Emerging policy issues require us to rethink the role of ‘Value of Time’ in Transport Appraisal

THERE ARE MANY criticisms of the dominant position that calculated travel time savings have often occupied in transport appraisal, both because of much broader objectives of public concern, and because travel speed in itself is only one criterion of transport efficiency.

But there is also room for rethinking another aspect of the value of travel time savings, due to economic analysis of the relationship between value of time and personal income for individual travellers. Department of Transport and other research has traditionally indicated that the amount of money people are prepared to pay to save time increases with income. But Treasury and other research has meanwhile suggested that the utility to them of the money they pay to this end decreases with income. Integrating these two findings needs a new focus on an old idea, the ‘marginal utility of time’.

This is not to pose a simple choice between efficiency and equity, but a recognition that just as the utility of money depends on disposable income, so does the utility of time savings depend on disposable time, ie how busy, pressured and flexible an individual’s life is. My contention is that the interaction between money and time pressures fundamentally influences the true value of time. Our proper understanding of the significance of this needs a new research strategy, which pays attention to the interaction of time pressures and income pressures, social as well as economic, and the enjoyable or wasteful conditions under which time is spent.

The big picture

There is widespread agreement in principle now that transport policy and infrastructure must have concern for the big issues – local and global environments, local and national economies, efficiency, fairness and quality of life. But there is not agreement on how these concerns translate into the daily business of forecasting and project assessment, pricing and traffic management, and the balance between cars, lorries,buses, trains, footways and cycleways.

Readers of Local Transport Today and TAPAS will have noticed a mood of concern among many of the magazine’s new and longer term contributors, who have played a leading role in thoughtful discussion of transport policy.

In the last issue, Professor Nick Tyler, of University College London, challenged ‘hypermobility’ as a damaging addiction[1]

“… especially over the past 100 years, we have designed our world around the idea that such movement should not only be possible, but desirable, and eventually, unavoidable and so needs to be intrinsically catered for. The measured objective seems to be the quantum of “movement”, not the quality of life. So provision for movement has been prioritised, even where the quality of life suffers as a result”

In the same issue John Dales[2] criticised the deep disconnect between Government rhetoric on growth and the street level, family level, local level impacts on better neighborhoods and the quality of the public realm. In the previous edition Professor David Metz[3] discussed the implications of ‘Travel Demand saturation’ as an alternative diagnosis of the current and future prospects for transport’s role in the economy. And Professor Glenn Lyons (2025)[4] gloomily commented that

“If I was sat 30 years ago at the start of my transport planning career and someone had described the reality of 2025 to me, it would have seemed preposterous. It would have sounded like a future in which humanity was entering another dark age - characterised by a world stumbling into an unknown and unfathomable future, governed by forces outside anyone’s real control.”

Bringing these thoughts together, last month[5] LTT and TAPAS Editor Peter Stonham wrote an editorial reporting views that the Government’s ‘driving ambition to achieve economic growth is… sweeping other important things out of the way’. He noted suggestions that

“Growth that involves speeding climate change, or winning high stakes international technology market races, is not sustainable, and instead may lead to the current generation taking resources away from future generations at a time when some acceptance is perhaps overdue that constant growth may not be possible.”

The Value of Time and Economic Efficiency

One of the features of any discussion of this kind is its implications for the methods of appraising the value for money of transport investments, in which the ‘value of time savings’ has for decades had a critical role. This has sometimes been treated in a simplified way as follows:

This is deliberately oversimplified, but nevertheless calculations of how much time would be saved by transport infrastructure investment have been at the heart of making the case for approval of specific projects, and often the value of time savings has been the biggest single element of suggested benefits, sometimes (but no longer) even large enough to appear to justify the scheme on their own.

But now we need to reduce the dominance of time savings

It is now widely agreed that factors other than travel speed are important to travellers themselves, and can be decisive. However, there is not always agreement about which factors are the most important, as discussed at successive Round Tables of the European Conference of Ministers of Transport and International Transport Forum[6].

