n/e/r/a

THE COSTS TO BUSINESS FROM POSTAL DELAYS A Report to Postwatch Prepared by NERA and RAND Europe* September 2002 London

Authors NERA John Dodgson Jan Peter van der Veer RAND Europe Andrew Daly Charlene Rohr Peter Burge * Rand Europe contributed to Chapter 5 only

n/e/r/a

National Economic Research Associates Economic Consultants 15 Stratford Place London W1C 1BE Tel: (+44) 20 7659 8500 Fax: (+44) 20 7659 8501 Web: http://www.nera.com An MMC Company

n/e/r/a TABLE OF CONTENTS EXECUTIVE SUMMARY

i

1.

INTRODUCTION

1.1. 1.2.

Scope of the Report Structure of the Report

1 1 2

2.

POSTAL SERVICES FOR BUSINESS CUSTOMERS

2.1. 2.2. 2.3.

Introduction Key Royal Mail Products for Business Royal Mail Traffic

3.

POSTAL STANDARDS AND PERFORMANCE

3.1. 3.2.

Royal Mail Standards Royal Mail Performance

4.

INTERVIEWS WITH MAJOR POSTAL CUSTOMERS

4.1. 4.2.

The Interviews Interview Results

5.

THE STATED PREFERENCE SURVEY: METHOD AND RESULTS

5.1. 5.2. 5.3. 5.4. 5.5. 5.6. 5.7. 5.8.

Introduction The Stated Preference Survey Design Of the SP Questionnaire The Interviews The Use of Royal Mail Services by the Companies Recruited Customer Priorities and Satisfaction Direct Estimates of the Impact and Cost of Delays Valuations of Postal Reliability

6.

THE OVERALL COSTS OF POSTAL DELAYS

6.1. 6.2.

Methodology Results

3 3 3 5 7 7 7 9 9 9

15 15 16 18 20 21 22 24 25 30 30 31

Ref: Anne Mace/8/17-Sep-02/F:\Anne.Mace\msoffice\winword\NERA Personnel\John Dodgson\Postwatch Final Report_jd17092.doc

n/e/r/a LIST OF TABLES Table 2.1 Table 3.1 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9 Table 5.10 Table 5.11 Table 5.12 Table 6.1

Proportion of Mail Sent/Received by Business/Residential Customers (% of Total Mail Volume) Royal Mail Performance, 2001/02 Variables and Levels used in the SP Experiments Average Number of Letters Sent Per Month by Any Postal Service Nature of Business by Segment Users’ Identification of the Most Important Aspect of the Service Provided to Them by Royal Mail Users’ Level of Satisfaction with the Service Provided by Royal Mail Priorities in Choosing Postal Service Users’ Assessment of Whether Delays in Postal Services Affect Their Business Estimates of Annual Cost of Delivery Delays to Business Users’ Assessment of Whether Delays in Incoming Mail Affect Their Business Estimates of Annual Cost of Delivery Delays of Incoming Post to Business Valuation for Changes in Advertised Day of Mail Delivery (pence) Marginal Valuations for Additional Changes in Reliability, i.e. Percentage of Post Delivered by the Advertised Day (pence) Costs of Postal Delays in 2001/02

6 8 19 21 22 22 23 23 24 24 25 25 28 28 31

LIST OF FIGURES Figure 2.1 Figure 2.2 Figure 5.1

Royal Mail Traffic from Different Services, 1999/00 Royal Mail Revenue from Different Services, 1999/00 Example SP Choice Pair

5 6 20

n/e/r/a

ACKNOWLEDGEMENTS

NERA would like to gratefully acknowledge the assistance of the following organisations in writing this report. However, all responsibility for the report rests with ourselves.

Consignia Mail Users Association Postcomm Postwatch The many business customers who participated in the surveys

n/e/r/a

Executive Summary

EXECUTIVE SUMMARY Scope of this Report 1.

This report has been commissioned by Postwatch to provide estimates of the total costs of delays and other service quality problems to Royal Mail’s business post users in the United Kingdom. Based on the stated preference approach to estimating the values of service quality attributes, NERA estimates that the cost of delays compared with existing performance standards was in excess of £ 150 million in 2001/02.1 NERA estimates that the total cost of delays was in excess of £ 450 million in 2001/02.

Postal Services for Business Customers 2.

About 86 per cent of current mail items are sent by businesses, and only 14 per cent by residential customers. Of the mail sent by businesses, about one third is sent to other businesses, and the remaining two thirds to residential customers. As well as the first and second class standard tariff letter services, Royal Mail provides a range of services specifically for businesses. The most important of these are the Mailsort 1, 2 and 3 services, which provide discounts from standard tariffs for pre-sorted prepared mail meeting minimum traffic volumes. Presstream services provide equivalent first and second class services for newsletters, periodicals or journals. In 1999/00 the various business services accounted for 48 per cent of total Royal Mail volumes, and for 44 per cent of revenues.

Postal Standards and Performance 3.

Royal Mail service standards are set out in Licence Condition 4 of Consignia’s licence from Postcomm. Performance against the targets is published by Consignia on a quarterly basis. In the 2001/02 Financial Year Consignia’s target for first class stamped and franked mail was to deliver 92.1 per cent by the next working day, but performance was only 89.9 per cent. Consignia’s target for first class postage paid impression mail was to deliver 90.0 per cent by the next working day, but performance was only 81.4 per cent. Delivery performance in other categories of mail also fell below the targets.

The Views of Major Customers 4.

1

NERA sought the views of major postal users through a series of interviews. These users’ major concern was with reliability. The firms interviewed were concerned both that delivery standards were not met, and that they are not high enough. In

For the purpose of this report, Royal Mail’s performance is the cumulative performance for each service for the entire year. This is distinct from the figures on which service standards are judged for the purposes of Consignia’s licence, which are only based upon performance in February and March.

i

n/e/r/a

Executive Summary

regard to first class mail, users were aware that “ next day” delivery did not necessarily mean this, and that they were not getting what they paid for when they used first class mail services. For large mail users, there was high reliance on Mailsort 2 and 3 services, and a variety of concerns about timing and reliability of mail deliveries. Firms had made attempts to measure the service quality they were receiving, and had often found that it fell short of national published performance statistics – though there was sometimes too a frustration about the value of conducting these exercises. Other concerns were with: delays to response services; timing of delivery and collection of mail from the firms’ premises; percentage of mail arriving with first delivery; the impact of actual or threatened industrial disputes at Royal Mail; and the service provided by Royal Mail’s account management. 5.

Although we asked the firms interviewed if they could provide estimates of the costs of postal delays, there were only two cases where the firms were able to provide specific financial estimates of the costs of particular impacts. In other circumstances, the firms had not made estimates of the costs: reasons included the difficulty of separating out the impact of different factors that might have caused a mail advertising campaign to have been relatively unsuccessful, or the fact that the firm would not be able to claim back costs incurred from Royal Mail so there was no value in making the calculation. The lack of explicit estimates of postal delay costs was confirmed in our detailed stated preference survey (see below). Of the 302 firms surveyed in that survey, only ten per cent of those experiencing problems were able to put a financial range on the annual costs they incurred as a result of delays to outward mail, and only 15 per cent of those experiencing problems were able to put a financial range on costs as a result of delays to inward mail

Stated Preference Estimates of Service Quality Problems 6.

