The Dublin Taxi Market: 
Re-regulate or Stay Queuing?

Explanatory Notes

The views expressed in this paper are solely those of the authors. The authors would like to thank the Carriage Office, Dublin City Centre Business Association, Dublin Corporation, the Department of the Environment and the Irish Hotels Federation and Sean Barrett for providing us with information and Philip Lane, Suzanne O'Neill, Frances Ruane, Vivienne Ryan, and others for helpful comments and suggestions.  We are grateful to QRESEARCH for help with making the HTML version of the paper available over the WWW at WWW.JOEMAXI.COM.     Any errors are ours.

Contents

1. Introduction
2. The Dublin Taxi Market
   • 2.1 System of Regulation
   • 2.2 Features of the Dublin Taxi Market
   • 2.3 Analysis
3. The Economic Literature on Taxi Regulation
   • 3.1 Price and Fare Controls
   • 3.2 Entry to the Market
   • 3.3 Quality
   • 3.4 Regulatory Capture and the Political Economy of Regulation
   • 3.5 Summary
4. International Lessons from De-regulation
   • 4.1 Prices
   • 4.2 Entry to the Market
   • 4.3 Summary
5. Public Policy towards Taxis
   • 5.1 Social Benefits of Taxis
   • 5.2 Efficiency in the Taxi Market
   • 5.3 Accessibility
6. Re-regulating of the Dublin Taxi Market
   • 6.1 Retain Price Controls
   • 6.2 Improve Quality Standards
   • 6.3 De-regulate Entry to the Market
   • 6.4 Establish a New Regulatory System
   • 6.5 Moving towards Re-regulation
7. Conclusion

• References
• A Appendix: Legislation from 1961
• B Appendix: Current Regulated Fares
• FootNotes

1. The excess demand in the Dublin taxi market is outlined.
   • Entry was regulated in 1978 (when the number of taxis was 1,835). Taxi numbers have increased only once since, by 139 in 1992 after a detailed review of the market. The number now stands at 1,976.
   • Demand for taxis has grown enormously over that period, fuelled by economic growth, especially in the recent "tiger" years (see Figure 1 on Page 6), demographic changes and increased tourism. Although taxi usage has increased, and the number of hackneys has ballooned from 800 in 1992 to over 3,000 now, excess demand for taxis persists.
   • A survey in mid-December 1995 revealed that only 23 per cent of calls succeeded in obtaining a taxi, with an average wait of 48 minutes for the taxi.
   We estimate that the market could support at least twice the existing number of taxi vehicles.
2. The cost of this excess demand is high. The cost to taxi users in terms of higher fares is in the region of 12 million pounds per annum and this does not include the cost of waiting for a taxi.¹ There is also the cost of non-use in terms of business not undertaken and reductions in public safety.
3. Public policy should encourage efficiency in the taxi market. In particular, we identify two important measures.
   • (a) First, it should recognise that in the taxi market supply needs to adjust to meet demand. For a market to develop properly, a combination of full-time and part-time employment is likely to be required (as in restaurants and retailing). Higher fares at peak times would encourage part-time or shift-time operators to supply these peaks, resulting in flexible supply of the market.
   • (b) Second, high licence fees prevent effective supply in the market, constitute an inefficient form of taxation, and ultimately encourage the use of private cars instead of public transport. The provision of an adequate taxi service also assists other policies such as the campaign against drink-driving. For these reasons, the initial licence fee and any renewal or other fees should be set at a low or zero level. The current proposal to charge £15,000 is thus ill-advised.
4. The international economic literature on taxi markets has analysed the regulation of quality, entry, and price. The main points that emerge from this literature are as follows.
   • (a) The arguments in favour of regulating entry are weak. More over, any system of regulation could create more problems than it would solve and would be an inevitable target for "capture" by the industry it regulates.
   • (b) There is less agreement as to whether the regulation of price is desirable. Those advocating regulation do so on the basis that customers cannot easily compare different fares and that this reduces the normal incentive that suppliers in any market have to compete by undercutting high prices. Attention has also focused on the minority, but often the most vulnerable, that might be exposed to exploitation. On the other hand, some argue that competition in the telephone booking market can work effectively and that customers could compare different fares at taxi ranks. We consider this more applicable to hackneys than taxis.
   • (c) There are compelling reasons why the quality of both vehicles and drivers needs to be regulated.
5. De-regulation of taxi markets has become common internationally and the actual experience of de-regulation is broadly consistent with what the economic literature suggests.
   • (a) The de-regulation of entry has had a consistently positive effect on supply, except that in some instances quality standards have fallen.
   • (b) The de-regulation of price has had ambiguous effects. This may be due to weak price competition (as some authors predict) or to catch-up because fares were previously regulated at too low a level.
   • (c) De-regulation of entry is likely to be most successful if accompanied by closer regulation of quality standards.
6. We argue that the existing regulatory system has failed and that it has prevented the market from developing. We recommend complete re-regulation of the Dublin taxi market. Briefly this would comprise.
   • (a) De-regulation of entry to the taxi market.
   • (b) Higher quality standards and better enforcement to cope with the increased supply of taxis that would result from de-regulation.
   • (c) An improved system of complaints.
   • (d) The retention of the control of taxi fares.
   • (e) A new system of independent and transparent regulation with the clearly stated objective of regulating fares so as to ensure no excess demand in the market.
   Such a system would bring enormous private and social benefits to consumers, business and tourism in Dublin and would increase employment as the taxi market expanded. We consider that the cost of improving the regulatory system would be an efficient use of public resources.
7. With regard to the de-regulation of entry, we advocate a staggered approach to increasing the number of licences with the following features.
   • (a) A second licence should be given free to each existing licence holder.
   • (b) Licence holders who purchased their licences within the last 5 years should receive this second licence immediately. The remaining licence holders should receive a second licence after 2 years.
   • (c) There should be no restrictions on the trading of taxi licences, subject to buyers meeting quality standards.
   At the end of 5 years, full de-regulation of entry should occur and a licence should be available to any qualified applicant. This approach has the advantage of compensating the existing licence holders on average in proportion to any perceived loss to them resulting from deregulation.
8. We do not consider that compensation should be paid to taxi licence holders. A taxi licence has economic value solely because of the monopoly profit it enables the holder to earn in its use or by its sale. It would be contrary to good economic policy to reward monopolists in this way.
9. The services provided by taxis and hackneys are substitutes for some users but not for others. As a result, regulation should continue to permit both services to co-exist. Because of the limited substitutability and the use of telephone booking for hackneys, there is no compelling reason to regulate prices in the hackney market. In addition, continued regulation of taxi fares will ensure greater competition in the hackney market.
10. The question of taxi accessibility for persons with disabilities has not been adequately addressed in Ireland. This issue is beyond the scope of this paper, but there is a need to formulate overall policy in this area and we outline several options for discussion. However, regardless of any new policy development, de-regulation of entry in the market at present would have a significantly positive effect on meeting the demands of those with disabilities. The current excess demand for taxis encourages cherry-picking (refusing less economic fares) so that the supply of taxis to those with disabilities is even worse than for the remainder of users. With an excess supply of taxis in the market, those (few) taxis that are accessible would be willing, if not encouraged, to supply a service to those with disabilities.

1 Introduction

In order to make recommendations about the taxi market, we first outline the international economic literature on regulation and, in particular, that work which pertains specifically to taxi markets. We then examine the experience of de-regulation in cities abroad which had previously regulated their taxi markets in a similar way to Dublin. The results of these investigations suggest that there is little rationale for regulation of the number of taxi licences and that where the restriction on numbers has been removed, there has been a dramatic improvement in quality of service in terms of reduced waiting times for customers. However, the doubt expressed by many authors concerning the ability of price competition to operate in the taxi market is evident in that the effect of de-regulation on prices has been ambiguous. De-regulation has also led to concerns about quality of vehicles and drivers and experience suggests that any de-regulation of entry or price needs to be accompanied by re-regulation of quality standards.

In section 5, we address a number of issues that relate to government policy towards the taxi market and the role of the taxi service in overall transport policy. The questions of peak loading and flexible supply have implications for using part-time or shift employment in the taxi market. We outline how efficiency of matching relates to the level of fares and argue that the taxi market should not be used as a source of government revenue, no matter how tempting. We also examine such questions as the distinction between taxis and hackneys, drink driving and the accessibility of taxis for persons with disabilities.

In Section 6, we argue for a comprehensive re-regulation of the market,

              Table 1: Regulatory Functions and Agencies Resposible

involving tighter quality standards, increases in taxi numbers and a new and more responsive and efficient mechanism for regulation of the market. In recognition of the political reality that the de-regulation of entry to any market raises difficult issues, we examine a number of different options for such de-regulation. We suggest that the best approach would be one of staggered de-regulation, in which new licences are awarded to existing taxi drivers over some years according to how recently they purchased a licence. This would compensate existing licence holders and simultaneously ensure a strong commitment to full de-regulation.

2 The Dublin Taxi Market

2.1 System of Regulation

• Quality:

  There are quality standards for both drivers and vehicles. Vehicles must be approved both in terms of size and seat capacity and in terms of their safety and roadworthiness.⁵ Drivers must possess a Public Service Vehicle (PSV) license which requires a once-off test of driving skills and a written test of knowledge of the city.⁶ These standards were introduced at a national level by 1961 legislation and, for Dublin, are enforced by the Carriage Office (the Garda Commissioner).⁷ Regulation thus focuses on the technical standards of the vehicle and driver and not on more general quality issues such as comfort of seating and cleanliness. Because of enforcement by the Carriage Office, court-quality evidence would be required in order to remove a sub-standard operator. Metering is regulated by a different agency.

