Measuring Trust in Retail Banking in China

Christine Ennew
Nottingham University Business School
On Behalf of The Financial Services Research Forum

1
INTRODUCTION
It is widely acknowledged that trust plays a central role in the way in which
financial services organisations present themselves to their customers and this is
particularly apparent in retail markets. The combination of intangibility, product
complexity, and the long term nature of many products means that customers
face high levels of risk in making purchase decisions; they will often have
difficulty in judging product performance and will need to trust financial services
institutions (FSIs) to offer products of an appropriate type and quality. The
importance of trust for an efficient and effective financial sector cannot be
understated. Trust underpins retail customers’ willingness to engage with the
financial services sector. If the financial services sector is to fully mobilise private
savings to support economic development and is to provide individuals with the
means of managing risk and maintaining financial well being, it must command
the trust of its customer base.

Although clearly important, our understanding of the concept of trust can be both
variable and imprecise. Its importance as a means of ensuring stability within
social systems and facilitating economic transactions is widely accepted; its
precise meaning is open to rather more debate and rigorous empirical evidence
is in short supply. This paper reports on work undertaken by the Centre for
Global Finance at the University of Nottingham Ningbo, China in conjunction with
the Financial Services Research Forum to measure levels of trust in retail banking
in China. The paper begins with an overview of trust in financial services and
then moves on to present a framework for measuring trust. The next section
presents the result of a large scale survey and examines both levels and drivers
of trust. The paper closes with a summary and conclusions.

TRUST AND FINANCIAL SERVICES

In a business context, much of the discussion about the meaning of trust has its
origins in literature relating to organizations and organizational analysis. Within
this body of research, trust has attracted the interests of psychologists,
sociologists, economists and management researchers; as a consequence there
are a variety of different approaches to understanding the concept of trust and a

2
variety of different definitions1. What is apparent from the many different
definitions and approaches to trust is that there are certain key themes that
emerge and appear to be recognised as integral to the concept of trust2;

trust depends on the existence of risk – if the outcomes of a particular

action are certain, then there would be no need to trust;

trust depends on interdependence between actors – if actors are not
somehow dependent on each other, there is no need to trust;

trust is associated with vulnerability – risk and interdependence create
vulnerability;

trust involves confident expectations about future behaviours – an actor
will only accept vulnerability in the presence of strong expectation of the
positive future behaviour of another actor;

some form of trust is likely to be inherent in most relationships – few
relationships are or can be characterised by complete certainty or
complete contracting.

Thus, a common view of trust would suggest that it is concerned with an

individual’s willingness to accept vulnerability on the grounds of positive
expectations about the intentions or behaviour of another in a situation
characterised by interdependence and risk. Existing perspectives on trust
suggest that this willingness to be vulnerable may arise from a calculation of
costs and benefits, an individual’s existing predisposition to trust, a detailed
knowledge and understanding of exchange partners and/or a faith in social
systems and institutions.

Thus, to the extent that there is vulnerability, risk and interdependence

associated with the purchase of financial services, then there will be a role for
trust3. From the perspective of retail consumers, the buying process for financial

1
Rousseau, D M, Sitkin, S B, Burt, R S, Camerer, C (1998) ‘Not So Different After All: A
Cross Discipline View of Trust, Academy of Management Review, Vol. 23 (3) p393-104
2
Sheppard, B., and Sherman, D., (1998), “The Grammars of Trust: A Model and General
Implications”, Academy of Management Review. Vol. 23, No. 3, pp. 422-437.
3
Berry, L., (1995), “Relationship Marketing of Services – Growing Interest, Emerging
Perspectives”, Journal of the Academy of Marketing Science. Vol. 23, pp. 236-245.

3
services can be complicated because of the variety and complexity of the products
available and the difficulties of making comparisons between different products.
Despite the relatively recent development of the retail banking sector in China,
there is, nevertheless, a considerable variety of traditional banking products
available alongside the variety of insurance based products that are increasingly
being distributed via bank branch networks. Even what might be regarded as
relatively simple products, such as bank accounts, savings products and loans may
appear complicated to many consumers and difficult to compare. In addition,
financial services are generally considered to be uninteresting to large sections of
the retail market and, accordingly, many customers are unwilling to invest time
and effort in trying to understand such products. The difficulties associated with
understanding financial services are compounded by the inability to judge how well
they will perform in the future; the consumer can only assess the product once it
has been bought and even then, that assessment may take a number of years as
is the case with products designed for the longer term. With many protection
products, performance evaluation may only take place in the event that a
customer needs to make a claim and for many customers, this maybe never.