As part of this study we conducted a stated preference survey of a selection of all business mail users. Stated preference is a widely-used technique to value attributes of service quality in a wide range of sectors, and has been used by Consignia to value service quality in postal services in the UK. The stated preference questionnaires were designed by RAND Europe in conjunction with NERA, and a pilot and the main telephone surveys were conducted by the market research company Market Probe. The sample of customers covered the main types of service used by business customers, namely first class mail, second class mail, Mailsort 1, Mailsort 2, Mailsort 3, and second class response services. Interviews were conducted with 50 (or more) firms in each of these six sectors.

7.

The most important aspect of the service provided by Royal Mail was reliability, identified as most important by 43 per cent of respondents. Speed of delivery, identified by 24 per cent of respondents, was next, followed by price (16 per cent), security (9 per cent), and time of delivery (8 per cent). In terms of priorities in choosing postal services, 52 per cent said that their choices were based on quality of ii

n/e/r/a

Executive Summary

service and price, 25 per cent on the proportion of mail delivered by a set date, 13 per cent on the proportion of mail delivered on a set date, and ten per cent on price. When asked to quantify their satisfaction with Royal Mail services, 28.8 per cent of respondents said that they were very satisfied with Royal Mail services, 54.3 per cent said they were quite satisfied, 5.6 per cent said they were quite dissatisfied, 1.7 per cent said they were very dissatisfied, and the remaining 9.6 per cent were neither satisfied nor dissatisfied. 8.

Fifty five per cent of respondents reported that delays in outgoing mail directly affected their business. Sixty three per cent of respondents indicated that delays to incoming mail affected their business.

9.

Results from the stated preference interviews were used to estimate discrete choice models of the factors affecting choice of each of the six types of mail service. This yielded estimates of the values placed on percentage changes in reliability, and of the values of changes in the advertised day of delivery. The relationships between valuations for different types of mail conform to theoretical expectations. Marginal valuations of changes in reliability are shown in Table 1. Table 1 Marginal Valuations per Item for Additional Changes in Reliability, i.e. Percentage of Post Delivered by the Advertised Day (pence)

Per % point

Response services

First class*

Mailsort 1

Second class+

(p)

(p)

(p)

(p)

(p)

(p)

0.82

0.44

0.32

0.27

0.17

0.85

Mailsort 2 Mailsort 3

*

For the purpose of this report ‘First class’ includes First class Printed Postage Impression (PPI).

+

For the purpose of this report ‘Second class’ includes Second class Printed Postage Impression (PPI).

The Overall Costs of Postal Delays 10.

The report estimates two broad types of total cost, based on two measures of the performance standard, namely:

·

cost of delays compared with current target quality of service performance; and

·

cost of delays compared with an ideal (ie 100 per cent) quality of service performance.

iii

n/e/r/a 11.

Executive Summary

Total costs for each category of mail were derived as follows:

Cost Cost of of delay delay in mail in mail category category in in 2001/02 2001/02

Valuation Valuation per per % % point improvement point improvement in in reliability reliability

12.

X

=

Gap Gap between between actual actual performance performance and and performance performance standard standard

X

Volume Volume of of mail mail in in the the mail mail category category sent sent by by business business users users

Table 2 shows NERA’s estimates of total delays in Financial Year 2001/02. Table 2 Costs of Postal Delays in 2001/02 Costs of Not Meeting Existing Performance Standards

Costs of All Delays

(£ m)

(£ m)

First class

92

288

Second class

15

41

Mail Type

Mailsort 1

4

18

Mailsort 2

20

43

Mailsort 3

4

16

Second class response

4

10

Other

18

41

Total

156

456

13.

The total cost of delays compared with existing standards is estimated at £ 156 million a year.

14.

The total cost of delays compared with ideal standards is estimated at £ 456 million a year.

iv

n/e/r/a 1.

INTRODUCTION

1.1.

Scope of the Report

Introduction

This report was commissioned from NERA by Postwatch, the UK consumer postal watchdog. The report provides estimates of the total annual costs imposed on UK businesses and business customers of delays arising from postal services provided by Royal Mail. Royal Mail is one of the brands of Consignia, the national publicly-owned postal operator in the United Kingdom. Royal Mail collects and delivers letters and packages, provides direct mail services, and designs and produces the UK’s stamps and philatelic products. We used two main techniques to undertake this study. First of all NERA carried out detailed interviews with 12 major Royal Mail customers, concentrating on firms in the mail order sector, those in the financial sector, and utilities providing public services. The aim of these interviews was to provide an initial assessment of the types of mail used by major business users and of their concerns about mail service quality. The lessons from the earlier of these interviews were then used to help design questionnaires for the second stage of the study. The second stage of the study involved the use of the Stated Preference (SP) approach to provide valuations of different attributes of service quality by the full range of Royal Mail’s business customers. Stated preference is now a widely-used technique to value attributes of service quality in a wide range of sectors. It has already been used by Consignia to value service quality in postal services in the UK. The sample of customers for our SP survey covered the main types of service used by business customers, namely standard first class mail, standard second class mail, Mailsort 1, Mailsort 2, Mailsort 3 and second class response services. Stated preference questionnaires were designed by RAND Europe, in conjunction with NERA, and the main and pilot telephone surveys were conducted by the market research company Market Probe. Market Probe conducted a total of 302 interviews in the main survey, 50 in each of the six service sectors.2 Results of the SP survey were then analysed to provide valuations of service quality attributes. These valuations were then aggregated over the whole of the business sector to provide our estimates of the total cost of delays in the Financial Year 2001/02.

2

There were 52 in the first class mail segment.

1

n/e/r/a 1.2.

Introduction

Structure of the Report

The structure of the rest of this report is as follows:

·

In Chapter 2 we provide an overview of the market for postal services for business customers in the UK, detailing the main products, and information on Royal Mail’s traffic and revenue from different products and customers.

·

Chapter 3 provides information on current service standards, as set out in Consignia’s licence from the postal regulator Postcomm, together with statistics on current performance.

·

Chapter 4 presents the results of our interviews with business customers.

·

Chapter 5 explains how we undertook the SP survey of business users, and summarises the results.

·

Chapter 6 explains our methodology for estimating the total annual costs of postal delays using the Stated Preference approach, and provides our estimates of these costs.

2

n/e/r/a

Postal Services for Business Customers

2.

POSTAL SERVICES FOR BUSINESS CUSTOMERS

2.1.

Introduction

This Chapter is divided into two main sections. Section 2.2 provides an introduction to the main services provided by Royal Mail. Section 2.3 summarises the publicly-available information on Royal Mail’s traffic and revenue from the services and from business customers.

2.2.

Key Royal Mail Products for Business

2.2.1. The range of products There are two main classes of mail available in the UK for all customers, namely first class mail, whose delivery standard is next (working) day delivery, and second class mail, whose delivery standard is delivery by the third working day after posting. For purposes of setting targets, first class mail is split into first class stamped and franked, and first class Postage Paid Impression. Similarly, second class mail is split into second class stamped and franked, and second class Postage Paid Impression. In addition there is a variety of services which are available to volume mailers who are able to guarantee minimum volumes and meet certain standards for the preparation of mail. These include Mailsort 1, Mailsort 2, Mailsort 3, Walksort, Door to Door (unadressed items), Presstream 1, and Presstream 2. Royal Mail also provides response services for customers, primarily Freepost and Business Reply. A number of key services are described in more detail in the subsections below. 2.2.2. Mailsort Mailsort is a service that offers discounts for large mailings that are sorted and prepared by the sender, as well as 90 per cent accurately postcoded. Mailsort can be used for sending a specified number of items (normally 4,000 letters), all of the same size and weight. The following Mailsort services are available:

·

Mailsort 1: for mailings targeted for delivery the next working day after posting;

·

Mailsort 2: for mailings targeted for delivery within three working days;

·

Mailsort 3: for mailings targeted for delivery within seven working days; and

·

Mailsort 3 deferred: enables a mailing to be prepared and lodged with Royal Mail until an agreed delivery start date. The mailing is targeted for delivery within four working days of this delivery start date.