• Prices:

  The prices (or fares) that taxi drivers can charge are strictly limited. This is done by specifying the fare per distance travelled, the fare per minute waiting, and other fixed charges such as that for baggage, unsociable hours etc. Appendix B gives details of the price controls currently in force. The legislation allowing for price controls was the 1961 Act, and the last major change in the regulated prices was in 1985, although there have been minor changes since. In September 1995, the task of regulating prices was transferred from the Department of the Environment to to the relevant local authorities. Any change in the level of fares allowed is a reserve function of the authority which means that it must be voted upon by its elected members.⁸

• Entry:

  Restrictions on the number of taxi licences were introduced in 1978.⁹ The number of licences currently stands at 1,974. Licence numbers are decided by Dublin Corporation on its own behalf and on behalf of the other local authorities.¹⁰ Again this is a reserve function. Licence renewal fees are charged and the transfer of a licence from one operator to another incurs a transfer fee of £3,000.

Perhaps the most extraordinary aspect of these regulations is not their specific controls (which are similar to those in many other cities) but rather the very segmented structure of the overall system of regulation. Price and entry are regulated by four different authorities, each of which must vote on any important measure, while quality standards are determined by the Carriage Office. In consequence, the effective regulation and management of the market would require both highly sophisticated co-ordination between the different agencies and unity of purpose among them. The system does not appear to provide for such coordination and the regulation that results is likely to be inefficient.¹⁵ More generally, there appears to be no formal link between the regulation of the taxi market and public transport and infrastructure for the city generally. For this reason, the present system does not permit proper management of the taxi service in Dublin.

2.2 Features of the Dublin Taxi Market

                      Table 2: Taxi Licences in Dublin since 1970

Taxis 1,389 1,232 1,395 1,437 1,561 1,664 1,835 1,974

Source: Dublin Corporation.

The effective supply of taxi services has, however, increased over the period as taxi usage has increased. There are approximately 12,000 licensed drivers and, at present, it would appear that most taxis licences are worked to full capacity with two or more drivers operating each taxi.

Annual data on numbers of hackney¹⁸ licences are not available, but their number has increased dramatically over the last 5 years from a figure of about 800 in 1992 to over 3,000 today.¹⁹ Since August 1997, there has been a moratorium on new hackney licences which is likely to aggravate the problems of excess demand.

Because the supply of taxis is restricted, we cannot observe or measure the demand for taxi services. Any estimate of demand must, therefore, be based on proxy variables that are correlated with demand for taxis.

Gross domestic product measures the overall level of economic activity and this has more than doubled since the ceiling on taxi numbers was introduced in 1978. Figure 1 graphs an index of the number of taxis and GDP (setting the 1970 values at 100 for ease of comparison). It illustrates how this measure of demand has out-stripped supply over the period. As noted above, effective supply is higher than the number of taxis because taxi productivity has increased. However, demand may also be underestimated because taxi productivity has increased. However, demand may also be underestimated because prices have not changed, demographics have changed and people may spend a higher proportion of income on taxis. Simple extrapolation based on GDP and other conditions in 1978 would suggest that 4,200 taxis are needed to supply the current level of demand.

Demographic changes have also contributed to increased demand for taxis. Table 3 shows the population of Dublin (in thousands) in 1977 and 1996. The population has increased by 121 thousand (13 per cent). However, within the age cohorts of 25 years of age and upwards that are most likely to use taxis, there has been a 65 per cent increase in population from 378 thousand to 626 thousand. Thus a GDP-based estimate of demand might underestimate the true demand. These population figures do not include population expansion around the periphery of Dublin which could also have increased the demand for taxis. Changes in economic activity that have accompanied these demographic changes, such as the increases in nightclubs in the city centre, may also contribute to extra demand for taxis.

               Table 3: Changes in Dublin Population: 1977-1996

1996 234 197 326 199 101 626 1057

1977 291 267 165 145 68 378 936

Change -57 -70 161 54 33 248 121

% change -20 -26 98 37 49 65 13

Source: Labour Force Survey Estimates of the Population of Dublin. Figures are in thousands.

Table 4: Passenger Numbers at Dublin Airport: 1979-1996

Year 1978 1996 increase % increase

Airport Traffic 2.5m 9.1m 6.6m 262

Source: Aer Rianta.

Table 4 indicates that traffic at Dublin Airport has almost quadrupled in the last 19 years from 2.5 million passengers in 1978 to over 9 million in 1996. This is another factor which might suggest that the type of economic growth which the Dublin economy has experienced might be disproportionately biased towards activities which require the use of taxis.

The other source of information on excess demand for taxis is the extent of taxi availability in the city. A private survey²⁰ during 1990 and 1991 found that ranks were often vacant, particularly at peak times but that taxis rarely had to wait for a customer. On 15th December 1995, Dublin Corporation surveyed taxi availability between 10am and midnight by telephoning companies and taxi ranks. They made 126 calls in 14 hours. Although this was a Friday on a busy day of the year, the results still indicate enormous excess demand for taxis. Only 23 per cent of calls succeeded in obtaining a taxi and the average expected waiting time was 48 minutes. Of the other 77 per cent of calls, 33 per cent had nothing available and 44 per cent either engaged or did not answer. In only one instance was there a taxi waiting at a rank. This was almost exactly two years ago and it is clear that queuing for taxis continues to be a problem. ²¹

Although the existence of queues of people waiting for taxis is evidence of excess demand, the length of those queues does not provide information about the size of that excess demand. This is because some people who would use a taxi may not be willing to do so if it involves queuing.

Another feature which illustrates the excess demand for taxis is the transfer value of a vehicle licence. No official record is made of the price at which transfers occur (although records of transfers are made). Informal data suggest that this steadily increased during the 1980s (see Barrett (1991)) and is now in the region of £70,000 to £90,000.²² This suggests a monopoly profit in the market in the region of £150 million (see Table 5 on page 11 below). This monopoly profit is entirely due to the restriction in the number of licences.

Although the number of hackney licences has increased in response to this excess demand, hackneys are not perfect substitutes for taxis.²³ Hence excess demand for taxis is likely to worsen unless the supply of taxis is increased.

It is difficult to measure quality of service in the taxi market, but there have been complaints of both cream-skimming and overcharging (especially from the airport). The 1990/1 survey found that the quality of vehicles was generally good but that there was a significant minority of poor quality vehicles. There is no indication that this situation has improved.²⁴ It is not clear that the quality of vehicles and the standard of hygiene within them, although perhaps meeting the technical standards discussed above, are at the level that consumers would otherwise choose.

There is currently a trend towards the charging of increased licence fees by Dublin Corporation. The renewal of a taxi licence (required every two years) incurs a charge of £450 since May 1997 compared to £100 before. At the time of writing, there is a proposal to increase the number of licences by 200 and each of these licences will cost £15,000, compared with £100 for those issued in 1992.²⁵ This trend is maintained with regard to hackneys where the initial fee for a new licence increased from £100 to £1,000 in May of this year.

2.3 Analysis

The motivation of the original 1961 legislation was to protect consumers by regulating quality of vehicles and drivers, controlling prices and establishing taxi ranks. It would be difficult to argue against the desirability of quality standards in the taxi market. Controls on prices might also be appropriate, especially in a newly emerging market, where there might be fears of over-charging (discussed in Section 3). Moreover, the use of taxi ranks is a sensible measure that reduces mis-match in the market. In fact, the current problems in the taxi market appears to have their origin solely in the restriction of numbers in 1978.

The introduction of entry regulation followed a 1976 report on taxi and hackney services in Ireland commissioned by the National Prices Commission and undertaken by Hyland Associates (1976). They found acute public dissatisfaction with the quality of the taxi service with complaints of creamskimming (refusing uneconomic hirings), over-charging, broken meters and high fares. They accepted the arguments put forward by taxi drivers and recommended that the number of licences be restricted. They rejected a proposal that higher fares be allowed at peak times on the basis that this would create an incentive to concentrate services at peak times at the expense of other times.²⁶

Recommendations for re-regulation of a market that do not have regard to the future development of the market quickly become outdated. This happened to those of Hyland Associates (1976). The consequences are serious given how difficult it can be to alter such regulations. However, even judged by contemporaneous standards, their recommendations are inadequate and might illustrate the phenomenon of "capture" whereby regulation is determined by and on behalf of the regulated industry. In particular, concerns about quality could have been addressed by a tightening of quality standards and improvements in their enforcement. The fact that the price in a market will tend to rise at peak times was also true in 1978 and a regulatory system that prevents this from happening is not acting in the interests of consumers.

As Figure 1 on page 6 indicates, the introduction of entry controls was extremely effective and the taxi shortage that ensued led to a new review, this time by an Inter-Departmental Committee which reported in 1992. This group also concluded that the removal of entry controls would impair quality standards. In reaching this conclusion, it had examined evidence from UK cities where entry de-regulation had resulted in reductions in quality of service.²⁷ However, it did recommend that the actual number of licences be increased and, as a result, 139 new licences were issued.