1. Some focus on the evidence that time is not as overwhelmingly influential in individual travel decisions as has been thought. Comfort, convenience, noise levels, air quality, opportunity for exercise, access to telecommunication signals and the conditions to use them, social status, desire for variety, desire for habitual reassurance, enjoyment of company or conversely privacy, conditions suitable for undertaking productive work or private enjoyment during travel, access to refreshments and toilets, and many other factors affect choices. By putting too much emphasis on duration, and not enough on the other factors, there will be too much emphasis on speed.

2. There is an unresolved argument about the evidence that the overall time spent travelling has changed little as a result of faster travel, and in some analyses not at all. People do not ‘save time’ so much as use some or all of that ‘saved’ time to travel more frequently or to more distant destinations. There will be benefits to them of doing so, of course, but there will also be costs, for example by reduced accessibility for others due to consequential land use changes, and by the generation of extra traffic which in general erodes, or in certain circumstances outweighs, the intended time savings. The eventual gainers will not necessarily be the same people at all as those whose hours of time saving were assessed in an appraisal: the initial advantages to them will have an effect, for example, of changing land prices, resulting in a shift of benefit from travellers to land-owners. This argument claims that there is less, or no, reduction in the time spent travelling and it is therefore misleading to value the benefit in units of time: it should be measured in terms of the overall balance of change in the quality of life, and perhaps accessibility.

3. There are some travel activities whose benefits are greater the longer is spent on them (within limits). Then time savings can have a negative value. This especially relates to the health benefits of active travel notably walking and cycling, but can also apply to that category of travel where the experience of travel (e.g. sightseeing) is the intrinsic benefit sought. In these circumstances simply reducing the time spent will not be an advantage. Walkers and cyclists may well choose to take a longer, slower route specifically to spend more time on them, or to enjoy a ‘nicer’ outing. Tourists will not want to pay more for a briefer tour. One of the intriguing lesser weaknesses of the analysis of the Stonehenge A303 road proposal was the observation that a significant part of the local congestion was due to motorists choosing to slow down as they passed, to enjoy the view of the monument!

4. Similarly, some journeys, particularly by train, are very suitable as a working environment by using on-line connectivity or suitable tables, power sources, lighting and ride comfort. Those who do this often (including some academics) find that there are discrete amount of time which can be usefully spent on marking scripts or writing papers for conferences. An hour or more may be useful, but less doesn’t give sufficient time for focus. In these circumstances arriving at your destination earlier is not a benefit, and can be irritating. The same argument is increasingly being offered if autonomous vehicles offer the opportunity of other activities while travelling by car without needing to drive.

5. Some transport improvements are specifically addressed to providing higher standards of comfort. A feature of the interaction between comfort and speed is that empirical work has found that less comfortable travel, other things being equal, has higher values of time (that is, people are prepared to spend more money to save an hour of uncomfortable travel than comfortable travel). Therefore one of the effects of higher comfort is to lower the value of time savings. One can say that achieving a lower value of time is a mark of the success of such improvements. Therefore a progressive reduction in the value of time, achieved in this way, is a benefit, and a mark of successful progress.

6. The reliability of values of time savings used for appraisal is fundamentally dependent on the validity of calculations of the effects of changes in speed of travel on the actual choices and behaviour of individuals and companies. I would argue that some simplifying assumptions routinely made that lorries and vans have the same demand for travel, whatever changes are made in speed, are so far from the truth that no sensible judgements can be made about the economic value of the time savings they are offered.

7. There are other, online, ways of engaging with economic and social activities than travelling to them, and the other ways are increasingly improving and sometimes superior, whatever speed increases are achieved.

8. There is increasing concern that promised time savings are of rather short duration, as traffic growth erodes them and congestion increases to its initial level or more. Established appraisal methods (which use a counter-factual ‘what would happen otherwise’ as the basis for defining benefit) are able to ignore this issue, but it remains an issue of continued public concern.

Taken together, all this additional insight very properly should change the way that time spent in travel is appraised, and reduce the dominance of time savings per se.

Equity and Efficiency

There is one other critical new feature, of increasing importance, that will change the way we use the money value of time. That is a shift in how we distinguish between equity and efficiency.

