The economics of investment subsidies — draft terms of reference.

I am looking for an academic economist (or economists) who:

1. Has a deep understanding of the theory of microeconomics and welfare economics

2. Is comfortable with moving from theory to messy reality, keeping in mind how policy recommendations could be affected by features of reality that are absent from tractable economic models

3. Can communicate the implications of theory to non-technical readers.

4. Wants to influence practice

This would be paid consulting project. Contracting terms etc. remain to be determined. At this stage I am just trying to gauge interest. If you are interested contact me at pcarter at bii co uk.

Background.

Guidelines that govern the allocation of blended concessional finance (investment subsides) by development finance institutions specify that the investment must, in some way, be a response to a market failure. Market failures create the possibility of social returns that exceed private returns and hence justify a subsidy.[1]

More generally, impact investors (some of whom do not regard themselves as providing finance on concessional terms) seek to make investments that are beneficial to society, in some sense. Impact is often described with reference to the UN sustainable development goals. The idea that impact investing is something meaningfully different from purely profit-seeking behaviour could be seen as implicitly resting on the existence of market failures.

The development finance community is already familiar with externalities such as carbon emissions and learning from innovation, but other potential justifications for allocating subsidies may exist. In particular, market failures are about efficiency, but the SDGs and other impact objectives typically incorporate distributional concerns.

This project will present practitioners (people who decide the allocation of subsidies) with an accessible discussion of the justifications for subsidies, that expands on existing practice which is based on a small set of textbook market failures. The goal is to improve capital allocation decisions by sharpening our understanding of the rationale for subsidising investments.

Outline of the project.

1. A review of theory of socially optimal investment in the absence of market failures.

2. Extension to incorporate the objective of maximising a social welfare function in which distribution matters

3. Examination of externalities that are less commonly used as justification for subsidies: job creation in the presence of surplus labour and (potentially) complementarities in production networks and competitive behaviour (e.g. new entrants in markets where incumbents extract rents)

Socially optimal investment in absence of market failures.

Before looking at a case for subsidies based on externalities and distributional objectives, it would be helpful to review what “impact investing” might mean in their absence. Is there a case for impact investing, and potentially subsidies, that does not rest on the existence of market failures?

A half-remembered version of econ 101 probably goes something like this: profit maximising firms incur costs to increase production (including costs categorised as investments) until the marginal cost of increasing output equals marginal revenue. The marginal consumer’s willingness to pay results from their equalisation of marginal utility across the goods and services available to them in the market. Suppose we define “impact” per dollar of investment as change in welfare, or perhaps change in consumer surplus. Does a market equilibrium in the absence of market failures mean that the marginal social benefits of all investments equals their marginal costs, and it makes no sense to say one investment is more “impactful” than any other? Or, in the absence of externalities and so forth, might we still expect some variation in the impact return to investment across firms and entrepreneurs? Why?

Of course ability to pay varies across individuals, so in what sense is a market equilibrium socially optimal? Does it mean that social welfare is maximised? If not, could “impact” be defined as taking us closer to maximal social welfare? If, in the absence of market failures, the “impact” per dollar of investment varies across investment opportunties, then how can we achieve socially optimal investment without varying the cost of capital for those opportunities? In other words, in the absence of market failures, is there a case for “impact investing”, where that is understood to mean tolerating lower financial returns on selected investments, to have greater social impact?

It seems intuitively unlikely that the market works so that the impact of a dollar of investment in luxury fashion will be equal to impact of an investment in innovative basic healthcare. Models of investment often contain the feature that all investments that offer (risk adjusted) returns above a certain hurdle rate will be undertaken [citation needed]. If we only consider the individual private benefits of investments, does that necessarily achieve socially optimal allocation of capital? Does it matter that the return on the marginal dollar of investment (capital) is not quite the same thing as marginal cost, which is what firms set equal to marginal revenue? Why should we expect the available private expected financial return per marginal dollar of investment in new eye care technology and garden furniture, for example, to bring about the socially optimal levels of investment in their production?

Intuitions about impact probably involve distributional concerns, and ideas about a hierarchy of needs, that are absent from the concept of efficiency in a simple economic model. We might imagine a model with two types of consumer (rich and poor) and firms that produce two types of good (basic needs and luxuries) in which the market equilibrium is efficient, and the “social returns” on marginal investment in both types of firm are equalised, but it would be useful to explain what that does — and doesn't — mean.