The Inter-Departmental Committee (1992) recognised that taxis and hackneys provide different services and recommended that licensing arrangements for both be provided. Although it recommended maintenance of regulation of price and entry in the taxi market, it did not consider such regulations appropriate for the hackney market. As Figure 1 indicates, there has been an explosion in the number of hackneys in the intervening period.²⁸

The most serious lesson from this outline is that regulation of entry to a market is extremely difficult to fine-tune and almost impossible to reverse if it is not implemented successfully. By offering monopoly profits to existing suppliers in the market, it establishes a natural opposition to any increase in numbers, so that the system fails to respond to demand. This was a problem, even before the "tiger" years, but has been chronically exacerbated by the strong growth of Dublin city this decade. In effect, the city struggles with a taxi infrastructure and level of service that was considered appropriate, by taxi drivers, for conditions some 20 years ago.

Estimating the extent of excess demand is difficult. Simple extrapolation based on GDP would, as noted above, suggest a figure of 4,200 taxis.²⁹ However, demographic change and the growth of tourism might suggest that demand is considerably higher. In addition, more taxis would be needed if capacity utilisation fell. A figure in excess of 4,000 taxis is also consistent with the fact that a combination of 2,000 taxis and 3,000 hackneys cannot satisfy the current excess demand in the city.³⁰ Hence it is plausible that a de-regulated market might be supplied by over three times the existing number of taxis.

The effects of this regulatory failure is considerable and can be broken down into the cost on taxi users and the cost of non-use of taxis. The cost on taxi users takes the form of high fares and high waiting times. Table 5 indicates the monopoly profits in the Dublin taxi market, based on two possible values for licences on the secondary market. A licence holder has to cover capital costs of 5,000 to 7,000 per annum before any wage income or operating costs are met and this represents an unnecessary charge on users of taxis. This aggregate figure is in the region of 12 million per annum. However, this only measures the cost of the fare, and does not include waiting time. The cost of non-use of taxis relates to the fact that taxi (and other) business is not undertaken and the fact that safety may be impaired. Putting a monetary value on these losses would be difficult, but it would appear that 12m per annum underestimates the extent of the problem.

The main beneficiaries are been the holders of taxi licences. In particular, those who have sold a taxi licence have converted the entire monopoly profit into liquid form. Those who continue to operate the licence obtain this profit as an annual return to the licence.

There appears to be little point in another review on the Dublin taxi market if it simply recommends that new licences be issued. First, the level of demand in the market suggests that the number of licences would need at least to double. Thus incremental change is not adequate. Second, regulators do not have the information required to know the `correct' number of licences. Third, even if we could determine this `correct' number, this would change within a short period, and the system would likely fail to catch up as at present.

   Table 5: Secondary Market Value of Taxi Plates in the Dublin Taxi Market

Total 75,000 148.2m 85,000 168.0m

Annual (8%) 6,000 11.9m 6,800 13.4m

Annual (7%) 5,250 10.4m 5,950 11.8m

This table gives the monopoly profit based on two possible secondary market values for taxi plates, £75,000 and £85,000, and gives annual profit flows for two possible interest rates, 8% and 7%. Figures are in Irish pounds.

Fundamental reform of the taxi licensing system is required. In order to examine the form that this should take, we first survey the international literature in the area and then examine the varied experiences of de-regulation in other cities. We then continue with an examination of special issues before returning to outline options for re-regulating the Dublin taxi market.

3 The Economic Literature on Taxi Regulation

3.1 Price and Fare Controls

Diamond (1971) showed how monopoly pricing could prevail in a market if customers have search costs.³² This is because a supplier who undercuts the others would not attract extra sales because it is costly for customers to search for such lower prices.³³ In the taxi market, these search costs would include the cost of waiting to find a second price, the psychic costs of saying no to a high-price driver and the risk of losing a relatively good offer.

The size of any search costs depends on the number of taxis available. If there is excess demand (scarce supply), price comparisons will be more difficult to make. The size of the search costs also depend on the location of the customer. They are likely to be high for someone hiring a taxi at a random point on the street. For a customer at a taxi rank, better price comparison might be possible, but this could be impaired by baggage or by the social convention that the customer should hire the first taxi in the rank.³⁴ If different taxis charge different hiring fees and mileage rates, customers may not be able to compare prices very easily. Something that has received little attention in the literature is the ability of customers to make price comparisons when a service is priced by formula.³⁵ Even if taxis advertised hiring rates and mileage cost on their doors, customers might not be able to compare what fare they would actually pay as they are purchasing a relatively complex multi-dimensional product.³⁶

On the other hand, for customers hiring by telephone, waiting costs and psychic costs are lower, and the main search cost is that of a telephone call.³⁷ Fares are usually quoted a composite for the journey, rather than as a formula. In addition, companies may have an incentive to acquire a reputation for setting low prices and this would work better with telephone bookings where the customer chooses the company than with taxi-ranks where the matching is random. Overall, therefore, one would expect more intense price competition for the telephone hiring market than for the markets for random or taxi-rank hirings in the street.

This point holds regardless of whether taxi users are well-informed or not about prices in the market. The second point relates to the over-charging of those who are poorly informed. This could include visitors to the city, infrequent users³⁸, or those clearly disadvantaged by heavy baggage or a journey through a less safe part of town. The distinction between over-charging and monopoly pricing is subtle. Over-charging could involve charging more than the monopoly price that would be charged to an informed customer, or could take the form of price discrimination, where people are charged according to their perceived willingness to pay. As a result, a cogent argument for price regulation has been to avoid over-charging at airports and train-stations when relatively un-informed and disadvantaged travellers are hiring taxis.³⁹

The traditional solution to these problems has been the regulation of taxi fares and the installation of meters. The regulated price depends on the distance and time of the journey plus any hiring or other fixed charges (see Appendix 1 for an example for the Dublin market). The meter allows the customer to verify the distance travelled which is usually the largest component of the fare.

Some economists contend that regulation of maximum fares is not necessary if there is free entry to the market.⁴⁰ The basis for this argument is that search costs are not high and price comparisons by customers will force taxi suppliers to compete on price. This argument is more convincing with regard to telephone hirings, particularly in the United States where the majority of hirings are by telephone,⁴¹ large taxi fleets exist, and local telephone calls have zero marginal cost.⁴²

In summary, there would appear to be some theoretical rationale for believing that competition among taxis would work in the market for telephone hirings, but less confidence in the ability of price competition to operate in the market for street hirings.

3.2 Entry to the Market

We deal first with the effects of free entry on the cost of supplying a taxi service. Excess capacity in the taxi market is valuable in that it reduces customer waiting times. Thus it does not present the same type of welfare costs as in other markets. In other words, customers would be willing to pay for some excess capacity.⁴⁵ To the extent that fixed costs⁴⁶ in the taxi market are relatively low⁴⁷, any productive inefficiency arising from excess entry would be minor. In many cities, taxi services are supplied by part-time employment, indicating that an economic return can be obtained without operating at optimal capacity. Moreover, in many markets characterised by very high fixed costs such as cement production, there are no restrictions on entry. Thus there would appear to be little reason to be concerned with excess entry in the taxi market.⁴⁸

The fact that more taxis would increase traffic congestion is merely a manifestation of a more fundamental problem with the pricing of public infrastructure. As such, it applies to private cars as much as to taxis and it would be invidious to single out taxis for special attention without a systematic approach to road pricing. Indeed, insofar as taxis provide a useful public service, it could be argued that they be favoured, as they are in many cities with access to bus lanes.

Over-crowding at taxi ranks is likely to occur if there are high fixed costs of operating a taxi and the economic return per hiring is high. This creates intense competition among drivers. Other policy measures, however, can mitigate this problem. If licence fees are set at a low or zero level, the cost of leaving a taxi idle is not so high. A reduction in the hiring fee also reduces over-crowding. We argue below that a good regulatory system should use excess supply or demand to determine whether fares should use excess supply or demand to determine whether fares should be reduced or increased. Such a system would put additional collective pressure on taxis not to over crowd taxi ranks, as this would precipitate a fare reduction.⁴⁹

The impact of de-regulation on quality standards is difficult to assess a priori, but it is possible that de-regulation of entry could reduce quality of service in the absence of quality standards. This is not an argument against de-regulation of entry, but rather an argument for improved quality standards if entry is de-regulated. In particular, the improvement of quality standards should focus on ensuring equivalent standards for part-time and full-time work.

More generally, there is a considerable body of argument against regulation of entry to markets on the grounds that such regulation cannot improve the market, even if the market fails.⁵⁰ First, the regulation of entry to an industry can have a serious impact on competition within the industry. It may facilitate collusion among suppliers, especially as this is not affected by the usual threat of entry. Moreover, if licences are not fully tradeable inefficient operators would have less reason to minimise costs.⁵¹ Thus entry restrictions could result in additional productive inefficiency, and possibly also X-inefficiency.⁵²

Second, regulation of entry will create rents in the market in the form of positive values on taxi licences. If licences are fully tradeable (see footnote 51), then the rents will be high. As a result, suppliers in the market may engage in socially wasteful rent-seeking behaviour in order to prevent re-regulation of the market when this becomes necessary. Third, as costs and demand conditions changed over time, the regulatory system might not adjust, creating further welfare losses.

There is much evidence in taxi markets of the validity of these concerns with regulation of entry. The fact that taxi licences trade for positive (and sometimes high) values is direct evidence that the restriction of entry leads to prices above cost and illustrates the level of economic rents and market power created by this licensing.