In discussions of transport policy, the recent tradition is of a rather cautious presumption that there is a tension between ambitions of equity and ambitions of efficiency, and both should be pursued. This would be reinforced by a presumption that an economy cannot long continue to be efficient if its governance is perceived as inequitable. But there are two separate traditions, on the treatment of the value of travel time savings in official transport economics, and on the declining marginal utility of income in official Treasury economics, which are now converging on a synthesis of some key aspects of equity and efficiency. A richer view of equity, and a more multi-dimensional view of efficiency, can be reconciled.

Declining marginal utility of money

The tradition of economic studies positing that the value of an additional pound to a wealthy person is less than to a poor person is very long, with references going back to the 19th Century, notably Marshall (1890) ‘The richer a man becomes the less is the marginal utility of money to him’[7]. The latest statement of the application of this principle is set out in the Treasury’s ‘Green Book’ for 2022[8]. It stated that

“5.70 Distributional weights are factors that increase the monetary value of benefits or costs that accrue to lower income individuals or households. They are based on the principle that the value of an additional pound of income may be higher for a low-income recipient than a high-income recipient.

“A3.5 A review of international evidence provides an estimate of the marginal utility of income at 1.3[9]. This is used by DWP [Department of Work and Pensions] in distributional analysis. The estimate of the marginal utility of income can be used to calculate welfare weights to adjust costs and benefits.”

I have not tracked down the first application of this principle in Treasury guidance, but it is not new. Assessing ‘Equalities Effects’ was an obligation under the Equalities Act 2010 and the Green Book refers to this as the legitimation of treatment of distributional impacts.

Empirical analysis has repeatedly made estimates of the size of this difference, and the most recent meta-analysis of published results has been assembled in a major review article by Acland and Greenberg (2023)[10] of evidence from the US and UK, which calculates distributional weights to be applied in cost benefit studies. They report that the appropriate values for the weighting factor would be based on an elasticity of 1.6, ie each 1% increase in income is associated with a 1.6% decrease in the marginal utility of income. (They suggest sensitivity testing within the range 1.2 and 2.0.). They also point out that as well as affecting the relative advantages of different groups, this affects longer term appraisal of total discounted benefits, since if future average incomes rise, each successive increase will also give rise to the corresponding reduction in marginal utility. Acland and Greenberg’s argument is that

“a dollar’s worth of consumption is typically considered to have a greater impact on the welfare of the poor than on the welfare of the wealthy–due to diminishing marginal utility of income–which means that any given quantum of welfare is represented by a smaller number of dollars if it accrues to a poor person than if it accrues to a wealthy person. As a result, BCA is actually biased against the poor and thus may promote inequitable strategies for increasing aggregate welfare.”

The Value of Travel Time Savings and Income

Application of money values to travel time savings has been a practice of the Department for Transport and its predecessors ever since the 1950s (and possibly earlier), with a major statement of principle in 1970[11], and a research programme of major empirical studies updated every decade or so[12]. This history is very relevant in knowing ‘how we got to where we are today’, and a fuller version of the historical background is given in an Appendix.

More materially to this discussion, it has led to discussion on the values which should be used in forecasting and in appraisal published and regularly updated in the Department’s Transport Analysis Guidance TAG[13].

A key theme of that guidance is that it is consistently shown (from multiple national studies across different countries) that the amount of money people are prepared to pay to save travel time increases with income. This was shown initially by observing people’s choices between some cheaper, slower alternatives and some faster, more expensive alternatives, and recently mainly by their stated responses in simulated choice experiments. For business travel time the main procedure has been to assume that businesses behave in effect by valuing their employees’ time in relation to the salary and related overheads. For appraisal of time savings for individuals on their own personal and leisure travel, the procedure has long been not to accord richer individuals a higher value of time, in spite of their own stated preferences, but to adopt an average level, sometimes called an ‘equity’ value.

Thus discussion recently has therefore concerned two findings of the effect of income

1. The Treasury guidelines that note empirical evidence that the marginal utility of money to richer people is less than to poorer people.

2. DfT research that has shown that richer people are prepared to pay more of their money to save travel time.

So taken together, the richer people are indeed prepared to pay more of their money to save time, but they do so with money which they value less. So the obvious question is then, do these results, taken together, mean that the utility of time savings to the rich is less than the utility of the same time savings to the poor? What happens to the value of time if we apply the distributional weights to the stated willingness to pay money for time savings?

