremendous amount of cognitive effort to derive a policy from it, and another huge regulatory effort to turn the mortgage market more competitive, not to mention the high uncertainty regarding the policy’s outcome. On the contrary, the “some bug” explanation has efficacious power in the sense that once the bug is identified, the cognitive effort required to derive a policy is relatively small. More so, After fixing the bug, the expectation that the market will operate more smoothly is reasonable and clear. Therefore, it is no surprise that the BOI’s economists selected an explanation which perhaps has less explanatory goodness, but has a clear-cut policy derivation.
Numerous questions arise from the above argument. We might ask what is the inferential scheme of this process, discuss the proper manner in which efficacy should be assessed, including mapping the various efficacious virtues as first or second order (a policy-relevant reflection over an explanation’s usefulness) and much more. Meanwhile, whatever virtues should be added to the explanatory arsenal, the take-home message is that policy cannot be taken philosophically for granted. Policy, and policy decision-making in particular, is not “research in practice”, but a distinct field that deserves a designated philosophical attention.