Overall, therefore, the theoretical literature suggests that regulation of entry is likely to be a poor policy but that regulation of prices may be required. Some authors (Beesley (1973) ) who have developed specific models conclude that price regulation without entry regulation is desirable. Beesley & Glaister (1983) go further and argue that regulation of entry is a bad policy. De Vany (1975) finds that free entry is better, provided that price regulation is set at the right level, while authors such as Hackner & Nyberg fail to consider the political economy arguments against regulation when they argue for entry controls.

In summary, there would appear to be little or no public interest benefit in restricting numbers in the taxi market. If anything, restrictions on entry may be detrimental to the public interest, resulting in sub-optimal supply.

3.3 Quality

The main issues are the level of the qualitative standards that are set and how the regulation of quality responds to changes in the market (such as expansion). We note two points. First, higher quality standards will reduce the number of suppliers in the market. Hence any qualitative regulations have inherent quantitative implications, especially if there is free entry to the market. In other words, higher quality standards increase costs and, to keep supply at a given level, prices would have to rise. Second, de-regulation of price or entry would require either higher qualitative standards or tighter enforcement of existing standards. Anything that makes the market more keenly competitive will increase the incentives of suppliers to skimp on quality. Thus one would expect de-regulation of price or entry to lead to a reduction in standards if no change in standards was made.

3.4 Regulatory Capture and the Political Economy of Regulation

For the taxi market, the regulator might need to have information about costs in order to regulate price or entry. If the suppliers of taxi services convinced the regulator that costs were higher than they really were, then the regulator might implement a higher fare structure, giving the industry some monopoly profit or rent. Capture could also occur with regard to entry, where the interest of the industry to restrict entry would conflict with the consumer interest in having free entry.

The possibility of regulatory capture means that regulation may not improve on the free market. This does not mean there should be no regulation. Rather, regulation should only be used where strictly necessary and, in such cases, the institutions of regulation should be designed so as to minimise the possibility of capture. Several features of the institution would reduce the possibility of capture. First, the regulator should be given a clear objective: for example, to regulate price in the interests of consumers. Second, the process of regulation should be completely transparent. Thus any arguments presented to the regulator by the industry should be public, and any decisions of the regulator should be accompanied by published reasoning that explains how the decisions relate to the stated objectives.⁵⁴

A third issue that has implications for the political independence of the regulator is the political economy of regulation. Consider a regulatory measure that results in less profits for an industry but gains for consumers that, on aggregate, outweigh the losses to industry. An example might be a reduction in local telephone charges. Such a measure would be in the common good and so it is desirable that it be undertaken. However, if market variables are determined by the strength of political or lobbying power, then it is unlikely that the correct levels will be chosen. This is because the industry always consists of a small, focused and well-organised group, whereas the consumers who benefit are typically a large and diffuse group. Each firm in the industry has a lot to loose from tighter regulation and will exert great effort against the regulation. Each consumer, by comparison, has only a small gain and, because consumers are less organised, the greater aggregate gain of consumers will not be reflected in any lobbying. Moreover, such lobbying is socially wasteful, and a system of regulation that encourages it would compound the problems.

The political economy analysis of regulation recognises the importance of economic vested interests in determining political outcomes. One solution is to have a regulator that is independent from the political process (but answerable to politicians). Given this independence combined with transparency, the regulator can implement measures that are good for society. Moreover, the industry has less incentive to lobby because it knows that it is less likely to be successful. Politicians quite properly appoint the regulator, and insist on transparency of process but they are removed from the day-to-day decisions.⁵⁵

The taxi market exemplifies these political economy arguments. The suppliers in the market are a relatively small, focused and highly vocal group whereas consumers comprise a large and diffuse population.⁵⁶ A system of regulation based on decision-making by elected representatives is likely to give rise to a large amount of lobbying and inevitably decisions that reflect the interests of taxi suppliers. Thus good regulation necessitates political independence and transparency.⁵⁷

3.5 Summary

With regard to entry, de-regulation would be expected to increase supply and reduce waiting times. This increases the regulatory burden with respect to quality and management of the market and successful de-regulation of entry would need to be accompanied by measures to ensure quality standards and the efficient organisation of the market. We now go on to examine the experience of de-regulation where it has occurred.

4 International Lessons from De-regulation

4.1 Prices

Authors who have studied the effects of de-regulation on price have without exception (to the best of our knowledge) reported increases in nominal fares. Teal & Berglund (1987) analysed the effects of de-regulation in seven US. cities with population above 250,000 and found fare increases in all except Sacramento. Their benchmark was the level of fares in regulated cities and they found that fare increases in de-regulated cities were higher than in those that maintained regulation.⁶¹ This suggests some level of monopoly pricing may have occurred. Fares in Sweden increased in real terms by varying amounts, except in medium-sized municipalities where fare per distance decreased. Fares in New Zealand fell in real terms but increased in nominal terms. Morrison (1997) analysed the two largest taxi companies in the Wellington market and found that the consumer was paying less in real terms in 1994 than in 1989 when the market was de-regulated. These results are supported by at least four surveys of fare changes which have been conducted in Wellington since de-regulation. For over 60 per cent of companies fares for certain journey types increased nominally but fell in real terms.

Another effect of de-regulation is the creation of price dispersion in the market. We distinguish between price dispersion that is common to all suppliers (for example, higher prices at certain times of the day) and price differences among suppliers for identical products.

In terms of the variation in prices that is common across the market, the main distinction would appear to be between street and telephone hirings, with a greater increase in the price of street hirings relative to telephone hirings. This effect is particularly pronounced at (but not confined to) taxi ranks in airports where higher fare increases were found in San Diego, Seattle, Phoenix and New Zealand.⁶² This is consistent with the theory that suggests price competition in the market for street hirings is likely to be weaker. Other variations in the fare according to the time of day and the number of passengers were noted but, taken in isolation, these offer little insight into the proximity of fares to underlying costs.

The variation in the prices charged by different suppliers for identical products may offer greater insight. In an intensely price competitive market, such dispersion would be low, although perhaps not quite at the zero level imposed by regulation. However, if suppliers offer different services, then greater dispersion would still be consistent with keen competition. The main dimension in which the service could be differentiated in the taxi market would be if some companies offered lower waiting times.

The evidence is that de-regulation has led to large increases in price dispersion among companies. In both Seattle and San Diego the second largest company offered fares 15 per cent below those of the market leader (for telephone hirings). For New Zealand, Morrison (1997) looked at six companies operating standard taxi services in Wellington in 1994. Fare schedules varied with four companies offering a hiring charge of $2, one $1.40 and another $1.30. Morrison, controlling for differences in waiting times, calculated the prices of three journeys for each company and found variations of up to 34 per cent between the different companies for identical journeys.⁶³ This suggests that average prices in the market are above the competitive level and that price-cutting does not increase market share.⁶⁴

Overall, the evidence on price de-regulation is not overwhelming in support of reduced prices. Most authors who report increases in nominal prices offer little analysis of how those new nominal prices relate to underlying costs. In particular, most authors appear to have ignored the fact that a fall in the value of taxi plates would exert downward pressure on prices.⁶⁵ Where meaningful comparisons are possible, they suggest that monopoly price-cost mark-ups may be established or persist after de-regulation of price.

4.2 Entry to the Market

In New Zealand the effects were more dramatic with a tripling of taxi companies (200 per cent increase) in some locations and a doubling of taxi numbers in Wellington. In all cities waiting times fell significantly. Again the price of taxi plates on the secondary market in these cities was considerably lower than in the Dublin market at present. For example, a taxi licence in New Zealand traded for about NZ$25,000 (approximately IR£10,000).

Several authors have examined the effects of de-regulated entry on quality and level of service. Several types of reduction in quality are reported. In San Diego and Seattle, trip refusal/no-show rates increased suggesting cream-skimming. The Inter-Departmental Committee (1992) examined four anonymous UK cities and reported over-crowded and illegal taxi ranks. Other concerns have focused on driver knowledge of the city and on language and communication problems. However, these authors do not examine whether entry de-regulation was accompanied by active measures to maintain or improve quality.

   Table 6: Changes in Taxi Supply following De-regulation of Price and Entry

New Zealand: Auckland 120 New Zealand: Christchurch 300 New Zealand: Dunedin 150 New Zealand: Wellington 105 320 Sweden: large towns 16 Sweden: medium towns 25 Sweden: small towns 10 US: Indianapolis 7 US: Kansas City 18 US: Phoenix 83 US: Oakland 38 US: Sacramento 56 US: San Diego 127 US: Seattle 33 50 US: Tucson 33

In New Zealand quality standards improved with de-regulation. Topographical (i.e. route-finding) tests were introduced and some companies instituted internal levels of quality control over drivers with training programmes. This resulted in the emergence of a private driver-training academy. Enforcement powers were also increased with compliance officers being given permission to remove unsatisfactory drivers and vehicles from the road immediately. The New Zealand Ministry of Transport⁶⁷ in 1991 reported increases in service quality with consumers being provided with more information Morrison (1997} reported a wider range of niche services and better geographic coverage.

Few authors give attention to the issue of part-time drivers, although de-regulation in New Zealand was accompanied by an increase in part-time employment. In London, where entry is not regulated but fares are controlled, all drivers must take the same "knowledge" test. The market is supplied by a large proportion of part-time drivers.⁶⁸ There is no evidence to suggest that this has impacted negatively on quality.