Even in the case of products whose performance can be assessed over a rather
shorter term period, the consumer may find difficulties in assessing product
performance. For many products, performance depends on both the skills of the
product provider and the future performance of the economy as a whole. The
performance of similar product types may vary considerably according to the time
period over which they are assessed and the timing of initial purchase. Poor
performance might be due simply to bad fortune and timing. This forces the
consumer to rely heavily on credence qualities – on trust and confidence in what a
provider does and has done4. Indeed, it has been suggested that the main
function of trust is to reduce uncertainty thereby making risks manageable and
simplifying choice5.

4
Diacon, S R and Ennew C T (1996) `Can Business Ethics Enhance Corporate
Governance : Evidence from a Survey of UK Insurance Executives, Journal of
Business Ethics vol 15 pp 623-634
5
Morrison, D., and Firmstone, J., (2000), “The Social Function of Trust and Implications
for E-Commerce”, International Journal of Advertising. Vol. 19, pp. 599-623.

4
Based on this explanation of the purchasing process, the conditions that make
trust important are clearly in evidence in relation to financial services;

Risk – consumers identify significant risks associated with financial

services because they are effectively surrendering some control over their
finances to a third party. Risks are associated with poor product
performance, which may be due to the poor quality of the product but
could equally be due to misfortune. Risk is inherent in financial services
but is compounded by consumer’s typically low levels of understanding
and the impacts of uncontrollable factors.

Vulnerability – since financial services can and do have a significant impact

on the consumer’s well being, a poor performing product can have a very
significant impact on individual customers. For example, savings product
with poor returns or an insurance product that does not provide the
expected cover can easily have significant adverse impacts on consumers.

Interdependence – the functioning of financial markets means that in

general, individuals need the services of a specialist intermediary to deal
with their financial needs. More significantly, product variety and
complexity mean that the customer is dependent on a financial services
organisation for advice and the more limited the customer’s understanding
of financial services, the greater the dependence on a financial services
provider or a financial adviser.

The specific features of financial services and the importance played by front line
staff and sales staff means that trust may be based in the organisation, the
brand or the individual; traditionally, trust in the individual has always been of
considerable significance to the industry. Increasingly though, as methods of
distribution change, the importance of the brand as the basis for a relationships
and a basis for trust is increasing6.

6
Dall’Olmo-Riley, F (2000) ‘The Service Brand as Relationship Builder’, British Journal of
Management, vol 11(2) pp 137-150

5
DEVELOPING THE TRUST INDEX
While there is widespread recognition of the importance of trust, attempts to
measure trust all too often rely on simple measures of the concept, using
perhaps single statements and often simple yes/no answers. Research in the
area of trust (and the related concept of trustworthiness) highlights the richness
and complexity of both concepts. Accordingly, in attempting to evaluate and
monitor trust and trustworthiness in the financial services sector, the Financial
Services Research Forum sought to develop a more robust and rich measure. The
measurement framework, which is outlined in Figure 1, proposed that consumer
trust in an organisation may be low level or cognitive (ie based around notions of
reliability and dependability) or high level or affective (ie based around notions of
being concerned about the best interests of the customer).

Figure 1: A Framework for Measuring Trust in Financial Services

Consumers’ disposition to
Expertise and trust
Competence

Integrity and
consistency

Organisational Trust in the

Communications trustworthiness organisation

Shared values
Cognitive Affective

Concern and
benevolence

In line with marketing studies which have reported close associations between
these forms of trust it is suggested that cognitive trust can lead to affective trust.
Consumer trust is also related to individual characteristics, reflecting work within
the psychology tradition which would argue that consumers may have different
dispositions to trust. The other major determinant of trust is organisational

6
trustworthiness which is determined by expertise and competence, integrity and
consistency in behaviour, effective communications, shared values and concern
and benevolence.

