Almost all government agencies are in cost-cutting mode at present, under instructions from the incoming government. All sorts of things they, or the previous government, thought were nice to have and some things perhaps they thought were really rather useful indeed seem to be going by the wayside.

But at the Reserve Bank they are planning a new nice-to-have.

The Bank has a consultation process open at present on a proposal to “invest in” (that is public sector for “spend”) a new survey of business people, asking about their specific numerical expectations for a subset of macroeconomic variables. What is striking, in this climate of general public sector austerity, is the utter absence of any analysis of what gaps this survey is designed to fill and how material they might be for monetary policymakers. That was one thing in the era of bloat and lavish increases in public spending – in which the Bank fully participated over the last few years – but it really shouldn’t be acceptable at present.

On the substance, I’m sceptical of asking specific numerical questions to a wide range of businesses about specific macroeconomic variables (up, down, or the same questions just might be useful) and of whether there are really information gaps at all. I put in a fairly short submission, as follows:

Submission to the Reserve Bank on proposed Business Expectations Survey

Michael Reddell

23 March 2024

This is my personal submission in response to the Reserve Bank’s consultation document on the proposed Business Expectations Survey.  I have been a long-term user of business and household surveys, was involved in various refinement to the existing (and now) expert survey of expectations the Bank runs, and was a user of such material for several decades as a senior monetary policy adviser in the (former) OCR Advisory Group.   I am now also a monetary policymaker in another country, so am very much attuned to the perspectives and interests of policymakers.

There appears to be no prior background document referred to, so I am working on the assumption that the current document is all the information/analysis that the Bank is choosing to provide.

On that basis, it is quite astonishing that there is no analysis of the case for (or against) a survey of the sort the Reserve Bank proposes to spend additional public money on.   As you will be well aware, there is a plethora of surveys in existence (which wasn’t the case several decades ago), including the monthly ANZBO, the monthly BNZ PMI and PSI, the quarterly QSBO, and of course the Bank’s own survey of (now) expert expectations –  to name just the more-prominent of the business surveys.   The Bank used to be a large funder of the QSBO, and I assume that that is still the case  (and I used to, and would still, champion the case for some slight extensions to the QSBO, notably around wages).

The survey you propose is of a subset of macroeconomic variables.  You will no doubt be aware that the existing Survey of Expectations began with a much larger range of respondents, including business and union people, and was eventually slimmed down towards its current form in part as it was realised that many of the business respondents had no particular reason to have formed specific expectations for many macroeconomic variables, or even in some case to be aware of the specific measures being asked about.   I am aware, of course, that your current proposal is for a more-limited subset of questions, and in the case of near-term inflation it is a fairly commonly asked question even of households.  But do you really believe that many business respondents are likely to have well-developed thoughts on what the inflation rate might be in 10 years’ time (or even that they have risen or need to have even implicit views given that very few nominal contracts exist)?  Similarly, how many of your respondents will be familiar with the details of the specific wage series you ask about, or will recognise a difference between annual GDP growth and annual average GDP growth (let alone know the numbers).  

It is much more common for surveys of the wider business community to use tendency questions and focus reporting on net balance results.  Absent any compelling analysis or evidence from abroad, I’d have thought that was likely to be a much better way to go, if you really believe there is value to the MPC in yet another business survey (typical targeted respondents seem much more likely to have some sense as to whether growth or wage inflation might speed up or slow down over the period ahead than to have meaningful numerical expectations.  In respect of labour market slack existing questions in other survey (“difficulty of finding labour”) seem to cover the ground these sorts of respondents could typically generally offer (few will, by contrast, have particular to care what the specific HLFS definition of unemployment is).

But, more generally, where is the evidence of gaps in the (survey) data that the MPC needs to make good monetary policy?   This is posed as a serious question.  I’m not aware of such gaps, but perhaps you are and could have elaborated on this point for submitters.  Material monetary policy mistakes have (and no doubt will again, in the nature of imperfect individuals and institutions) be made, but is there really any evidence that it is (a subset of) business expectations of macroeconomic variables that is lacking, or which led the MPC astray in the last few year?  I’d be surprised, but am certainly open to evidence.

It was perhaps surprising that there was no indication in the consultation document of the cost of (a) developing and (b) administering the proposed survey, which makes it even more difficult to assess whether it is plausible that there will be net benefit for taxpayers (as ultimate funders of the Bank).

On a couple of specific items in your document:

• It was rather surprising to see several references to “respondent burden”, including to justify leaving out very small businesses from the possible sample base.   It is a highly relevant consideration when the state uses the coercive powers SNZ in particular has to insist on private firms and citizens completing surveys, but as you note this proposed survey will be quite as voluntary as any of the other business surveys, and it isn’t obvious that –  to the extent you are likely to get any useful data from this survey – businesses with 5 employees will have anything less to offer than those with (say) 15, again bearing in mind that it is macroeconomic variables you are asking about.  Moreover, since you propose to revolve the panel, no one potential respondent would be asked for response all that often or for that long.    More generally, I note the very low response rate you expect (“10-20%”) and given the severe selection bias that probably introduces one is again left wondering about the strength of the case for the survey at all.
• I’m also unpersuaded by the (very sketchy) case made for the exclusion of primary industry firms.  It might be common to exclude such firms, but (as you recognise) this is an economy in which the primary sector is of some considerable importance, and (again) these are macroeconomic questions you are proposing to ask, and it isn’t obvious why primary industry respondents would be any less equipped to answer such questions than other firms their size. I note that you say that primary industry firms are not generally price-setters, but it is a galaxy of macro questions you are proposing to ask, not ones about price-setting intentions (and primary sector firms employ, borrow, fx hedge (or not), and so on).

Finally, and while I recognise that the Reserve Bank’s current funding agreement does not run out until next year, in the current climate in which government agencies pretty much across the board are being expected to exercise considerable spending restraint, often cutting established functions and activities, it seems extraordinary (but consistent with a longstanding culture of Reserve Bank exceptionalism –  and yes, I used to share and even champion it) for the Bank to be proposing a new ongoing spending commitment on what is, at best, a nice-to-have, supported by no serious underlying analysis of the need for this new survey, or even of the Bank’s ability to continue to fund it (against other priority claims) if and when the Minister of Finance finally catches up with the Bank and adjusts its authorised spending levels.