Entry de-regulation affects the concentration of supply in markets where taxis are owned by companies. For the US, taxi companies consolidated or increased their share of the market. For example, Teal, Berglund & Nemer (1984) found that 40 per cent of independent cabs in the airport market in Phoenix left the industry during the 15 month period after de-regulation and, in San Diego, one third of all taxi operators not affiliated to the two largest companies left the market within 18 months. On the other hand, new companies emerged in New Zealand, with the number of taxi companies in Wellington increasing from 5 in 1989 to 21 in 1994 (Morrison) indicating a reduction in concentration.

Vehicle productivity (i.e. vehicle kilometres carrying a customer relative to total vehicle hours supplied) declined with de-regulation of entry in all cases. This effect was particularly pronounced in cases where overcrowding emerged at taxi ranks as these taxis had a low proportion of fares to time searching for fares.

Entry de-regulation has typically eliminated the secondary market for taxi licences, with the price of a licence being determined solely by the administrative fees set by the licensing authority. Although de-regulation does cause a once-off reduction in the value of plates, there is no evidence on the effects of de-regulation on the earnings of suppliers. In New Zealand, no compensation was offered to existing plate holders.

Finally, a consistent feature of regulation of entry is the development of hackney-style services to meet the excess demand for taxis. This has happened in New York (where the number of taxi licences continues to be regulated) and in the United Kingdom. De-regulation in UK cities led to a reduction in the demand for hackney services, with evidence that many hackneys converted to taxis.

4.3 Summary

5 Public Policy towards Taxis

5.1 Social Benefits of Taxis

Public policy towards transport usually aims to encourage the use of mass conveyance (buses etc.) rather than private cars which impose greater social costs in terms of congestion and pollution. This has traditionally been achieved by subsidies to cover fixed costs so that marginal cost pricing can encourage higher usage of public transport.⁶⁹

This argument does not apply directly to taxis, but public policy may wish to encourage taxi usage for several reasons. First, for many consumers, transport by private car may be a closer substitute for a taxi than public transport. For this reason, taxis and mass public transport may be complements rather than substitutes.⁷⁰ For example, a bus ride in one direction might be matched by a taxi journey in the other. A better provision of taxi services would then increase usage of public transport.⁷¹ It might also reduce the pressure on city centre parking.

Second, the taxi market has an important role to play in promoting safety. Most importantly, legislation against drink driving will not be effective if the taxi service is poor. For many, public transport is not available as an alternative to self-driving. Even if it is available, the fact that it does not deliver to one's home makes it an undesirable substitute (especially for women). As a result some people who drink will substitute towards private cars, not because it is cheaper but because there are no realistic alternatives. This imposes a high cost on society in terms of road safety. Another safety argument for more taxis relates to intense competition among customers for taxis late at night. This scramble for taxis might lead people who have been drinking taking excessive risks in order to obtain a taxi. Taxis may also assist generally in crime prevention and detection, providing a loose network of information for the police.

Third, there is a category of people for whom taxis are the only possible means of transport. The reason why these people have no access to other forms of transport often relates to a fundamental disadvantage such as infirmity, disability which prevents them from driving or using mass public transport. Government policy that encouraged the use of taxis would benefit such groups disproportionately. This would have positive distributional effects, especially if the disadvantages were correlated with low income. Massey & O'Hare (1996) note that these groups suffer disproportionately from longer waiting times in regulated markets and thus are the people who benefit most from de-regulation of entry.

These arguments suggest that public policy should favour taxis over private cars. This is already done in one important respect, namely that taxis in Dublin as in many cities have access to bus lanes. This preferential access to public infrastructure represents an implicit subsidy to taxis. However, it might be argued that the state should offer further preferential treatment to taxis. This could take several forms. For example, the costs of administering the licensing system could be borne out of general taxation, rather than directly within the market. Arnott (1995) goes further and argues that taxis should be subsidised in order to attain marginal cost pricing, but this might not be an effective policy measure.⁷² However the fact that taxis provide social benefits suggests that the authorities should not use the market as a source of taxation revenue. In other words, fees for licences should at a maximum cover the administrative costs of the licensing system. The debate could then be focused on whether some of the costs of administering the licensing system should be borne out of general taxation thereby reducing further the fixed costs of taxi supply and enabling greater usage of taxis by having prices closer to the marginal cost of providing the service.⁷³

Favouring the use of taxis imposes, in turn, some public service obligation on taxis, namely the provision of an adequate and comprehensive service. Any re-regulation of the market should have regard to the provision of such service and the public service obligations of the taxi market.

Public transport can, in the other direction, generate unnecessary fluctuations and biases in the demand for taxis. This happens in Dublin with buses. Because public transport is poor between 11.30pm and 6.30am, this creates a peak, one-way demand for taxis. Similar problems arise because the service from the airport to the city centre is poor and because many bus routes have the city centre as their termini.

5.2 Efficiency in the Taxi Market

What do we mean by the efficient supply of taxis?

First, it is necessary that supply be adequate to meet the demand. A service, by definition, is produced at the time of its consumption. In markets where demand fluctuates (for example, restaurants, electricity, transport), the level of capacity required to satisfy peaks in demand will be excessive when demand is peak to off-peak times.⁷⁴ Examples include night-rate electricity and early-bird menus in restaurants. If fixed costs are not high, as in restaurants, then supply can adjust to meet demand and meals provided when people wish to consume them.

Two particular characteristics of taxis make it a market in which supply would adjust to meet demand. First, demand is very time-specific so that consumers may not easily substitute a taxi at 2pm say for one at 1pm. This means that peak-load pricing alone will not create smoothing of demand. Second, with low or zero licence fees, fixed costs would be low and unused capacity would not be expensive. Flexible supply in the taxi market can be achieved through working in shifts and, if necessary, supply by part time drivers, both of which would alter the intensity of car usage. Public policy should encourage such measures in the market and, in particular, should not impose high licence fees on taxis.

At present in Dublin (see page 3 above), regulations prevent efficient supply, both by the restriction of taxi numbers but also by the requirement that drivers be available for full-time work. High licence fees also discourage flexible supply by making unused capacity more expensive. This adds further to the arguments made in the previous sub-section for low or zero administrative fees for taxis.

The second aspect of efficient supply relates to taxi productivity.⁷⁵ As noted above it is inevitable that there will be empty taxis and waiting customers simultaneously in the market. Empty taxis mean that costs are higher and government policy should attempt to ensure that this mis-match is minimised, without having excess demand for taxis in the market.

This can be achieved by common-sense measures that make it easier for waiting customers to find empty taxis. Examples include clearly visible taxi ranks where customers and taxis may meet, requiring taxis to be visible, permitting hirings while cruising, and facilitating radio contact. In this way, taxis and customers will meet more quickly and the proportion of two-way journeys will increase. In addition, policy towards mass transport should have regard to its effects on the fluctuations in the demand for y increases inefficiency by reducing the possibility of two-way trips.

Policies allowing both taxi ranks and street hirings make taxis more efficient than hackneys because they have a higher proportion of two-way traffic. If hackneys are used because taxis are unavailable, then the system would be inefficient, with higher costs ultimately borne by consumers. This is undoubtedly the case in the Dublin market at present (see Section 2 at page 6 above). The best policy to encourage the use of taxis would be to make them available, that is, to de-regulate entry.

So what should policy towards the hackney market be? In a de-regulated market, many existing hackney users would switch to taxis (and many hackneys would switch to taxis) and this would represent a cost saving to society. However, there would continue to exist a distinct demand for unmarked cars, especially by regular users and for specific purposes. Thus the system should permit hackneys and taxis to co-exist, and should not limit the numbers of each.

5.3 Accessibility

Accessible vehicles are unlikely to be provided by the market unless some incentive is given. This incentive could be given in a number of ways.

1. A regulation requiring all taxis to be wheelchair accessible,⁷⁶ with or without some public subsidy to cover the cost.
2. Voluntary provision of wheelchair accessible vehicles which would require public subsidy as an incentive. Depending on the level of the subsidy, some proportion of taxis would be wheelchair accessible.
3. Direct subsidy to the disabled accompanied by allowing higher regulated fares for wheelchair hirings.

The second option would be less costly but might enable cream-skimming as those hiring would have to specify in advance that they required an accessible taxi. On the other hand, if entry to the market were de-regulated, then accessible taxis would at least be prepared to deliver a service, unlike the present situation where they can occupy the entire day's supply of service with custom from able-bodied customers.

The third option is demand-based, and involves giving taxi customers the resources to pay for the extra cost involved. Higher revenues would encourage taxi drivers to acquire accessible taxis and, amongst those with such taxis, there would be intense competition for disabled customers, reversing the cream-skimming phenomenon in favour of those with disabilities (rather than just neutralising it). It would also be consistent with existing policy which subsidises the purchase and use of private cars by wheelchair users. It might, however, appear discriminatory and administration might be difficult. However, the use of modern technology might overcome both of these potentially serious objections.⁷⁸

We do not propose to make recommendations in this regard, but rather to outline some options for addressing an important issue which is related to the regulation of the taxi market. One thing, however, is clear. As long as there is chronic excess demand for taxis, those who are disabled are likely to suffer most. Conversely, de-regulation of entry in the market will have an immediate and positive effect on the ability of persons who are disabled to hire taxis that are accessible, regardless of longer-term measures to improve accessibility.⁷⁹

6 Re-regulating of the Dublin Taxi Market

1. Price controls should be retained;
2. Quality standards should be improved;
3. Entry restrictions should be eliminated completely over time; and,
4. A new regulatory system should be established to regulate both price and quality.