In applying the Treasury distributional weights, would we end up with a ‘reverse income effect’, which in discussions could be crystallised with an emphatic rejection of the form

“It would be manifestly illogical to value time benefits to pensioners more than to wealthy productive adults. This would harm economic growth, which surely is a more important priority than equity”.

In this case there appears to be a tension between meeting objectives of equity and objectives of efficiency. This is partly, but not fully, resolved by noting that the idea of distributional weights in this case has not actually been justified by principles of equity, but from a different definition of efficiency, one based on the idea that an efficient economy is one which maximises the economic utility of its citizens. That the marginal utility of money is negatively related to income is a statement of empirical validity, not a political or value judgement.

What this is really saying is that GDP growth as currently defined is not a very good measure of efficiency or total welfare.. In terms of the objective to maximise consumer welfare, application of distributional weights would be more efficient, not less.

However, there is another way of resolving this problem. In recent years the tradition has been to estimate values of time directly by stated preference experiments measuring people’s ‘willingness to pay’ for time savings – see, for example, Batley et al (2018)[14]. This is convenient as it appears to avoid the need to identify the marginal utilities of time and money separately, and simply focuses on the money value of travel time savings directly. But this is at the sacrifice of a fuller understanding of the underlying separate reasons why, in some circumstances and for some people, time pressures or money pressures may be subject to different constraints.

There is therefore a weakness in the statements above about the relationship with income. We can say with some confidence, from the history of economic thought, and now from the Treasury, that the marginal utility of money reduces with income. And we can say with some confidence, from DfT, that the amount of money people are prepared to pay to save travel time increases with income.

So there is a missing link in the argument, because it has not identified how the marginal utility of time varies with income.

The most important clue to resolving the conflict is the understanding that what we call ‘the value of time’ is not a single quantity, but the ratio of two separate economic concepts, the marginal utility of time and the marginal utility of money. These are independent in the sense that each of them can vary under the different influences from a wide range of economic and social factors, of which income is only one, and not always the most important. There will be links between the two marginal utilities, via relationships with wages, incomes, and the trade off between work and leisure.

My proposal[15] has therefore been that - just as we treat the primary influence on a person’s marginal utility of money, as being the total amount of (disposable) money they have – we should logically treat the primary influence on the marginal utility of time as being the total amount of (disposable) time they have, ie essentially how leisured or constrained their lives and timetables are. Then the trade-offs between travel cost and travel time will vary according to, for example, employment status, family or household pressures, age, and other such variables. The motive for this is not mainly to pursue social welfare benefits at the expense of economic efficiency, but to redefine economic efficiency to take account of the differences in importance of time savings and money savings to different people, in proportion to the conditions of their lives, and their actual role in achieving additional beneficial economic or social outputs, or enhancements in personal health or wellbeing.

In 2018, the International Transport Forum convened a Round Table conference on the value of time initially titled ‘Zero Value of Time’ and its overall summary is published under the title ‘What is the Value of Saving Travel Time?’[16]

In the overall summary and conclusions it reproduces a hypothesis I suggested to resolve what I consider to have been a mistake in focussing the distributional question essentially as an issue of income. It uses the device of identifying people as ‘time rich’ and ‘time poor’ analogous to ‘money rich’ and ‘money poor’. (Both of course will show a continuous variation, but it is easier to explain in the form of a two-by-two matrix). (See Table)

In this framework, the stated ‘value of time’ for a person with little money and little available time would be low, and the stated ‘value of time’ for a rich person with plenty of spare time would be high, but both are indeterminate and could even appear identical. But is it manifestly obvious that the real usefulness of money or time savings for the two people is fundamentally different. This identifies the real weakness treating the essential source of variation in the value of time from a relationship with income. There are some people with low incomes whose time is much more important, to themselves and/or to society as a whole, than some other people with high incomes, and I do not think it is inefficient, or uneconomic, to recognise this in transport provision.

This approach is directly derived from the proposition that the value of time is the ratio of the marginal utility of time and money, and therefore must be understood as influenced by the interaction of the two dimensions.