Multi-item Likert-scales were develop to measure each of the above concepts,

working with the following definitions (see Appendix):

Trust Consumers’ trust in a financial services institution (FSI).

This is an attribute of consumers. Trust may vary across
consumers because of different experiences and
personality traits even where perceptions of
trustworthiness are similar.

Trustworthiness The extent to which an FSI is perceived as being worthy of

trust.
This is an attribute of the FSI; it is central to the image
and reputation of the institution and can be managed by
both internal policy and practice and through external
communications.

Benevolence The extent to which an FSI is concerned about its

customer’s interests from a customer perspective.

Integrity The extent to which an FSI is honest and consistent in

what it does from a customer perspective.

Ability/Expertise The extent to which an FSI is seen as having the

necessary skills and ability to deliver its services from a
customer perspective.

Shared values The extent to which consumers believe that an FSI has
values similar to their own.

Communications The extent to which an FSI communicates well/effectively

from a customer perspective

7
In addition, data was also collected in relation to the concept of system trust, to
assess the extent to which individuals believed that the financial system could be
relied upon to protect them and their interests. Although not formally part of the
model, the system trust concept is included in the detailed analysis in the next
section.

CUSTOMER TRUST IN RETAIL BANKING IN CHINA

Using the measurement instrument developed by the Financial Services Research
Forum, a survey of retail bank customers in China was carried out during
December 2007 and January 2008. A total of 1000 customers in the Yangtze
River Delta region were interviewed about their relationship with their bank; this
area was chosen as being one of the most economically developed regions of
China and thus one in which consumers were likely to have a greater diversity of
banking needs, thus ensuring that the data collected related to broad and diverse
banking relationships. Respondents were asked very specifically about their
relationship with their bank and were also asked to provide comparative ratings
of other non-financial institutions (employer, supermarket and mobile phone
provider) to provide a benchmark against which the performance of the banking
sector could be evaluated.

As Table 1 shows, the sample was spread across the key metropolitan areas of
the Yangtze Delta, with the largest number of respondents coming from
Shanghai.

Table 1: Sample Distribution by City

Count Col %
City Shanghai 400 40.0%
Nanjing 150 15.0%
Suzhou 150 15.0%
Hangzhou 150 15.0%
Ningbo 150 15.0%
Group Total 1000 100.0%

8
There is evidence of a slight bias towards female as table 2 shows, although this
is not large enough to give any cause for concern.

Table 2: Sample Distribution by Gender

Count Col %
Gender Male 456 45.6%
Female 544 54.4%
Group Total 1000 100.0%

The sample is dominated by relatively younger customers as Table 3 shows, with

some 50% of respondents being under 35. While direct comparisons between the
sample and the population of bank customers is not possible, there are no
reasons to believe that this pattern is not representative.

Table 3: Sample Distribution by Age

Count Col %
Age group 18 to 24 years 154 15.4%
25 to 29 years 200 20.0%
30 to 34 years 167 16.7%
35 to 39 years 131 13.1%
40 to 44 years 137 13.7%
45 to 49 years 121 12.1%
50 to 54 years 70 7.0%
55 to 59 years 11 1.1%
60+ 9 .9%
Group Total 1000 100.0%

As would be expected, the sample is dominated by relatively higher income

groups with over 50% of respondents earning in excess of RMB5,000 per
month1.

1
Renmimbi, the Chinese currency, currently about 13 to the £

9
 Table 4: Sample Distribution by Income

Count Col %
Monthly RMB 0-999 2 .2%
household RMB1,000-1,499 13 1.3%
income RMB1,500-1,999 17 1.7%
RMB2,000-2,499 61 6.1%
RMB2,500-2,999 41 4.1%
RMB3,000-3,499 121 12.1%
RMB3,500-3,999 59 5.9%
RMB4,000-4,499 96 9.6%
RMB4,500-4,999 63 6.3%
RMB5,000-5,499 134 13.4%
RMB5,500-5,999 50 5.0%
RMB6,000-6,499 76 7.6%
RMB6,500-6,999 31 3.1%
RMB7,000 or above 236 23.6%
Group Total 1000 100.0%

Table 5 shows the distribution of respondents by bank, with the dominance of 3

of the big four being clearly in evidence. The Industrial and Commercial Bank of
China (ICBC) has the largest number of respondents followed by the Agricultural
Bank of China (ABC) and China Construction Bank (CCB). China Merchants Bank
(CMB) and Bank of China (BofC) have similar numbers of respondents, with Bank
of Shanghai (BofS) and Bank of Communications (BofComm) being the smallest
of the 7 identified banks.