6.1 Retain Price Controls

With regard to hackneys, however, there appears to be less basis for regulating price and we recommend that fares in this market remain unregulated. The fact that customers book by telephone and in advance and are often repeat users means that price competition may be more effective. In addition, if taxis are in plentiful supply, this will constrain the ability of hackneys to charge high prices.⁸⁰

6.2 Improve Quality Standards

1. A new approach to quality standards in the market should be intro duced. This should focus not just on the technical safety of the car and driver but also on other characteristics that affect quality of service. These might include the age, comfort and hygiene of vehicles and the knowledge and interpersonal skills of taxi drivers.
2. A new complaints procedure should be introduced for dealing with al legations of over-charging, cream-skimming or poor quality of service. This procedure should be sufficiently simple that the most disadvantaged and tourists could have easy access to it.⁸¹ In particular, creamskimming, opportunistic exploitation or meter tampering should be dealt with severely as they damage public confidence in the market.
3. More resources should be allocated to the enforcement of regulations. This is particularly important in the transition period when the supply of taxis will increase substantially.
4. Similar standards should apply to both taxis and private hire vehicles.
5. The issue of accessibility of taxis should be addressed.

6.3 De-regulate Entry to the Market

We therefore suggest that there should be no specific restrictions on the number of taxi licences and that a taxi licence should be given to any suitably qualified applicant. Similarly, there should be no restriction of hackney numbers and the current moratorium on hackney numbers should be lifted.

With regard to the conditions under which licences are awarded, we would recommend the following.

1. Licences should be issued to qualified applicants regardless of whether they intend to work full-time or part-time in the market.⁸²
2. The licence fee should, at most, be set to cover the costs of adminis tering the regulations and that there should be absolutely no element of taxation in the licence fee. In contrast, some consideration should be given to the subsidisation of the administrative costs out of general taxation in order to keep the licence fee as low as possible.⁸³

6.4 Establish a New Regulatory System

The existing system of regulation is deficient in that it is segmented, poorly coordinated and incapable of adjusting supply as the market grows, with the result that the taxi market in Dublin is poorly managed and inappropriately regulated. A new regulator or regulatory agency needs to be established with the following features.
1. The regulator should have overall responsibility for the market, including the setting of controls on fares, the issuing of licences to qualified applicants, the maintenance of quality standards and the role of the taxi market in the overall transport infrastructure of the city. The regulator might use the Carriage Office (or other body) to monitor technical standards, but this would only be one component of overall quality regulation.
2. The regulator's objective should be clearly stated. An objective which

   was to ensure no excess demand for taxis (at any time of the day, week or year) would maximise the welfare of consumers. This would require a procedure for measuring excess demand or supply and adjusting hiring charges to bring forth more supply at times when it is needed.⁸⁴ In this way, part-time operators might be encouraged to supply the market at peak times.

3. The regulator should be free to make regulatory decisions independently of the political process, but should be politically accountable. He or she would be appointed by national or local government and would be required to make an annual report outlining his or her decisions. In particular, suppliers of taxi services should have no role in any aspect of the appointment or operation of the regulator.
4. The procedures adopted by the regulator should be open and transparent. In particular, there should be a process by which interested parties can make their views known and any regulatory decision should require written and published reasoning that makes clear how the decision relates to the regulator's objectives.
5. The regulator should operate a new complaints system so that the information on the market is concentrated in the regulator's hands.
6. A complaints system should be established with penalties for breaches in the standards. In particular, cream-skimming, opportunistic exploitation or meter tampering should be dealt with severely as they would damage public confidence in the market. Again, this should apply to both taxis and private hire vehicles.

6.5 Moving towards Re-regulation

It would be wholly inappropriate for the government to compensate existing license holders directly for the value of their plates. The value of a taxi plate on the secondary market reflects a monopoly profit in the market. Government policy (following from our membership of the European Union and as enunciated in the 1991 and 1996 Competition Acts⁸⁵) toward monopolies in markets with a single monopoly supplier is not just to eliminate this monopoly profit, but also to impose heavy fines (up to 10 per cent of turnover) for abuse of any such dominance. In this case, the only difference is that the monopoly position is occupied by a group of suppliers. De-regulation accompanied by such compensation would be equivalent to admitting that a crime was committed and then rewarding the criminal. Moreover it would set a dangerous precedent for other markets, increasing the incentives for suppliers to lobby for the licensing of entry to their markets.⁸⁶

There is precedent in the other direction. When licensed haulage was de-regulated in the 1978, each existing licence was converted into six licences and full de-regulation of entry followed in 1986. Barrett (1991, pages 86-89) shows how this resulted in expansion of the freight market as its efficiency increased.

We therefore propose that the suppliers of existing taxi services be actively included in the de-regulating process and that every existing licence holder be issued with a new licence. This should be done in a staggered manner, with those holders who bought a licence in the last 5 years being issued with a new licence first, and remaining holders issued with a second licence after 2 . Licences should be freely tradable (no transfer fee) subject to quality standards and full de-regulation should follow in 5 years from now.

This scheme has a number of attractive features. First, the number of licences would increase immediately by approximately 500 and by a further 1,500 in 2 years from now. This would gradually ease the excess demand in the market, although we believe that 2,000 new licences would not be sufficient to satisfy existing demand for taxis. Second, the resale value of the first 500 on the secondary market will be greater than that of the second tranche of 1,500 so that those drivers who bought a licence in recent years will be compensated. Because excess demand for taxis is likely to exist even after the issue of 2,000 new licences, licences on the secondary market still trade at above the administrative fee for the issue of a licence for a further three years. Third, the scheme would be immune to any collusion amongst taxi drivers. For example, an agreement not to use or sell second licenses would be difficult to enforce not be stable because the first licence holder to sell would get the highest price.⁸⁷ Fourth, progressive de-regulation by issuing new licences to persons who do not currently hold licences would be tantamount to offering these persons a windfall monopoly profit.

The only danger with this approach is that the new licences could remove the pressure for fundamental reform of the market so that the supply of taxis can be determined by demand factors. For this reason, it is important that the new regulatory system be put in place at the time that this staggered de-regulation commences and that the date for the full de-regulation of entry be pre-announced. This will enable the market correctly to value licences in the transition period and will ensure full de-regulation to the advantage of this society.

7 Conclusion

On the other hand, we are not convinced by either the arguments or the evidence in favour of de-regulation of price and hence we recommend that fare controls continue. Because the existing system of regulation has not worked, a new system of regulation is needed. This should both regulate and manage the taxi market to the benefit of the customers of taxis and hence to society. Although many cities have de-regulated and re-regulated their taxi markets, few outside of New Zealand have done it successfully. We believe that the re-regulation of the Dublin market along the lines we suggest would not only provide a comprehensive and high quality service to the people who live in and visit this city, but could also serve a blueprint for the re-regulation of taxi markets in other cities.

Arnott, R. (1995), `Taxi Travel Should Be Subsidised', Journal of Urban Economics 40, 316-333.

Barrett, S. D. (1991), Transport Policy in Ireland in the 1990s, Gill and Macmillan, Dublin.

Beesley, M. (1973), `Regulation of Taxis', Economic Journal 83(329), 150- 172.

Beesley, M. & Glaister, S. (1983), `Information for Regulating: The Case of Taxis', The Economic Journal 93(371), 594-615.

Chamberlin, E. (1933), The Theory of Monopolistic Competition, Harvard University Press, Cambridge, Mass.

Coffman (1977), `The Economic Reasons for Price and Entry Regulation of Taxicabs: A Comment', Journal of Transport Economics and Policy 11, 288-297.

Corry, N. (1991), A fare go? The deregulation of the taxi industry, Research Essay 1/91, Geography Department, Victoria University of Wellington.

De Vany, A. S. (1975), `Capacity Utilisation under Alternative Regulatory Restraints: An Analysis of Taxi Markets', Journal of Political Economy 83, 83-94.

Diamond, P. (1971), `A model of price adjustment', Journal of Economic Theory 3, 156-168.

Douglas, G. W. (1972), `Price Regulation and Optimal Service Standards', Journal of Transport Economics and Policy pp. 116-127.

Fingleton, J. (1997), `Standards of Competition in the Irish Economy', Journal of the Statistical and Social Inquiry Society of Ireland. forthcoming. Also available as Trinity Economic Policy Paper No. 97/1 from the Department of Economics, Trinity College, Dublin.

Hackner, J. & Nyberg, S. (1995), `Deregulating Taxi Services: A Word of Caution', Journal of Transport Economics and Policy pp. 195-207.

Hackner, J. & Nyberg, S. (1996), `Vanity and Congestion: A Study of Reciprocal Externalities', Economica 66, 97-111.

Hyland Associates (1976), Study of Taxi and Hackney Services, The Stationary Office, Dublin.

Inter-Departmental Committee (1992), Report of the Inter-Departmental Committee to review the operation of small public service vehicles, The Stationary Office, Dublin.

Klemperer, P. (1995), `Competition when Consumers have Switching Costs: An Overview with Applications to Industrial Organization, Macroeconomics, and International Trade', The Review of Economic Studies 62(4), 515-41.

Leibenstein, H. (1966), `Allocative Efficiency vs. X-inefficiency', American Economic Review.

Massey, P. & O'Hare, P. (1996), Competition Law and Policy in Ireland, Oak Tree Press, Dublin.

Ministry of Transport (1991), `Taxi Industry-Progress Report, Wellington', Ministry of Transport, Wellington.

Morrison, P. (1997), `Restructuring effects of deregulation: the case of the New Zealand taxi industry', Environment and Planning 29, 913-928.