What is to be done? Proposals for a research programme

A weakness of empirical work on the distributional aspects of values of time has been that the influence of pressures of income has been much more emphasised than the effects of pressures of time. Sometimes this has meant in effect assuming the value of money is constant and equal for everybody. I suggest a three-part research strategy is needed to look into this more deeply.

These key elements would be

1. There is a potential useful source of insight by reanalysis of earlier studies in cases where there was suitable segmentation of the sample such as that by Anable (2005) and since. This makes prima facie sense in terms of the life-style constraints on time use: this would include some sensible definitions of employment status (full time, part time, not employed, retired) and sensible distinctions of family circumstances (single parent, number and ages of children), and some tentative measures of other constraints of health, social obligations and conditions of circumstances affecting time of day constraints, as a first pass to develop likely influences.

2. There is also a considerable literature of previously ignored social factors, such as research by Chatterjee et al (2018) on the changing travel patterns of young adults in particular, and of new patterns of home-working and technical influences on the need for travel. Most of this work has not explicitly focussed on values of travel time, but it has identified different patterns of choice related to the marginal utility of time savings which help to explain why the trends in their choices are different from those which have been thought dominant for earlier generations, a rich source suitable hypotheses for testing.

3. A third strand, most important in the longer run, would be a new generation of value of time studies which are specifically and intentionally focussed on separate estimation of the two sets of marginal utilities. This would enable unification of the separate strands of research being undertaken on time savings and road pricing.[19]

4. The two different marginal utilities then offer a much more coherent way of including consideration of ‘in travel’ time that do not fit into values of time savings at all, as in the discussion above of reducing the dominance of time savings. Not all time spent on travel is a disutility.

5. This more closely reflects the concerns of earlier approaches to the role of household time scheduling in the ‘activity models’ drawn from the time-space geography rather than transport modelling, eg by Jones et al (1983)[20], whose importance has never been fully recognised.

It is very likely that the central empirical task would be development of a set of categories for a form of segmentation which as closely as possible corresponds with the different real-world pressures on time and money which drive differences in the value of time, and therefore more closely explain different choices and priorities.

Conclusion

For these reasons, I hope you may agree with me that a fresh look at this topic would indeed be timely.

It is surely true that in recent years transport conditions and individual behavioural priorities within them, and more widely, have changed, and there are many situations where the stated willingness of people to spend money to save travel time do not reflect the different benefits of the conditions of travel. But also, the stated willingness to pay for time savings has to be moderated by the evidence that the value of money is less as income increases. This itself changes estimates of the value of time savings.

A better measure of the value of time should therefore have to take into account how much time people have to dispose – how much free time, the nature of work and family pressures, the flexibility of their lives – just as their willingness to pay also depends on how rich or poor they are.

References and Links

1. Tyler, N (2025’ ‘Hypermobility, a damaging addiction of modern life, Local Transport Today 19.2 2025 and on TAPAS https://tapas.network/78/tyler.php

2. Dales, J (2025) ‘Going for Growth’: what does it mean to the people in the street, https://tapas.network/78/dales.php

3. Metz D (2025) Travel demand saturation means that we should accept our road system has reached maturity, and act accordingly https://tapas.network/74/metz.php

4. Lyons G (2025) Looking for the light in a dark age https://tapas.network/74/lyons.php

5. Stonham, P (2025) Should growth really ‘trump everything’, Editorial Opinion, LTT 908 3.2.2025 and TAPAS, https://tapas.network/76/stonham.php

6. This summary of these is taken from Goodwin P (2019) The influence of technologies and lifestyle on the value of time, International Transport Forum Discussion Papers, No. 2019/03, OECD Publishing, Paris, Roundtable 176 https://www.oecd.org/en/publications/the-influence-of-technologies-and-lifestyle-on-the-value-of-time_9811ba36-en.html

7. Marshall A (1890) Principles of Economics, 8th Edition Macmillan, London, 1938

8. https://assets.publishing.service.gov.uk/media/6645c709bd01f5ed32793cbc/Green_Book_2022__updated_links_.pdf
   

9. The reference given for this is an earlier study by Layard R, Mayraz G, & Nickell S et al. (2008) “The marginal utility of income” Journal of Public Economics, Vol. 92, pp. 1846-1857.