Table 5: Sample Distribution by Bank

Count Col %
Bank ICBC 387 38.7%
ABC 167 16.7%
CCB 124 12.4%
CMB 79 7.9%
BofC 82 8.2%
BofS 32 3.2%
BofCom 45 4.5%
Others 84 8.4%
Group Total 1000 100.0%

Table 6 provides summary information on the overall measures of trust and

trustworthiness. In contrast to many trust measures, which rely on a simply
yes/no comparison, the approach adopted here measures degree of trust. The

10
trust/trustworthiness measures were constructed by averaging across responses
to a series of statements and then scaled so that the maximum possible score is
100 and the minimum is 0. To interpret these figures, a score of 100 would mean
that all respondents strongly agreed with every statement on trust. A score of 75
would indicate that on average respondents moderately agreed with the
statements on trust and a score of 50 would indicate that respondents neither
agreed nor disagreed with the various statements.

Table 6: Whole Sample Measures of Trust

Base Higher Trustwort

Overall trust level trust level trust hiness
Mean 74.30 75.00 72.89 74.31
Std Deviation 15.28 15.68 18.74 16.24
Minimum .00 .00 .00 .00
Maximum 100.00 100.00 100.00 100.00

Thus, in looking at Table 6, overall consumer trust is slightly below the figure of
75 suggesting that on average, respondents are moderately trusting of their
bank. Base level (cognitive) trust is significantly larger than high level (affective)
trust as might be expected – that is to say, respondents are more convinced
about the reliability/dependability of banks and less convinced about the extent
to which banks have their interests at heart. Respondents’ perceptions of the
extent to which banks are trustworthy is significantly higher than the reported
level of overall trust. This would suggest that the reputation that banks project
may promise more trust than consumers are willing to offer, perhaps reflecting
variations in individual dispositions to trust. However, the observed difference is
relatively small.

The factors that are believed to influence trust are outlined in Table 7 and the
figures should be interpreted in the same way as the figures relating to the
aggregate measures. Based on these results, banks attract their highest ratings
in relation to system trust, ability/expertise and integrity and are weakest in
relation to communication and shared values, both of which display a high
degree of variability relative to both expertise/ability and integrity.

11
 Table 7: Whole Sample Drivers of Trust

Ability/exp Shared Communi

Benevolence ertise Integrity values cations System trust
Mean 74.61 75.91 75.74 71.53 70.87 78.05
Std Deviation 15.25 15.18 15.83 17.09 17.60 15.88
Minimum .00 12.50 15.00 .00 .00 .00
Maximum 100.00 100.00 100.00 100.00 100.00 100.00

Figure 1 provides a comparative analysis of trust across leading banks in the

sample. ICBC and Bank of China record some of the strongest scores across
different measures of trust and trustworthiness while China Construction Bank,
Bank of Shanghai and Bank of Communications receive rather weaker customer
ratings.

Figure 1: Overall Measures of Trust

80.00

75.00
Trust index

Overall trust

70.00 Base level trust

Higher level trust
Trustw orthiness

65.00

60.00
ICBC ABC CCB CMB BofC BofS BofCom

Bank

Figure 2 combines trust ratings for banks with the ratings reported for
comparator institutions, providing clear evidence that customers display
relatively higher levels of trust in their bank, a finding that is consistent with
results from surveys undertaken in the UK, Malaysia and Oman.

12
 Figure 2: Overall Measures of Trust
80.00

75.00
Trust index

Overall trust
Base level trust
70.00
Higher level trust
Trustw orthiness
65.00

60.00
C

fC
BC

fS

er
CB

e
om

t
ke
M
AB

on
Bo

Bo

oy
IC

fC

ar
Ph
pl

rm
Bo

Em

pe
Su
Institution

An overview of the individual drivers of trust is presented in Figure 3, broken

down by bank. Consistent with the aggregate figures, customers of ICBC and
Bank of China provide high ratings for their bank across most drivers of trust and
both communications and shared values are uniformly weak. System trust, which
relates to overall trust in the financial system is consistently highly rated across
all banks suggesting a high degree of confidence on the part of individuals.