Peltzman, S. (1976), `Towards a More General Theory of Regulation', Journal of Law and Economics 14, 109-48.

Posner, R. (1974), `Theories of Economic Regulation', Bell Journal of Economics 5, 335-58.

Schuurmans-Stekhoven, J. B. (1996), `Who will Hail Taxis in a Deregulated Market? A Comment on Gaunt and Black', Economic Analysis and Policy 26(2), 177-183.

Shreiber, C. (1975), `The Economic Reasons for Price and Entry Regulation of Taxicabs', Journal of Transport Economics and Policy 9, 268-279.

Shreiber, C. (1977), `The economic reasons for price and entry regulation of taxicabs', Journal of Transport Economics and Policy 11, 298-304.

Stigler, G. (1971), `The Theory of Economic Regulation', Bell Journal of Economics 2, 3-21.

Suzumura, K. (1995), Competition, Commitment and Welfare, Clarendon Press, Oxford.

Teal, R., Berglund, M. & Nemer, T. (1984), Urban Transportation Deregulation in Arizona, Report No. UMTA-CA-11-0021-1, U.S. Department of Transportation, Washington, DC.

Teal, R. F. & Berglund, M. (1987), `The Impacts of Taxicab Deregulation in the USA', Journal of Transport Economics and Policy pp. 37-56.

Viscusi, W. K., Vernon, J. M. & Harrington, J. E. (1995), Economics of Regulation and Antitrust, 2nd edn, MIT Press, Cambridge, Massachusetts.

Williams, D. (1980a), `Information and Price Determination in Taxi Markets', Quarterly Review of Economics and Business 20(4), 36-43.

Williams, D. J. (1980b), `The Economic Reasons for Price and Entry Regulation of Taxicabs: A Comment', Journal of Transport Economics and Policy 14(1), 105-112.

S.I. No. 191, 1963
This instrument allowed maximum fares to be set by local authorities and further extended the legislation relating to the licensing of drivers and vehicles [Article 9(1)].

S.I. No. 273 of 1968
The legislation relating to the licensing of drivers and vehicles was extended, particularly with relation to the requirement that drivers have a certain familiarity with local Geography [Article 5(2)].

S.I. No. 200, 1970
A written undertaking would have to be provided to show that if the vehicle were licensed, it would be available on at least five days of the weekend for at least forty eight hours in each week [article 4]. It also required that a driver make his services available for at least forty hours a week [Article 8].

S.I. No. 24, 1976
This amendment allowed for a pick up charge for telephone initiated hires.

S.I. No. 160, 1976
This amendment transferred the power to set maximum fares to the Minister for Industry and Commerce.

S.I. No. 292, 1978
This amendment allowed local authorities to determine entry [Article 5].

S.I. No. 226, 1978
This amendment extends S.I. No. 200 of 1970 by stipulating that, in addition to the forty eight hour rule, that the vehicle be available for hire for at least six hours between 8 AM. and 8 PM. five days a week [Article 2].

S.I. No. 273, 1983
This amendment prohibits hackneys from initiating hires by radio or telephone while in a public place [Article 2].

S.I. No. 272, 1991
Dublin Corporation allowed to issue 100 new taxi licences. The criteria for the issue of such licences are updated [Article 4(2)].

S.I. No. 32, 1992
The power to set maximum fares reverts to the Minister for the Environment.

S.I. No. 32, 1992
Dublin Corporation allow for the issue of 50 wheelchair taxis licences. This instrument also sets out regulations regarding the minimum dimensions of wheelchair taxis and other such regulations. Further amendments are to be found in S.I. No. 358 of 1992, S.I. No. 29 of 1993, S.I. No. 193 of 1997.

S.I. No. 136, 1995
Among other things this amendment allows local authorities to declare and extend taximeter areas [Article 7]. Also, the responsibility of issuing new taxi licences and to set maximum fares was transferred to the four local authorities [Article 32(1)].

      £1.80 For distance not exceeding 12 mile or time 4 minutes. 
      10p For each additional 18th mile or 1 minute.

Additional charges are also permitted as follows:

                    Additional Passengers        40p. 
                    Luggage                      40p per item. 
                    Unsocial hours               40p. 
                    Animals                      40p per animal. 
                    Public Holidays              80p. 
                    Pick Up Charges              £1.20. 
                    Dublin Airport               £1.30.

²We use the term re-regulation instead of de-regulation in order to emphasise that we are not advocating a completely unregulated taxi market.

³Hackneys are private hire vehicles. They are, in principle, unmarked and may not hire at taxi ranks or ply for trade. Most bookings occur by telephone or at hackney offices.

⁴The area of taximeter operation is also limited. In 1995, the Dublin Corporation taximeter area was increased from a radius of ten miles from the city centre to fifteen miles (sic). See Statutory Instrument No. 136 of 1995, Schedule 1. In March of this year, the taximeter area was changed to include the four electoral areas (see footnote 10).

⁵Statutory Instrument No. 136 of 1995 Section 17.

⁶Statutory Instrument No. 273 of 1968, section 6. Also as set out in Statutory Instrument No. 292 of 1978, section (4)(a)(i), the character and previous conduct of an applicant is taken into account.

⁷Road Traffic Act of 1961, especially Part 7. The Minister for the Environment was em powered to license vehicles and drivers, to enforce regulations, and to specify the location of taxi ranks. These tasks are, in turn, delegated to the Garda Commissioner.

⁸This function was transferred to local authorities by Statutory Instrument No. 136 of 1995 (see Schedule 7).

⁹Statutory Instrument No. 292 of 1978.

¹⁰The four local authorities in Dublin are Dublin Corporation, Fingal, Dun Laoghire Rathdown and South Dublin. The 1995 legislation (S.I. No. 136) enabled each to establish its own taximeter area, but they have agreed to cooperate in the formation of a single large taximeter area covering all four districts.

¹¹This provision was first introduced by S.I. No. 200 of 1970 and was reinforced by S.I. No. 226 of 1978 (see Appendix A).

¹²The most recent legislation relating to the licensing of taxis is set out in Statutory Instrument No. 136 of 1995 with additional legislation relating to wheelchair accessible vehicles in Statutory Instrument No. 193 of 1997.

¹³Hackneys are not permitted to ply for trade, use taxi ranks or use bus lanes. The quality of service may differ from that of taxis although the standards are, in principle, the same. See footnote 3 on page 1.

¹⁴An agreement exists between local authorities on the issuing of hackney licences that is similar to that for taxis (see footnote 10) so that Dublin Corporation acts on behalf of all four. A hackney licence issued by one local authority would be valid in all districts, unlike a taxi licence.

¹⁵For example, if Dublin Corporation wished to increase licence numbers and was concerned about quality falling, it would have to negotiate with the Carriage Office for improved quality standards or enforcement. However, there is no mechanism by which it can pay for this extra service, even if it was worthwhile.

¹⁶An increase of 150 was sanctioned, 50 of which were to be wheelchair accessible. However, there is a discrepancy in the figures and 11 taxi licences are unaccounted for.

¹⁷A few companies or individuals own or control groups of licences, but the number of licences affected is probably less than 25 per cent of total supply. Information is provided by Jerome Reilly in the Irish Independent on Saturday 22 November 1997.

¹⁸See footnote 3 on page 1. Hackney offices have sprung up around the city and operate in a similar manner to taxi ranks, except that they provide indoor waiting space.

¹⁹Dublin Corporation took over the regulation of the taxi market on September 1st, 1995. At that time it was told that there were approximately 1,400 hackney licences. It now believes the figure to be 3,000, although the Hackney Association estimates the current figure at 3,500. The Carriage Office, which issued licences prior to September 1995 is unable to provide data on hackney numbers.

²⁰Commissioned by the Dublin City Centre Business Association.

²¹An RTE Prime Time programme broadcast on 27th of November 1997 found people waiting at taxi ranks in the city centre for periods of between 90 and 180 minutes.

²²This is despite a transfer fee of £3,000 in the exchange of a licence from one owner to another.

²³See footnote 3 on page 1 above. They are better substitutes for those booking by telephone and who know the city.

²⁴In correspondence with the authors, the Dublin Branch of the Irish Hotels Federation noted that guests are sometimes refused taxis for short journeys. They also made the point that "the quality of driver, in respect of their job knowledge and qualities as `ambassadors for our city' is...found wanting".

²⁵As Table 2 indicates, this only affected 139 taxis. A justification given for that low fee was the requirement that new licensees be wheelchair accessible.

²⁶To some extent, this reflected the arguments of full-time taxi drivers that competition from part-time operators, particularly at peak-times, was detrimental to the quality of service.

²⁷Ironically, in some of these markets, the number of taxis trebled after de-regulation and there is no reporting of active measures to maintain quality standards.

²⁸Data on the number of hackney licences and the level of hackney fares relative to taxi fares are poor.

²⁹As we see in Section 4 below, de-regulation in taxi markets has resulted in increases in doubling the supply in some instances where the value on taxi plates on the secondary market was considerably lower.

³⁰Hackneys, because they cannot hire in the street, have a higher proportion of one-way journeys. On the other hand, taxi productivity is likely to fall following de-regulation.

³¹In addition, if entry is restricted, prices may be high due to scarce supply. Regulation of prices would then lead to queuing.

³²Douglas (1972) and Shreiber (1975) discuss price competition in the taxi market.