3

n/e/r/a

Postal Services for Business Customers

Mailsort discounts over standard rates depend on the level of sorting undertaken by the mailer. Typical Mailsort 1 and 2 discounts would be around 15 per cent off standard first and second class rates. Mailsort 3 discounts are higher, between 25 and 30 per cent off standard second class rates. A separate product within the Mailsort range of services is Walksort. Walksort offers additional discounts over the Mailsort rates if mail is pre-sorted by the mailer to within its actual “ postal walk”, including the route and the addresses within it. To qualify for Walksort, a mailing must be sent to at least one in ten addresses within its coverage area. Walksort is available with a one (Walksort 1) and a three (Walksort 2) working day delivery time and offers discounts of between 36 and 42 per cent off standard Mailsort 1 and 2 rates. 2.2.3. Presstream Presstream is a service similar in nature to Mailsort, but specifically designed for organisations sending newsletters, periodicals or journals. There are two delivery options:

·

Presstream 1: for next working day delivery after the day of posting; items must be sent at least four times per year; and

·

Pressstream 2: for delivery within three working days after the day of posting; items must be sent at least two times per year.

Because Presstream requires items to be mailed regularly, the discounts that are offered on the Presstream service are considerably higher than those on Mailsort (for a given weight category – often Presstream products will be heavier and therefore fall into a higher weight category). 2.2.4. Special delivery For time-critical mail, Royal Mail offers a Special Delivery service which guarantees a 12 noon delivery on the next working day for most parts of the UK. If an item arrives later than the deadline, postage is refunded. If items are lost or damaged, compensation can be claimed. In addition, it is possible to take out coverage for Consequential Loss at an additional charge. This provides compensation for financial loss that is separate from an item’s actual market value, incurred as a result of delay, loss of or damage to items. 2.2.5. Response services Response services are another important category of Royal Mail services for business. Response services are used when companies wish to enable (potential) customers to send mail to them without having to pay for a stamp.

4

n/e/r/a

Postal Services for Business Customers

The following key response services are available:

·

Business Reply. With this service, customers can be provided with envelopes or cards that are preprinted with the company’s address and with first or second class Postage paid.

·

Freepost. With this service, either prepaid envelopes of cards can be supplied to consumers, or consumers use their own envelopes or cards.

2.3.

Royal Mail Traffic

Figure 2.1 shows Royal Mail traffic in 1999/00 from different services, the last year for which we have been able to obtain publicly available detailed data.3 It can be seen that the three Mailsort services account for around 35 per cent of Royal Mail’s traffic, although Mailsort 1 traffic is a small proportion of total Mailsort. Figure 2.1 Royal Mail Traffic from Different Services, 1999/00

First class

3

Second class

Mailsort 1

Mailsort 2

Mailsort 3

Second class response

Other

Consignia have supplied us with 2001/02 data which we have used for our calculations in Section 6, but these more recent data are not in the public domain. The 1999/00 data were published by Postcomm in the report: An Assessment of the Costs and Benefits of Consignia’s Current Universal Service Provision: a Discussion Document June 2001, pp.33/34.

5

n/e/r/a

Postal Services for Business Customers

Royal Mail revenue in 1999/00 from different services is shown in Figure 2.2. The Mailsort services account for just under 30 per cent of Royal Mail revenue, with first class being the single most important product in revenue terms. Figure 2.2 Royal Mail Revenue from Different Services, 1999/00

First class

Second class

Mailsort 1

Mailsort 2

Mailsort 3

Second class response

Other

The proportion of mail sent and received by business and residential customers is shown in Table 2.1. The table shows that about 86 per cent of the total mail volume is sent by businesses, though businesses receive only about a third of the total mail volume. Of the mail sent by business, about two third is sent to residential customers and one third to other businesses. Residential users also send the majority of their mail to other residential users. Table 2.1 Proportion of Mail Sent/Received by Business/Residential Customers (% of Total Mail Volume) Received by

Sent by

Business

Residential

Total

Business

28

58

86

Residential

5

9

14

Total

33

67

100

Source: Postwatch Competition in Postal Services

6

n/e/r/a

Postal Standards and Performance

3.

POSTAL STANDARDS AND PERFORMANCE

3.1.

Royal Mail Standards

Royal Mail service standards are set out in Consignia’s licence from Postcomm, under Licence Condition 4. The Licence requires Consignia to apply service targets in its internal business planning and lays down minimum levels for these (see Table 3.1 on page 8). Consignia shall use all reasonable endeavours to meet these standards and is required to comply with a code of practice for identifying and addressing the causes for significant failures to meet the standards. Although the Licence does not explicitly specify penalties for not meeting the service, Consignia is required to develop a compensation scheme for customers affected by poor service quality. Every quarter, Consignia has to submit a progress report on its performance to Postcomm. For the purpose of this report, Royal Mail’s performance is the cumulative performance for each service for the entire year. This is distinct from the figures on which service standards are judged for the purposes of Consignia’s licence, which are only based upon performance in February and March. In addition, Condition 5 of Consignia’s licence lays down minimum standards for Consignia’s complaint handling procedures and requires the company to submit quarterly reports on the application of these procedures.

3.2.

Royal Mail Performance

Table 3.1 summarises service quality performance in the Financial Year 2001/02. As can be seen, Royal Mail did not meet any of its regulatory targets in this year and sometimes missed them by a significant margin. For example, while the headline first class target was missed by 2.2 per cent-point, the target for first class postage paid impression was missed by over 8.5 per cent-point. This affects businesses most as they are the main senders of postage paid impression items. Compared to the standard postage paid impression services, Royal Mail’s performance on Mailsort is somewhat better, though still well below the standards set. A similar picture applies to Presstream. On response services, by contrast, the targets were missed by very significant margins.

7

n/e/r/a

Postal Standards and Performance

Table 3.1 Royal Mail Performance, 2001/02

Product

1st class stamped and franked

Delivery standard (working days) 1

Delivery by target day Target Performance (%) (%) 92.1

89.9

Tail of mail (3 days after due date) Target Performance (%) (%) 99.9

99.8

Absence of loss/substantial delay Target Performance (%) (%) 99.0

100.0

2nd class stamped and franked

3

98.5

98.3

99.9

99.9

99.0

100.0

1st class postage paid impression

1

90.0

81.4

99.9

99.3

99.0

100.0

2nd class postage paid impression

3

97.0

94.4

99.9

99.5

99.0

100.0

Mailsort 1

1

92.1

90.0

99.9

99.7

99.0

100.0

Mailsort 2

3

97.6

95.5

99.9

99.7

99.0

100.0

Mailsort 3

7

98.5

97.9

99.9

99.8

99.0

100.0

Presstream 1

1

91.0

89.2

99.9

99.5

99.0

100.0

Presstream 2

3

97.6

96.2

99.9

99.9

99.0

100.0

1st class response services

1

90.5

78.1

99.9

99.7

99.0

100.0

2nd class response services

3

97.0

93.7

99.9

99.5

99.0

100.0

Source: Consignia Licence Condition 4: Paragraph 8a and Licence Condition 5: Paragraph 6 – National Report to Postcomm and Postwatch Quarter January 2002 – March 2002, p.45. Note: For the purpose of this report, Royal Mail’s performance is the cumulative performance for each service for the entire year. This is distinct from the figures on which service standards are judged for the purposes of Consignia’s licence, which are only based upon performance in February and March.