10. Acland D and Greenberg D (2023) The Elasticity of Marginal Utility of Income for Distributional Weighting and Social Discounting: A Meta-Analysis , Journal of Cost Benefit Analysis https://www.cambridge.org/core/journals/journal-of-benefit-cost-analysis/volume/0502589C38385DBC58CF4AF1DDC9B7CC

11. McIntosh, P.T. and D.A. Quarmby (1970), Generalised Cost and the Estimation of Movement Cost and Benefit in Transport Planning, MAU Note 179. Department of the Environment, London

12. https://assets.publishing.service.gov.uk/media/5a80bdeced915d74e62302f2/vtts-phase-2-report-non-technical-summary-issue-august-2015.pdf
    

13. https://www.gov.uk/guidance/transport-analysis-guidance-tag

14. Batley R, Dekker T & Stead I (2018) Worthwhile use of travel time and implications for modelling, appraisal and policy planning – the UK experience, ITF ‘Zero Value of Time’ Round Tablehttps://www.itf-oecd.org/sites/default/files/repositories/worthwhile-travel-time-use-uk-batley-dekker-stead.pdf

15. Goodwin P (2019) The influence of technologies and lifestyle on the value of time, International Transport Forum Discussion Papers, No. 2019/03, OECD Publishing, Paris, Roundtable 176 https://www.oecd.org/en/publications/the-influence-of-technologies-and-lifestyle-on-the-value-of-time_9811ba36-en.html

16. ITF (2019), What is the Value of Saving Travel Time?, ITF Roundtable Reports, No. 176, OECD Publishing, Paris. https://www.itf-oecd.org/sites/default/files/docs/value-saving-travel-time.pdf

17. Anable, J (2005) ‘Complacent Car Addicts’ or ‘Aspiring Environmentalists’? Identifying travel behaviour segments using attitude theory Transport Policy 12 (1) https://www.sciencedirect.com/journal/transport-policy

18. Chatterjee K et al Young people’s travel – what’s changed and why? https://assets.publishing.service.gov.uk/media/5a82a485ed915d74e3402d3e/young-peoples-travel-whats-changed.pdf

19. Anable J and Goodwin P (2018) Assessing the net overall distributive impacts of a congestion change, International Transport Forum https://www.oecd.org/content/dam/oecd/en/publications/reports/2018/10/assessing-the-net-overall-distributive-effect-of-a-congestion-charge_017287a4/7f4c51f2-en.pdf

20. Jones, P, Dix, M, Clarke M, & Heggie I (1983) Understanding Travel Behaviour, Oxford Studies in Transport, Gower.

21. My first entry to this topic was an unpublished conference paper I wrote in 1972, as a side interest before finishing my PhD thesis, called ‘Measuring the marginal utility of money’, in which I suggested interpreting the reciprocal of the value of time as a measure of the marginal utility of money. The very eminent economist Colin Clark (1905-1989), a pioneer of the concept of Gross National Product, who had come across my paper, sent me an encouraging letter and a draft copy of a revised version of his ‘The Marginal Utility of Income’ (Oxford Economic Papers, July 1973) in which he constructed an elaborate graph of the marginal utility of income defined as the reciprocal of the value of travel time, and I referred to this paper in my 1974 paper ‘Generalised time and the problem of equity in transport studies’, Transportation (3) 1-24. I don’t think his revised draft was ever published, though I still have the copy. J M Keynes described him as ‘a bit of a genius, almost the only economic statistician I have ever met who seems to me quite first class’. I wish I had met him.

22. Harrison, A.J. and D.A. Quarmby (1969), Theoretical and Practical Research on an Estimation of Time-saving: Roundtable 6, European Conference of Ministers of Transport, OECD, Paris.

23. McIntosh, P.T. and D.A. Quarmby (1970), Generalised Cost and the Estimation of Movement Cost and Benefit in Transport Planning, MAU Note 179. Department of the Environment, London

24. Goodwin P. (1976), Value of Time, Roundtable report number 30, European Conference of Ministers of Transport OECD Paris.