³³Klemperer (1995) surveys the literature on competition in markets with customer switching costs. Corry (1991, reported in Morrison (1997)) interviewed taxi-company managers and found a reluctance to lower nominal fares on the basis that the public "showed little initial awareness of intercompany price differences" and that they "would not respond in sufficient numbers to price reductions to compensate any given company for those fare reductions".

³⁴See Shreiber (1977).

³⁵This arises in many markets but the goods are often once-off purchases so that it may pay the customer to make detailed price comparisons. An example might be mobile phones where each supplier offers a range of price formulae.

³⁶The fare depends on time and distance, and other variables. Consider the following two relatively simple fare schedules. Schedule A: fixed fee of £1.20 plus 60p per kilometre and 15p per minute waiting. Schedule B: fixed fee of £1.60 plus 70p per kilometre and 14p per minute waiting. Even a highly informed customer might not know relevant information such as journey time.

³⁷It might be easier for taxi drivers to maintain collusion at a taxi rank as any supplier who undercut could immediately be identified by competitors. This might not apply to undercutting by telephone.

³⁸Teal & Berglund (1987) cite evidence that forty percent of demand for taxi services (in San Diego and Seattle) is by people who use taxis once a month or less.

³⁹For example Schuurmans-Stekhoven (1996) reported significant overcharging (and over-crowding) at taxi ranks at airports in New Zealand.

⁴⁰See Coffman (1977) and Williams (1980a, 1980b).

⁴¹Teal & Berglund (1987) report that between 70 and 80 per cent of hiring in US cities is by telephone.

⁴²Ironically, this argument works best if the taxi fleets are larger, that is, if the market is more concentrated.

⁴³The result was developed by Chamberlin (1933). For a recent account, see Armstrong, Cowan & Vickers (1994) or Suzumura (1995).

⁴⁴One type of congestion in the taxi market arises from a negative externality among customers. A customer who hires a taxi imposes higher waiting costs (and hence higher prices) on other users. Hackner & Nyberg (1996) develop a theoretical model of congestion generally and Hackner & Nyberg (1995) apply this to the taxi market. They find that with congestion, taxis can charge prices above cost.

⁴⁵De Vany (1975) models the price of taxis as being the fare plus the cost of waiting time along the following lines

where h is the monetary value of the time t spent waiting. Excess capacity would reduce the waiting time h (t) so that consumers would be better off even if the fare had to rise by a little.

⁴⁶Fixed costs are borne regardless of the volume of work undertaken by a taxi. They would include insurance (if it did not vary with workload) and any fixed or annual licence fee. Fixed costs would not include the costs of the car because the cost of depreciation varies with output, a second-hand market exists, and because cars can have alternative uses (e.g., domestic) even when "marked" as taxis. Other variable costs include labour, fuel, and other running costs.

⁴⁷Hackner & Nyberg (1995, page 204) illustrate that fixed costs are low and that sunk costs are even lower. Sunk costs refer to fixed costs that are committed to the market in the sense that they cannot be recovered on exit. For example, if licences are tradeable, then any licence fee is a fixed cost but not a sunk cost.

⁴⁸We deal with the effects of high licence fees below.

⁴⁹The corresponding problem of taxis leaving ranks under-supplied in order to get a fare increase would be unlikely to occur unless entry to the market was restricted. Thus good price regulation would require free entry.

⁵⁰See Armstrong et al. (1994, pages 106-111) and Suzumura (1995, pages 197-200) for an elaboration of these arguments.

⁵¹Making licenses fully tradeable in markets where entry is restricted encourages productive efficiency. An inefficient licensee bears a high opportunity cost associated with holding a licence as it is worth more to a more efficient potential supplier. Enabling trade would encourage the licenses in existence to be allocated to the most efficient potential suppliers.

⁵²X-inefficiency, a term coined by Leibenstein (1966), occurs when firms fail to minimise costs for a given level of output. In contrast, productive inefficiency occurs when the scale of output is inefficient.

⁵³See Stigler (1971), Posner (1974), and Peltzman (1976).

⁵⁴For example, the decisions of the Competition Authority contain detailed reasoning of this kind.

⁵⁵This model of independent regulation has long been used in the US and has more recently become common in other countries, most notably but not exclusively in the UK and New Zealand. In Ireland, the Competition Authority and the recently created Office of Telecommunications Regulation have been established as independent institutions for similar reasons.

⁵⁶In the Dublin market, for example, there are several thousand drivers, but several million users of services, especially if one includes business visitors and tourists.

⁵⁷Transparency is required to compensate for the reduction in direct political account ability. It would involve the clear specification of regulatory objectives, open procedures and written reasoning to accompany decisions so that one can observe how decision relates to the objectives.

⁵⁸Unless otherwise stated, the sources of information in this section are Morrison (1997) for New Zealand, Hackner & Nyberg (1995) for Sweden, and Teal & Berglund (1987) for the United States.

⁵⁹Conversely, higher prices would be expected if the regulated fare had failed to keep up with cost increases.

⁶⁰The distinction between the nominal and real (inflation adjusted) fares receives much attention from authors but is not a suitable benchmark for the initial adjustment. Even for subsequent movement in prices, it may be biased if taxi costs change at a different rate to average prices.

⁶¹Using fares in cities that continue to be regulated as a benchmark could also be biased, especially if the regulated fares are set at too high or too low a level.

⁶²Reported by Teal & Berglund (1987) for the U.S. and Morrison (1997) for New Zealand.

⁶³A 10km journey from the airport had a range across the companies of 30 per cent. An evening 8km journey booked by telephone and a three minute waiting time had a range of 15 per cent across the six companies. A 5km daytime ride had a range of 34 per cent across the companies.

⁶⁴Teal & Berglund (1987) found that companies generally did not see market share rise if they cut prices. This conforms with Corry's interview evidence for New Zealand noted in footnote 33 above.

⁶⁵This might be tempered by higher costs due to reduced occupancy rates.

⁶⁶Viscusi, Vernon & Harrington (1995, page 345) reports maximum values of $15,000 (San Diego) and $12,000 (Seattle). The only city where the value exceeds that in Dublin is New York ($210,000 in 1993).

⁶⁷Reported in Massey & O'Hare (1996).

⁶⁸Beesley & Glaister (1983) report that 15 per cent of all London taxi drivers have other full-time occupations.

⁶⁹An alternative policy, road pricing, although long recommended by economists as more efficient, is rarely employed. Subsidies may also have a positive effect on distribution, with disproportionate benefits for those on lower incomes. See Arnott (1995).

⁷⁰This argument contrasts with that of Shreiber (1975, page 275) who assumes that taxis and public transport are always substitutes. This is certainly the case for some (probably low) level of taxi fares.

⁷¹The complementarity could be in the same direction. For example, providing public transport late at night might increase city-centre activity in the small hours and create a critical mass effect that would increase demand for taxis also. Casual observation suggests that this occurred with the introduction of night buses in Dublin.

⁷²See Armstrong et al. (1994) for a comprehensive discussion of the advantages and disadvantages of public subsidies to cover fixed costs in this way. They also examine methods such as tendering that can mitigate some of the disadvantages.

⁷³For reasons outlined above, the auctioning of licences would be an extremely bad idea. Although it would maximise government income from the market, it would impose high fixed costs on suppliers in the market resulting in suboptimal use of taxis and inflexible supply.

⁷⁴Higher prices at peak times mean that consumers who are willing to pay suppliers for the extra unused capacity they generate can obtain the service without waiting.

⁷⁵As noted above, this is the proportion of time that a taxi plying for trade is actually in use. It contrasts with flexible supply which is the quantity of time that taxis ply for trade.

⁷⁶This is UK government policy and all taxis are to be wheelchair accessible by 2005. See the 1995 Public Transport Act, UK.

⁷⁷The cost might be slightly lower if all taxis are required to be accessible because of the benefits of mass production. This would particularly be true if the new UK standard for accessibility (see footnote 76) were adopted.

⁷⁸One idea is that disable persons might have a taxi-pass (unlike a bus-pass which is less useful). The subsidy could be lump-sum (for example, equivalent to the subsidy given to car owners) or based on usage. Swipe cards might be used to avoid deferred refunding and the problem of apparent discrimination.

⁷⁹It is possible that one taxi firm or dispatcher would specialise in accessible taxis if there was excess capacity in the taxi market.

⁸⁰Even though the costs of hackneys and taxis differ, the existence of price competition in the private hire market might enhance regulation of the taxi market by providing a benchmark or yard-stick for the regulator. See Armstrong et al. (1994, pages 74-77) for a discussion of yard-stick regulation.

⁸¹It should be well-advertised in taxis and hotels with a memorable telephone number.

⁸²The taximeter area should cover the entire city, however defined. In particular, there should be no division of the city into separate taximeter districts.

⁸³In this context, we would argue that the current proposal to charge £15,000 per new licence is ill-advised for public policy reasons and that it inhibits de-regulation of the market because there would be some moral force to the argument that Dublin Corporation should refund that money paid to it in the event of de-regulation.

⁸⁴Adjusting the mileage rate would also be possible.

⁸⁵The taxi market does not come directly under this legislation because the source of the monopoly is in the licensing system and the legislation does not protect consumers from monopolies created by government intervention. See Fingleton (1997) for a discussion.

⁸⁶This is not a hypothetical consideration. In 1996, pharmacists succeeded in persuading the government to restrict entry to that market. A similar problem would exist with pub licences.

⁸⁷Such an agreement would almost certainly offend against the 1991 Competition Act with the possibility of criminal sanctions.

Adverts