8

n/e/r/a

Interviews with Major Postal Customers

4.

INTERVIEWS WITH MAJOR POSTAL CUSTOMERS

4.1.

The Interviews

NERA has conducted 12 interviews with major postal users. These consisted of:

·

A major bank;

·

A major building society;

·

A major insurance company;

·

A major utility;

·

Three general mail order companies;

·

Two more specialised mail order companies;

·

A travel company;

·

A publisher; and

·

A mailing house servicing clients in both the financial and utilities sectors.

The semi-structured interviews covered three main areas:

·

The nature of the firm’s use of Royal Mail services;

·

The features of service quality of particular concern to the firm; and

·

The types of cost imposed by delays and other service quality problems

A blank interview form which shows the structure of the interviews is reproduced in Appendix A.

4.2.

Interview Results

4.2.1. Firms’ use of Royal Mail services It was clear from the interviews that letter mail should be seen as part of an overall process in which firms were engaged, be it sales of goods via mail order, or the retailing of services such as basic utilities, or provision of banking or other financial services. Concerns about service quality therefore relate to where mail fits into that process.

9

n/e/r/a

Interviews with Major Postal Customers

Some mail clearly is time sensitive, such as a customer’s request to a bank to stop a cheque or a request for a new credit card to replace one that has been lost, or a quote for new business that may be competing with requests from other firms, but much mail, such as invoices or marketing mailings can be planned in advance. Mailing houses are often used for bulk mailings. In particular mailing houses may be used for routine bulk items such as invoices or quotes for new business, whereas an organisation will send responses to individual non-standard customer enquiries themselves. Mail costs are seen as an important element of costs, and decisions appear to be taken centrally more than used to be the case. Firms are often unwilling to pay for “ next day” delivery – and are aware that “ next day” delivery does not necessarily mean this because much less than 100 per cent of letters sent by first class mail services will be delivered the next working day (this was mentioned especially with Mailsort 1 and Presstream 1). Of course, there are exceptions where firms are prepared to pay, eg postal financial accounts, responses to consumer requests for quotes or brochures. One firm had switched from mainly first class mail to second because second class was adequate for the service required and reliability was good – almost everything is delivered within three days and the cost saving was significant (although since the switch there had been some drift back to first class). 4.2.2. The features of service quality of particular concern to the firms The firms interviewed were concerned both that published delivery standards are not met, and that they are not high enough. Firms appreciate that first class mail does not necessarily arrive the next day: a number pointed out that they were paying for next day delivery if they used first class mail, but were not getting it, and that the target of 92 per cent was not good enough, especially as this relates to “ good” items only, that is those that are well addressed and meet other standards. A major financial institution also noted that the 92 per cent target for first class mail was too low. Another noted that the main impact for them of delays in first class deliveries was an increase in customer complaints about the service being provided. Mailsort 2 or 3 (including Mailsort 3 deferred) is of primary concern for many large mailers. Particularly important were issues of timing and reliability: For example:

·

Book club catalogues are sent out at regular intervals, and there is a limited time slot for members to respond to offers.

·

There is a limited period during which mail order catalogue offers are available.

·

Mail order firms may use an initial “ teaser” letter and a main mailing offer – there is an optimal gap between these two letters to maintain the potential customers’

10

n/e/r/a

Interviews with Major Postal Customers

interest, while customers will become confused if the teaser and main letters arrive on the same day, and especially if the teaser arrives after the main letter.

·

Preview mail order catalogues are time-sensitive since they are used to develop forecasts for numbers of items that will be ordered from the main catalogue – the mail order company concerned sends these preview catalogues by Mailsort 1.

·

A mail order company launching a major campaign will staff up call centres to deal with the calls, so if there are problems with the mail, staff may end up idle.

·

A mail order company has service level agreements with its customers that specify when remittances have to be delivered to final customers, so customers may complain if they do not receive their statements when they expect them to arrive.

·

Some mail order companies use co-ordinated media (e.g. TV and mail) campaigns; in which mailing timings need to be co-ordinated with the timings of TV adverts.

·

The utility company interviewed commented that reminder bills and disconnection notices are particularly time-sensitive and so are sent by Mailsort 1.

·

It is important that magazine subscribers receive their copy before it appears on newsstands, since this is seen by customers as one of the advantages of taking out a subscription.

·

Some magazines are also very time sensitive because they rely on advertiser revenue, and advertisers are unhappy if the magazine is delivered later than a competitor – this may be particularly important with weekly magazines.

We found that there was a general reluctance to pay for a higher price/higher quality service combination to ensure desired quality of service standards. An exception is during the Christmas period, where firms may upgrade the service they use (e.g. from second to first class standard letter tariff, or from Mailsort 3 to Mailsort 2). Other action includes sending out material (eg new chequebooks) earlier at Christmas. There was also concern over lack of knowledge of delays. Mailers cannot be sure individual items have arrived. There is also some difference between Royal Mail published quality of service statistics and mailers’ perceptions of what is being achieved for their own mail. Many of the firms we interviewed had made attempts to measure the service quality that they received, but it was interesting that many had not pursued this matter.

·

One firm had undertaken its own survey of mail sent to its own offices about 18 months previously and found that on-time delivery percentages were considerably below those published by Royal Mail – we were told only 60-70 per cent were delivered on time. Consequent discussions that the firm had with Royal Mail revealed that this was partly due to particular features of the way that the mail was prepared, and performance appeared to have improved – though the firm did not plan to carry out further surveys to check this. 11

n/e/r/a

Interviews with Major Postal Customers

·

The publisher interviewed had participated in studies to measure Royal Mail service quality using seeds, but this had been discontinued due to shortage of resources, and also because “ the service was not so bad to justify monitoring it”. There was still concern that Royal Mail was unable to tell them why magazines had not arrived, but the company recognised that the costs of tracking mail could well outweigh the benefits.

·

A mail order company paid for information to be provided on mail service performance, though noted that the information arrived too late to do much about it, and questioned whether even if the information were more up-to-date it would provide sufficient leverage over Royal Mail to do much about it.

·

A company selling products by mail order had employed consultants to monitor the quality of service it was getting, and they had found that this was consistently below that claimed by Royal Mail – but the company had discontinued a lot of their monitoring because Royal Mail had dismissed their consultant’s figures.

·

A very large mailer principally measures quality of service by sending and monitoring bills to its own home workers.

·

One mail order company undertakes random checks by phoning customers to see whether mail has arrived on time. In addition it holds regular meetings with Consignia to discuss mutual performance.

Response services are very important to some firms – though responses may be delayed because Royal Mail needs to count the items (and one firm had used them a lot less than before as a consequence of this). One firm pointed out that in part the associated expenditure is partly justified by better direction of mail within their organisation – use of response services would ensure that mail was delivered correctly to the organisation, so would not need to be re-directed if sent to the wrong department and/or geographical location. However, another company did not use pre-paid envelopes, and commented that if it did customers would use them for purposes other than the ones intended, for example to send general correspondence to the wrong address in the company. Firms are concerned that mail is delivered to their premises at the promised times to ensure efficient utilisation of their own staff. One firm told us that it was the timing of the delivery “ window” that was important. Both this, and other, firms said they would like their deliveries to be earlier in the day. Pick up times are also important – one mail order customer complained about late arrival delivery of trailers, which meant that their own staff had to be kept on later in the day, with consequent overtime payments, in order to fill them. Another respondent said that the biggest improvement that could be made would be later collection of Mailsort items. Firms also observed that there are differences in performance between different sorting centres. A major problem noted by one interviewee was the percentage of mail arriving with the first delivery. This had fallen considerably in the last nine months and presented particular 12