Whether or not impact investing turns out to be a meaningful concept in a market equilibrium in the absence of externalities, why should we presume to be in equilibrium? There are many potential reasons (so called “frictions” such as transaction costs, “lumpiness”, market power, incomplete markets, information deficiencies, behavioural quirks etc.) why private costs and social benefits might be misaligned in reality. There is a view (expressed here for example[2]) that if DFIs see their role as compensating for market failures, then they should know what the specific market failure is that an investment targets. Frictions could be labelled market failures, but social and private returns on investment could diverge because of the cumulation of frictions, and subsidies could be justified when social returns significantly exceed private, without appealing to a specific externality or market failure. But if so, how could those social returns be identified (how would they be modelled, what would they consist of)?

Distributional concerns.

If we want to maximise a social welfare function that incorporates distributional aspects, then (I presume) in theory if we have arbitrary distributional objectives, we could justify arbitrary investment subsidies on that basis. But we do not have arbitrary distributional objectives. The SDGs call for the eradication of poverty and universal access to health, education, water etc. To what extent should investment subsidies be justified on distributional grounds — how should subsidy allocation decisions be taken in practice?

There is a view that capital subsidies are not the right response to the desire to permanently change prices for distributional reasons. If we want everyone to be able to afford rail travel, the government should subsidise rail tickets, not give train operating companies cheap loans. Existing guidance on the allocation of blended finance emphasises the time-bound nature of subsidies — firms or markets might need help from cheap capital getting off the ground, but the objective is to create something that will not require ongoing subsidies (at least, not in the form of delivering sub-market returns on capital in perpetuity). DFIs have traditionally been reluctant to subsidise power investments, for example, purely to achieve lower power prices for consumers, because of concerns that establishing price points below that which unsubsidised suppliers could sustain would be ruinous to market development. Instead, subsidies are justified on the basis that once certain barriers are overcome (knowledge is generated, demand is established, economies of scale are achieved etc.) private markets will be able to operate.

If DFIs are going to start allocating concessional finance based on distributional concerns, without appeal to efficiency (externalities) then first they will need convincing that is the right thing to do and secondly they will need guidance in taking allocation decisions. What are the right and wrong ways of doing it?

Less commonly cited externalities

Here are a few potential externalities that could form a rationale for subsidising investment, if a convincing case was made.

1. Job creation. The idea that the creation of decent jobs can be an externality is not new (see Acemoglu, Rodrik). Most models of externalities assume full factor utilisation and do not incorporate unemployment and underemployment. Cost-benefit analysis of investments usually assumes that the wage represents the marginal opportunity cost of taking a worker away from an alternative occupation, and the benefits flow from goods and service produced. In reality there is under employment in low-income countries, and there are social benefits from having an economy that provides many decent jobs (as opposed to a polarised labour market with many poor jobs). How should subsidies for job creation be allocated in practice? Subsidies for firms with low output labour ratios, or higher labour shares of income?

2. Complementarities in production networks. We would like to understand whether the social benefit of investments that raise productivity are internalised in prices when there are strong complementarities in production networks. Do the prices faced by producers of intermediate goods fully capture the ultimate social benefits, and if not, why not? References here include Jones (weak links in chain) and Liu (industrial policy in production networks). If there is a positive externality to investments that raise productivity in “central nodes” of production networks in the presence of complementarities, what should DFIs look for when deciding how to allocate scarce concessional finance?

3. Dynamic effects. Thinking in terms of externalities can mean projects are appraised in terms of comparative statics — the outcome after the subsidy is more efficient. How do things change if we think about investments that are likely to initiate more dynamic processes? Perhaps development requires the creation of firms that have the capabilities to expand the scope of their activities, evolve and grow, without knowing in advance what investment opportunities these firms will identify and execute. We may observe that successful examples of development have involved the creation of such firms, and that these firms are absent in stagnant economies and hence wish to subsidies attempts at their creation, without appealing to a specific market failure. Does that make sense? How could such allocation decisions be taken?

4. Competition. Many markets in lower income countries are uncompetitive and new entrants that intended to take market share by innovating or cutting prices could have a large impact. Is that an externality? If the private financial returns from such behaviour do not fully reflect the social returns, is that is reason to subsidise new entrants that are expected to behave more competitively?

[1] DFI Enhanced Principles: https://www.ifc.org/wps/wcm/connect/a8398ed6-55d0-4cc4-95aa-bcbabe39f79f/DFI+Blended+Concessional+Finance+for+Private+Sector+Operations_Summary+R....pdf?MOD=AJPERES&CVID=npes1Dq IFC using blended concessional finance https://www.ifc.org/wps/wcm/connect/1decef29-1fe6-43c3-86c7-842d11398859/IFC-BlendedFinanceReport_Feb+2021_web.pdf?MOD=AJPERES&CVID=ntFHkEh

[2] https://publications.iadb.org/en/smart-development-banks