Economic appraisal – we shouldn’t start from here

Present methods of economic appraisal, as applied in transport and regeneration decisions, are wrong in some circumstances and inadequate in many...

Present methods of economic appraisal, as applied in transport and regeneration decisions, are wrong in some circumstances and inadequate in many. They should inform decision-makers about value for money – about how much bang the public will get from the expenditure of the public buck, but in many cases they measure the wrong bang, the wrong buck, or both. 

---

David Simmonds will be developing this theme at both European Transport Conference (6-8 September) and in a presentation for Modelling World (2-3 November)

---

In particular, present methods don’t support integrated land-use and transport planning, despite this being universally considered desirable. Place-based appraisal is at long last seen as necessary, after nearly a century of place-based policies and investments, but for this purpose “standard” methods are either unhelpful or downright misleading.  

There are two broad parts to the problem, which can be identified as problems of “assumptions” and of “values” respectively.

So far as the Transport Economic Efficiency (TEE) analysis in DfT Transport Appraisal Guidance (TAG) is concerned, the “assumptions” problem refers mainly to the fact that TEE assumes perfect markets in everything outside transport.

This would be fine if first of all the transport models we use captured all of the land-use responses that could affect the transport outcomes, and secondly there were no externalities associated with land-use responses.

But our transport models don’t do that (they are positively set up to assume fixed land-uses), and there are many such externalities. 

Moreover, we have a planning system which has to make decisions about both privately-financed development and about publicly-funded regeneration schemes, all of whose impacts on transport need to be assessed.

At the moment, there is no standard method which allows us to fully assess whether spending money on regenerating inner-city areas to build housing nearer to jobs is better value for money than spending on improved transport from more distant housing that will be cheaper to develop.

Such appraisals can of course be done separately, but the transport appraisal (following TAG) one will not capture the full effects on land-use, and the regeneration one (following DLUHC guidance) will capture little except the increase in land value on the housing sites (though this may improve following the recent revision of DLUHC guidance).  

The “values” problem is that the “values” measured in the “value for money” analysis don’t match the “values” underpinning the strategic case, which sets out why the intervention is needed and what it is meant to achieve.

This is most clearly seen in the TAG wider impacts guidance on “moves to more productive jobs”.  Levelling up is a clearly stated policy objective of the present government (and to some extent has been an objective of most UK governments since the 1930s), and equally clearly that means seeking to increase the prosperity of “left behind” areas. 

But TAG assumes that there are fixed differences in productivity between places, and can show large benefits from interventions that move jobs out of low-productive “left behind” places into higher-productivity “well in front” places.  

TAG (and the Green Book on which it leans) addresses this by arguing that proposals that do not satisfy strategic objectives such as “levelling up” should not be considered at all. That would be fine if all the interventions that planning has to consider could be neatly classified into ones which satisfy those objectives and those that don’t.

Life is not that simple

Life is not that simple: different proposals will deliver different degrees of levelling up for different levels of costs and of other benefits - not to mention different levels of progress on other strategic objectives, since levelling up is not the only one.  

The solution to this is to make sure that the underlying values of the strategic case are represented as explicit monetary values in the economic case – for example, by putting a higher value on bringing jobs (or better-paying jobs) into left-behind areas.

This may be controversial, but that can be addressed by distinguishing a “policy value for money” (which puts monetary values on achieving policy objectives) from a “conventional, objective-free value for money”.  

This does require moving from the present appraisal approach, which is traditionally “non spatial”, to an explicitly spatial one. (The recent TAG guidance on “place-based appraisal” recognises the issue but is not a sufficient response to the practical challenge.) This in turns requires a change from measuring the direct benefits of transport in terms of time savings between zones to measuring accessibility improvements for people (or businesses) in zones.

There is a substantial literature on the use of accessibility measures in appraisal, and they can be expressed in money terms. Critically, they can be combined with zonal estimates of the other benefits or malefits impacting on households, businesses etc as a result of whatever is being appraised, for example, changes in the quality of the local environment (from more or less traffic, more or less pollution) and changes in income (resulting for example from the changing availability of jobs).  

That is about the benefits to the individual and households affected by the proposal being appraised (directly or indirectly). What also needs to be added is the value that society more generally puts on redistributing prosperity, or wellbeing, towards deprived and left-behind communities – in effect, the externality of such changes as perceived by everyone else (or by the government on our behalf).  

This requires defining exactly what (forecastable) consequences are intended from “levelling up”, i.e. choosing indicators to measure progress towards strategic objectives, forecasting how much (or how little) they will arise from a particular scheme or programme, and putting money values on each of those consequences – for example, a premium of so much per additional job in certain areas, a premium of so much per additional dwelling in a different set of areas.

This will also be controversial, but no more so than putting value on saving lives through accident reduction – and some of the values can be estimated by analysis of past decisions to spend on relocating jobs or redistributing development.   

The point about putting money values on the forecast outcomes – in addition to all the existing calculations – is that the strategic objectives, and the degree of progress towards those objectives, can be reflected in the cost-benefit ratio.

Care should also be taken to consider what other costs or cost savings may be implied by that progress: for example, if a proposal or programme is forecast to encourage people to locate in “levelling up” areas rather than in already-prosperous areas, there will typically be savings in the costs of public services e.g. through not having to buy expensive land for schools and hospitals.  

That does not mean the additional values are to be hidden in the numbers; they should be explicitly set out, and could be used in calculating “adjusted BCRs” or “policy BCRs”, as wider economic impacts are already. 

The key step forward is that adjusted BCRs will then allow the value for money of a proposal to be assessed inclusive of levelling up and other strategic objectives, and in particular will allow alternative options or alternative schemes to be compared on a consistent and comprehensive basis.

By calculating the value of the “bang” we are really interested in (and taking account of the appropriate “buck”) we can make economic analysis relevant and helpful in a world that increasingly seeks to achieve results outside transport through interventions within transport.

Without such changes, the economic case in appraisal is doomed to become less and less relevant, and the strategic case to become more and more mired in special pleading for particular places.

---

David Simmonds will set a more formal and more detailed version of this argument in a paper for European Transport Conference (6-8 September) and in a presentation for Modelling World (2-3 November)