n/e/r/a

Interviews with Major Postal Customers

problems with payments, which could not be processed from the second delivery quickly enough to be paid into the banking system before 3.30 pm, with a consequent day’s delay in payment, and so additional interest costs. Similarly, customer correspondence might not be dealt with on a same day basis if it only arrived in the second delivery. This mailer also mentioned that the times at which the first delivery occurs are slipping (except for its largest receipt offices). Offices around the country used to get their first delivery mail by 7 or-7.30 am, but this is now slipping to around 9am. Homeworkers previously received their first delivery at around 8am, but this is now becoming 10 or 11am for most of them. Industrial disputes at Royal Mail can be a problem – especially when the customer is not informed. One mail order company specifically mentioned industrial relations disputes at Consignia as their biggest mail quality of service problem – though they noted a significant improvement in the extent of the problem since last year. A very large mailer noted that industrial action of more than two days would cripple their operation as their printing schedule was tight and so if mail was not printed they would not be able to clear a printing backlog. Even the threat of industrial action can divert considerable management time in developing contingency plans. Some firms were concerned about Royal Mail’s account management, including their ability to deal with operational issues or to communicate information on changes in the service supplied. A number of firms noted that this has improved in recent times. 4.2.3. The types of cost imposed by delays and other services quality problems There were only two cases where the firms interviewed could provide examples of estimates of the financial costs of delays. One was of the extra costs of using using Mailsort 2 rather than Mailsort 3 for statements at Christmas, and the other was of the costs of replacing missing issues of magazines. Other types of cost would be difficult to estimate e.g. the cost of losses of business as a result of mail losses as opposed to other reasons why a mail campaign has a poor response. Costs of dealing with customer complaints, or costs of late payment of invoices, are more straightforward, but we were not presented with estimates of costs of dealing with complaints that resulted from unsatisfactory mail service. If mail is not delivered and customers complain then the mailer will need to incur the cost of sending additional material – the publisher interviewed told us that persistent problems of this type were rare. We were told that on an annual mailing of tens of millions of items “ the number of serious delivery complaints probably runs into dozens”. The level of satisfaction with Royal Mail services varies between customers, and a number of the firms we interviewed made positive statements about the service they received:

·

“ Royal Mail are doing a good job, but the question is whether the price they charge represents value for money”.

13

n/e/r/a

Interviews with Major Postal Customers

·

A mail order company expressed the view that generally speaking the mail service is not bad.

·

One of the largest mailers noted that it did not have major quality of service problems with outgoing invoices.

One company was concerned that if Royal Mail volumes go down, fixed costs would have to be spread over fewer items, leading to a vicious circle as prices go up and volumes decline further.

14

n/e/r/a

The Stated Preference Survey: Method and Results

5.

THE STATED PREFERENCE SURVEY: METHOD AND RESULTS

5.1.

Introduction

As indicated in Section 1 of this report, one of the main techniques that we have used to estimate the cost to UK business of Royal Mail delays and other service quality problems has been a Stated Preference survey to provide valuations of different attributes of service quality by the full range of Royal Mail’s business customers. Before we discuss the design of this survey and its results, it is first necessary to set out the background to the choice for this approach. As we already pointed out in Section 4, there are only very few examples of companies directly calculating the cost of postal delays. Moreover, in our main survey, we found that only 10 per cent of the businesses experiencing problems could place a monetary value on the associated costs of delays to outward mail. We believe that there may be several reasons for companies not estimating the costs of postal delays, including:

·

the absolute magnitude of the costs may not justify the effort of quantifying them;

·

it is very difficult to quantify certain types of costs (e.g. loss of business); and

·

companies do not believe that calculating the costs serves any useful purpose, since they will not be compensated for the losses they incur.

We believe it to be likely that at least some of the companies that did quote a cost will not have made detailed calculations but are simply quoting an order of magnitude. It may also be the case that those companies that do quote a cost will be the most severely affected. In addition, certain types of costs to one company (e.g. loss of business) could actually be gains to other companies, so that the overall costs to the economy would actually be lower than the costs quoted by companies.

For these reasons, relying on direct estimates by companies of the cost of postal delays to estimate the total costs would produce completely unreliable results.

In view of the above, we believe that it is only possible to obtain an estimate of the costs of postal delays to businesses using a different approach, using customers’ valuation of service quality. In the appraisal of investment schemes, service quality (e.g. train delays) is an integral part of the evaluation. Service quality influences customers’ evaluation of a product, and a higher service quality means that the product becomes more valuable to them (and that they 15

n/e/r/a

The Stated Preference Survey: Method and Results

would be willing to pay more for it). An investment scheme may for example be judged to be worthwhile because of the reliability benefits it generates. Using a similar logic, a deterioration in service quality reduces the benefits of a product to consumers, reducing their willingness to pay for it. This would be counted as a cost in a social cost-benefit analysis.

Our approach to estimating the cost of postal delays to UK businesses has therefore been to analyse the benefits they derive from (and their consequent willingness to pay for) service quality. The current inadequate service quality levels reduce these benefits, thereby imposing a cost on the economy.

It is important to note the main limitation of the approach in the context of postal services. Unlike most other products in the economy, service quality in postal services may not just benefit the person who pays for the service. In some cases, postal delays impose costs on the receiver of an item as opposed to the sender. Since receivers of postal items do not normally pay for the postal service, it is not possible to obtain monetary valuations of postal service quality from them.4 Our approach focuses, with the exception of response services, solely on the senders of items and may therefore underestimate the true costs of postal delays somewhat.

5.2.

The Stated Preference Survey

Stated Preference (SP) is a widely-used technique to elicit valuations of attributes of service quality, such as the valuation of delays to road users. Consignia have themselves used SP to value attributes of Royal Mail service quality (such as reliability and delivery time) to business mail users as well as private customers. The broad approach used in the present study, of computer aided telephone interviewing (CATI), with initial contact by phone, and material sent in advance of the main interview, is consistent with that used by Consignia. SP techniques involve the presentation of hypothetical choice situations where each alternative, in this case the level of service of business post, is described by attributes describing the alternative, e.g. cost of delivery, delivery time, chance of delay, etc. Each of the attributes are described by a number of levels, e.g. costs can be varied from existing levels by a realistic range, say –50 per cent to 100 per cent. The attribute levels are combined using principles of experimental design to define different packages of services, which respondents evaluate in surveys. Respondents may be asked to rank the various packages, rate the packages using appropriate scales or choose one option from a pair or group of options. The latter approach is preferred on the basis that this is how respondents usually

4

In the case of response services, receivers do pay for the postal service and here, it would in theory be possible to obtain monetary valuations for service quality. We have therefore included second class response services in our survey.

16

n/e/r/a

The Stated Preference Survey: Method and Results

make choices in their everyday lives. Respondents, in this case senior business managers, would be asked to indicate which alternative they would choose: making trade-offs between the various attributes, e.g. more expensive postal costs for faster or more reliable delivery. The outputs from SP analysis provide estimates of:

·

the relative importance of each of the attributes, e.g. postal delivery time versus chance of delay; and

·

the trade-offs or marginal rates of substitution that respondents are willing to make between the attributes, providing indirect measurements of willingness-to-pay (in this case the willingness to pay to avoid delay).

A key advantage of SP analysis methods is that by requiring the respondent to trade between different variables (e.g. reliability and cost), they reduce the bias that can occur if respondents feel that they can make a political statement through their response (which may well be the case in a study like the present one). Following discussions with Postwatch it was decided to focus this study on six different segments of mail users:

·

First class service users;

·

Second class service users;

·

Mailsort 1 users;

·

Mailsort 2 users;

·

Mailsort 3 users; and

·

Second class response service users.

An SP questionnaire was designed to examine the valuations that businesses would place on different levels of postal service reliability. The SP surveys were undertaken using a telephone-fax-telephone methodology, where the person in the company who was responsible for postal decisions was contacted and invited to participate in a SP interview. An appointment date was fixed with respondents who were willing to participate in the survey and they were faxed (or e-mailed) an introduction letter describing the SP exercise and a set of SP choice pairs. The respondent was then called back to participate in the interview. Prior to the main survey, a pilot survey was conducted to test the questionnaire design and interviewing procedure.

17

n/e/r/a 5.3.

The Stated Preference Survey: Method and Results

Design of the SP Questionnaire

5.3.1. Outline of questionnaire structure The Stated Preference (SP) questionnaire was primarily designed to collect businesses’ valuations of reliability in postal services. The actual SP exercises, however, constitute only a section of the total questionnaire. Background questions are required to collect details about the characteristics of the respondent, which may be important in the modelling. These background questions also encourage the respondent to think about their current postal usage and the service characteristics of that service, prior to the main SP exercise. The questionnaire was therefore made up of a number of sections:

·

Recruitment questions in an initial questionnaire ;

·

Allocation of SP choice cards;

·

Overview of usage of postal service;

·

Users’ perception of their own service;

-

First class users’ perception of first class service;

-

Second class users’ perception of second class service;

-

Mailsort 1 users’ perception of Mailsort 1 service;

-

Mailsort 2 users’ perception of Mailsort 2 service;

-

Mailsort 3 users’ perception of Mailsort 3 service;

-

Response service users’ perception of second class response services;

·

Customer priorities and satisfaction;

·

Stated Prefence exercises;

·

Direct estimates of cost of delays;

·

Demographics of sample.

A copy of the full questionnaire is included as Appendix B. 5.3.2. Design of the stated preference exercises Following the investigation of a number of different Stated Preference specifications, it was decided to settle on a design which reflected the style of the level of service information that is readily available from the Consignia performance statistics. This offered significant benefits in the evaluation of the current service. It had the advantage also of being more 18

n/e/r/a

The Stated Preference Survey: Method and Results

accessible to respondents as it was consistent with the way that this information is presented in standard service updates. During the survey design, consideration was paid to the generation of choice experiments which contain sufficient information to make the choices interesting to respondents. A related issue is that of dominant choices, cases where one choice is obviously better than the other. In simple experimental designs with few variables there is a large number of potentially dominant choice combinations. In order to reduce dominant choices in the experiment, the reliability variables were supplemented by a variable presenting different delivery days. This also added an additional dimension of interest and allows potential future service packages that may be developed by Consignia to be examined. This also addressed concerns expressed by Postwatch that the project findings on reliability should be capable of addressing any changes in service that may emerge in the near future. The SP exercise reflected the style of service information that is readily available from Consignia performance statistics. The benefit of this presentation is that the information is presented in a manner that is consistent with the standard service updates. The SP exercise contained four variables:

·

advertised delivery day, which varied around the current service delivery day;

·

reliability of service, presented as two variables:

·

-

percentage of post that arrived by the delivery day;

-

percentage of post that arrived by the delivery day plus 3 days; and

cost of postage (per unit of mail) .

As well as adding an additional dimension of interest, the inclusion of the delivery day variable also deflected the respondent from concentrating solely on the reliability terms. The variables and levels used in the experiment are presented in Table 5.1. Table 5.1 Variables and Levels used in the SP Experiments Segment

First Class Second Class Mailsort 1 Mailsort 2 Mailsort 3 Response Users

Advertised Day of Delivery 1, 2 days 2, 3, 4 days 1, 2 days 2, 3, 4 days 6, 7, 8 days 2, 3, 4 days

% Post Delivered by Advertised Day

% Post Delivered by Advertised Day plus 3 Days

83, 89, 93, 95 83, 89, 93, 95 83, 89, 93, 95 83, 89, 93, 95 83, 89, 93, 95 83, 89, 93, 95

97, 98, 99, 99.5 97, 98, 99, 99.5 97, 98, 99, 99.5 97, 98, 99, 99.5 97, 98, 99, 99.5 97, 98, 99, 99.5

Cost (per unit posted) 25, 27, 28, 29p 17, 19, 20, 21p 22, 24, 25, 26p 14, 16, 17, 18p 14, 15, 15.5, 16p 17, 19, 20, 21p

19

n/e/r/a

The Stated Preference Survey: Method and Results

An example choice pair is presented in Figure 5.1 to demonstrate the format of the final choice cards. Figure 5.1 Example SP Choice Pair Choice 4

Which service would you choose for sending out your mail?

Service A

Service B

Advertised delivery time of 2 working days

Advertised delivery time of 3 working days

83%

will arrive within 2 working days

95%

will arrive within 3 working days

97%

will arrive within 5 working days

97%

will arrive within 6 working days

Service costs

14 p per item

Choose Service A

Service costs

Choose Neither

17 p per item

Choose Service B

Each respondent was presented with a set of eight choices pairs in their questionnaire, with varying advertised delivery times, proportions arriving within the advertised delivery day and the advertised delivery day plus 3 days, and cost of posting the item. Separate choice sets were developed for each of the six services:first class, second class, Mailsort 1, Mailsort 2, Mailsort 3, response services. As a result there was a total of 36 different choice sets, each comprising eight choice pairs. This provided a high level of variation in the data collected across the sample: a desirable property for model estimation. Following the recruitment of each respondent, a specific choice set was assigned to a respondent; this choice set was then faxed or e-mailed to the respondent with a covering letter. The letter described the purpose of the exercises, ensuring that the respondent understood which services they should be considering in the choices, and it provided a description of the variables that were to be considered in the choices. To ensure that this information was understood, the introduction was also read to the respondents during the interview just before they were asked to provide their answers to the SP choices.

5.4.

The Interviews

During the period 24th June – 12th July 2002 Market Probe conducted a total of 302 interviews of mail service users, using the SP questionnaire. These interviews were drawn from a database provided by NERA, and quotas of 50 interviews for each of the six different postal services were used to ensure sufficient data to allow examination of the variation in the

20

n/e/r/a

The Stated Preference Survey: Method and Results

valuations of reliability that may exist between these different groups of users. quotas were obtained for all the six different postal service segments.

These

The organisations surveyed came from the whole range of business users of Royal Mail services, so included government departments, local authorities, charities etc as well as purely commercial firms. The results therefore are not just based on responses of the largest business users of Royal Mail services.

5.5.

The Use of Royal Mail Services by the Companies Recruited

Respondents were asked in the main survey to provide the total monthly volume of letters sent from their location using any postal service. The responses to this question have been banded and are presented in Table 5.2 as the percentage falling in each band within each segment. Table 5.2 Average Number of Letters Sent Per Month by Any Postal Service Total

(%)

Do not know (%)

7.7

5.8

13.5

100.0

76.0

14.0

0.0

10.0

100.0

Mailsort 1

18.0

30.0

40.0

12.0

100.0

Mailsort 2

20.0

32.0

30.0

18.0

100.0

Mailsort 3

30.0

36.0

18.0

16.0

100.0

Second Class response services Total

66.0 47.4

28.0 24.5

4.0 16.2

2.0 11.9

100.0 100.0

1-1,000

1,001-10,000

10,001+

(%)

(%)

First Class postage

73.1

Second Class postage

(%)

It can be seen from Table 5.2, that across the sample as a whole, 88 per cent of the respondents were able to estimate their monthly volume of mail sent. As expected the high volume mailers are predominantly the Mailsort users. Table 5.3 shows the nature of the sector in which the respondents operated.

21

n/e/r/a

The Stated Preference Survey: Method and Results

Table 5.3 Nature of Business by Segment Financial Manufacturing Publishing Retail, Utilities Other Total Services Wholesale and Distribution (%) (%) (%) (%) (%) (%) (%) First Class postage

3.8

21.2

3.8

17.3

3.8

50.0 100.0

Second Class postage

12.0

Mailsort 1

14.0

16.0

4.0

12.0

4.0

52.0 100.0

6.0

12.0

10.0

6.0

52.0 100.0

Mailsort 2

10.0

-

14.0

6.0

4.0

66.0 100.0

Mailsort 3

2.0

2.0

12.0

18.0

2.0

64.0 100.0

Second Class response services Total

18.0 9.9

10.0 9.3

7.6

10.0 12.3

3.3

62.0 100.0 57.6 100.0

5.6.

Customer Priorities and Satisfaction

Respondents were also asked about their priorities in the postal service and their overall levels of satisfaction. Table 5.4 presents the aspects that the respondents believed were most important to their postal service, chosen from a pre-defined list. Across the sample as a whole the most important aspect was reliability, chosen by 43 per cent of all respondents. The time during the day in which the main delivery arrived and the security of the mail service were both given low ratings across all segments, suggesting these were not a primary concern for businesses. Respondents were also asked whether there was anything important missing from the list of service aspects. Only 3.6 per cent of the sample identified other priorities. Table 5.4 Users’ Identification of the Most Important Aspect of the Service Provided to Them by Royal Mail Price

Reliability Delivery Security time (%) (%) (%)

Total

(%)

Speed of delivery (%)

First Class postage

9.6

17.3

48.1

7.7

17.3

100.0

Second Class postage

16.0

32.0

38.0

8.0

6.0

100.0

Mailsort 1

16.0

26.0

42.0

8.0

8.0

100.0

Mailsort 2

20.0

22.0

50.0

4.0

4.0

100.0

Mailsort 3

24.0

18.0

40.0

8.0

10.0

100.0

Second Class response services Total

10.0 15.9

30.0 24.2

42.0 43.4

10.0 7.6

8.0 8.9

100.0 100.0

(%)

22

n/e/r/a

The Stated Preference Survey: Method and Results

When asked to quantify their satisfaction with the service provided by the Royal Mail, 28.8 per cent said they were very satisfied and 54.3 per cent said they were quite satisfied, while 5.6 per cent said they were quite dissatisfied, and 1.7 per cent were very dissatisfied. Just under 10 per cent were neither satisfied nor dissatisfied. Table 5.5 Users’ Level of Satisfaction with the Service Provided by Royal Mail Neither Quite Very Quite Very dissatisfied dissatisfied satisfied satisfied satisfied nor dissatisfied (%) (%) (%) (%) (%)

Total

(%)

First Class postage Second Class postage

0 2.0

3.8 2.0

5.8 8.0

53.8 58.0

36.5 30.0

100.0 100.0

Mailsort 1

2.0

16.0

6.0

58.0

18.0

100.0

Mailsort 2

4.0

4.0

12.0

50.0

30.0

100.0

Mailsort 3

0

4.0

14.0

54.0

28.0

100.0

2.0 1.7

4.0 5.6

12.0 9.6

52.0 54.3

30.0 28.8

100.0 100.0

Second Class response services Total

Following the Stated Preference exercises, the respondents were asked to identify their priorities in the postal service based around the types of decisions they were making in the SP choices. Table 5.6 shows that 51.7 per cent of the sample believed that in their real-life decisions they make a trade-off between the nature of the delivery and the cost of the service. Also of note is that more respondents were concerned with the delivery arriving by a set date rather than on a set date. Only 10 per cent of the sample stated that they made their decision on cost alone, with a large number of these responses coming from the Mailsort 3 segment who are choosing the cheapest available service in practice. Table 5.6 Priorities in Choosing Postal Service Largest proportion BY a set date (%)

Both Largest Cheapest proportion service nature and cost of ON a set available delivery date (%) (%) (%)

Total

(%)

First Class postage

25.0

25.0

1.9

48.1

100.0

Second Class postage

34.0

10.0

12.0

44.0

100.0

Mailsort 1

28.0

18.0

8.0

46.0

100.0

Mailsort 2

20.0

6.0

8.0

66.0

100.0

Mailsort 3

22.0

2.0

20.0

56.0

100.0

Second Class response services Total

24.0 25.5

16.0 12.9

10.0 9.9

50.0 51.7

100.0 100.0

23

n/e/r/a 5.7.

The Stated Preference Survey: Method and Results

Direct Estimates of the Impact and Cost of Delays

The survey asked about the cost of delays experienced by businesses, both in regard to outgoing and incoming mail. Table 5.7 shows that 55 per cent of respondents reported that delays impacted directly on their business. The reported effects were far greater for all the Mailsort categories. Table 5.7 Users’ Assessment of Whether Delays in Postal Services Affect Their Business Yes (%)

No (%)

Total (%)

First Class postage

36.5

63.5

100.0

Second Class postage

28.0

72.0

100.0

Mailsort 1

68.0

32.0

100.0

Mailsort 2

64.0

36.0

100.0

Mailsort 3

66.0

34.0

100.0

Second Class response services

68.0

32.0

100.0

Total

55.0

45.0

100.0

When asked to quantify the cost of these delays, few companies are able to place a figure on the problem; only about 10 per cent of the businesses experiencing problems could indicate a monetary value for these costs. These results are presented in Table 5.8. Table 5.8 Estimates of Annual Cost of Delivery Delays to Business £ 101£ 1,000 (%)

£ 1,001£ 10,000 (%)

£ 10,001 - £ 100,001 - Do not £ 100,000 £ 800,000 know (%) (%) (%)

Total (%)

First Class postage

5.3

-

5.3

-

89.5

100.0

Second Class postage

0.0

7.1

-

-

92.9

100.0

Mailsort 1

2.9

2.9

-

2.9

91.2

100.0

Mailsort 2

3.1

6.3

6.2

-

84.4

100.0

Mailsort 3

9.1

3.0

3.0

-

84.8

100.0

Second Class response services Total

3.6

3.0

2.4

0.6

100.0 90.4

100.0 100.0

Businesses were then asked whether they also experienced problems with delays in mail sent to them from other companies or customers. Results are shown in Table 5.9. This question demonstrated a lower number of companies experiencing problems as a result of delays in receiving mail, with 37 per cent of companies reporting problems. 24

n/e/r/a

The Stated Preference Survey: Method and Results

Table 5.9 Users’ Assessment of Whether Delays in Incoming Mail Affect Their Business Yes (%)

No (%)

Total (%)

First Class postage Second Class postage

23.1 30.0

76.9 70.0

100.0 100.0

Mailsort 1

38.0

62.0

100.0

Mailsort 2

50.0

50.0

100.0

Mailsort 3

44.0

56.0

100.0

Second Class response services Total

40.0 37.4

60.0 62.6

100.0 100.0

Again, businesses had problems quantifying the cost to their business of delays in the mail service. Results are shown in Table 5.10. Only 15 per cent of those reporting delays in receiving mail were able to quantify the costs. Table 5.10 Estimates of Annual Cost of Delivery Delays of Incoming Post to Business

(%)

£ 1,001£ 10,000 (%)

First Class postage

-

-

8.3

-

91.7

100.0

Second Class postage

-

13.3

13.3

-

73.3

100.0

£ 1-£ 1,000

£ 10,001 - £ 100,001 - Do not £ 100,000 £ 800,000 know (%) (%) (%)

Total (%)

Mailsort 1

-

5.3

-

5.3

89.5

100.0

Mailsort 2

8.0

4.0

4.0

4.0

80.0

100.0

Mailsort 3

9.1

4.5

4.5

-

81.8

100.0

Second Class response services Total

3.5

4.4

4.4

1.8

100.0 85.8

100.0 100.0

5.8.

Valuations of Postal Reliability

5.8.1. Results of the SP analysis Discrete choice models were estimated from the stated preference choices made by the respondents. The model estimation procedure assumes that each respondent evaluates each postal service profile on the basis of the characteristics presented and that these characteristics can be combined into a utility function that expresses the preference that an individual has for a given postal service profile, i.e.:

25

n/e/r/a

The Stated Preference Survey: Method and Results

UService A = badvert * advertised_A + b%by_day * %by_day_A + b%by_3_day * %by_3_day_A + bcost * cost_A UService B = badvert * advertised_B + b%by_day * %by_day_B + b%by_3_day * %by_3_day_B + bcost * cost_B UNeither

= bNeither

Additional terms and non-linear variations were tested during the model calibration process. Respondents are assumed to choose the alternative with the highest utility. It is the model coefficients (b) that are estimated in the model calibration procedure. By dividing the estimates of the service coefficients by the cost coefficient estimate, we observe the willingness to pay for the service components. Full technical details of the model estimation procedures and the results are provided in Appendix F. The SP data that were collected have proved a suitable basis for the development of satisfactory models indicating reasonable valuations of the key aspects of postal services, delivery speed and reliability, for most of the market segments studied. All of the coefficient values have reasonable values – of the correct sign – and nearly all are well estimated and significantly different from zero. For first class users the reliability service coefficients were estimated satisfactorily, but the cost coefficient was not sufficiently reliable. To deal with this problem the data from the first class and Mailsort 1 users were pooled to provide an estimate of the cost coefficient for both groups of next day service users jointly. However, no such assumption was made about the values of other coefficients in the models, and for Mailsort 1 valuations only the observations of Mailsort 1 users were used. For response users it has not been possible to attach monetary valuations to service aspects, although the relative valuations are well based. In this regard, it is interesting to note that only 10 per cent of response service users indicated that price was an important service aspect (see Table 5.4 for details), indicating low price sensitivity for this segment. For this reason, the valuations for response services users have been derived jointly using the data from the second class and the response service users together. The models measure sensitivity of cost for the different segments. No other variables were observed to affect cost sensitivity, although tests were made on company size, turnover and type of business. The linearity of the impacts of the delivery day and reliability variables were examined to see whether mailers value changes in particular aspects of service quality changes differently as the level of service quality provided changes. The only significant non-linear effect that was found was in the valuation of advertised day reliability for first class users, where there was some evidence that first class users do not attach significant value to reliability improvements that remain at a low level (between 83 per cent and 89 per cent), but that they are prepared to pay more for improvements if good levels of service have already been achieved. However, since we did not undertake experiments in the range 26

n/e/r/a

The Stated Preference Survey: Method and Results

where first class reliability exceeded 95 per cent, we were not able to investigate the extent to which marginal reliability valuations by first class mail users differed in this range. Consequently, for valuations of the total costs of delays we used marginal valuations for first class (as well as for other mail users) that did not vary by reliability level. Tables comparing observed and predicted choices were used to examine effects of other background variables on respondents’ choices. Three significant effects were observed:

·

increased valuations for reliability for second class post users in the utility sector;

·

increased valuations of speed of delivery for Mailsort 1 and 2 services for business in the financial sector;

·

increased valuations of speed of delivery and reliability for response services used in the retail sector.

The final models were subjected to jack-knifing tests (see Appendix F) to allow for the impact of correlated multiple responses from individual respondents. These tests did not lead us to change the model specifications, though some coefficient values were changed. 5.8.2. Summary of the final values Valuations of the advertised day of delivery and percentage of post received by the advertised day were derived from the choice models. The valuations of changes in the advertised day of delivery are presented in Table 5.11 All valuations are measured relative to the current advertised service. Faster delivery times are valued positively; slower delivery times are valued negatively.

27

n/e/r/a

The Stated Preference Survey: Method and Results

Table 5.11 Valuation for Changes in Advertised Day of Mail Delivery (pence) Advertised delivery day Day 1

First Class

Mailsort 1

Second Class

Mailsort 2

Financial Mailsort 2

Response Services

0.0

Financial Mailsort 1 0.0

0.0

Day 2

-6.8

-4.0

-7.3

5.6

1.9

3.6

8.9

Day 3

0.0

0.0

0.0

0.0

Day 4

-5.6

-1.9

-3.6

-8.9

Mailsort 3

Day 5 Day 6

0.9

Day 7

0.0

Day 8

-0.9

In regard to reliability, the valuations of marginal changes in the percentage of post delivered by the advertised day are presented in Table 5.12. Table 5.12 Marginal Valuations for Additional Changes in Reliability, i.e. Percentage of Post Delivered by the Advertised Day (pence) First class Mailsort 1 Second class Mailsort 2 Mailsort 3 Response services

Per % point

(p)

(p)

(p)

(p)

(p)

(p)

0.82

0.44

0.32

0.27

0.17

0.85

Table 5.12 reveals a broad consistency of results, in line with our own expectations, in that:

·

Values for improved reliability are higher for standard first class mail than for standard second class mail;

·

Values for improved reliability are higher for Mailsort 1 than for Mailsort 2, and higher for Mailsort 2 than for Mailsort 3;

·

Values for improved reliability are higher for standard first class mail than for Mailsort 1; and

·

Values for improved reliability are higher for standard second class mail than for Mailsort 2.

28

n/e/r/a

The Stated Preference Survey: Method and Results

Both the high values for reliability of first class mail, and the high values for day of delivery reported in Table 5.11, confirm its status as the premier service and the one on which users rely for important items.

29

n/e/r/a

The Overall Costs of Postal Delays

6.

THE OVERALL COSTS OF POSTAL DELAYS

6.1.

Methodology

This section explains the methodology for aggregating up to the total costs of postal delays in 2001/02. We estimate two broad types of total cost, based on two measures of the performance standard:

·

cost of delays compared with current target quality of service performance; and

·

cost of delays compared with an ideal (ie 100 per cent) quality of service performance.5

The way in which the total costs for each category of mail were derived is as follows:

Cost Cost of of delay delay in mail in mail category category in in 2001/02 2001/02

Valuation Valuation per per % % point point improvement improvement in in reliability reliability

5

X

=

Gap Gap between between actual actual performance performance and and performance performance standard standard

X

Volume Volume of of mail mail in in the the mail mail category category sent sent by by business business users users

There is an analogy between this and most estimates of road congestion, and of road congestion costs, where these are measured as the difference between actual journey costs or travel times, and the travel costs and times that would be achieved if there were no delays at all on the existing road network.

30

n/e/r/a 6.2.

The Overall Costs of Postal Delays

Results

Table 6.1 shows the results of these calculations. Table 6.1 Costs of Postal Delays in 2001/02 Costs of Not Meeting Existing Performance Standards

Costs of All Delays

(£ m)

(£ m)

First class

92

288

Second class

15

41

Mailsort 1

4

18

Mailsort 2

20

43

Mailsort 3

4

16

Second class response

4

10

Other

18

41

Total

156

456

Mail Type

The total cost of delays compared with existing standards is estimated at £ 156 million a year. The total cost of delays compared with ideal standards is estimated at £ 456 million a year.

31