Figure 3: Drivers of Trust

85.00

80.00

Benevolence
Trust index

75.00 Ability/expertise
Integrity
Shared values
70.00
Communications
System trust
65.00

60.00
ICBC ABC CCB CMB BofC BofS BofCom

Bank

13
Two variables likely to have a significant impact on trust and trustworthiness are
age and length of relationship because both may have a bearing on customer
knowledge and experience. Research in the UK has shown that age in particular
has a positive impact on recorded levels of trust with older consumers displaying
markedly higher levels of trust than younger consumers. As figures 4 and 5
show, there is evidence of a similar trend among customers of Chinese banks.
Trust evaluations dip for consumers in the 30-40 age group, with a pronounced
increase among those respondents who are over 40.

Figure 4: Overall Measures of Trust by Age

80

75
Trust index

Overall trust

70 Base level trust

Higher level trust
Trustw orthiness

65

60
Under 30 30-40 40-50 Over 50

Age

14
 Figure 5: Drivers of Trust by Age
85

80

Benevolence
Trust index

75 Ability/expertise
Integrity
Shared values
70
Communications
System trust
65

60
Under 30 30-40 40-50 Over 50

Age

Of course, age and length or relationships are closely related and an analysis by
length of relationship, as shown in Figures 6 and 7 displays a similar pattern with
higher levels of trust reported by those customers who had maintained longer
relationships with their bank.

Figure 6: Trust by Length of Relationship

80

75
Trust index

Overall trust

70 Base level trust

Higher level trust
Trustw orthiness

65

60
Under 3 years 3-5 years 6-10 years Over 10 years

Relationship Length

15
Detailed analysis of the relative impact of both age and relationship length
suggests that it is age which is the more important in explaining changes in
levels of trust. However, there is evidence that relationship breadth – ie the
number of products that a customer holds – does influence levels of trust.
Customers holding 3 or more products from a given bank report higher levels of
trust than those holding fewer that 3 products.

Figure 7: Drivers of Trust by Length of Relationship

85

80

Benevolence
Trust index

75 Ability/expertise
Integrity
Shared values
70
Communications
System trust
65

60
Under 3 years 3-5 years 6-10 years Over 10 years

Relationship Length

There is also evidence to suggest that trust ratings vary significantly by gender
with female customers reporting higher levels of trust than males. This pattern is
clearly in evidence across both the aggregate measures of trust and the drivers
of trust, as Tables 8 and 9 demonstrate.

Table 8: Overall Measures of Trust by Gender

Base Higher Trustwort

Overall trust level trust level trust hiness
Male Mean 72.37 73.26 70.59 72.26
Female Mean 75.91 76.46 74.82 76.03

16
 Table 9: Drivers of Trust by Gender

Ability/exp Shared Communi

Benevolence ertise Integrity values cations System trust
Male Mean 72.66 74.37 73.98 69.06 68.17 76.17
Female Mean 76.23 77.19 77.21 73.59 73.13 79.63

There is weak evidence of a relationship between income and trust with a slight
tendency for higher income earners to report higher levels of trust.

The analysis thus far highlights the importance of a variety of personal factors
that influence respondents’ assessments of levels of trust and trustworthiness in
relation to their bank. In order to understand what lies behind these perceptions,
the framework introduced earlier in this paper suggested that trustworthiness
would be determined by assessments of a bank in terms of ability/expertise,
benevolence, communication, integrity and shared values. Regression analysis
provides insights into the relative importance of these different drivers. Figure 8
shows the relative size of the coefficients from a regression analysis and it is
clear that benevolence and integrity are the most important influences on
trustworthiness. In contrast, communications, although highlighted earlier as an
area in which the banks generally were perceived to be weak, has virtually no
influence on perceptions of trustworthiness.

Figure 8: Drivers of Trustworthiness

0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
Benevolence Ability/expertise Integrity Shared values Communications

17
CONCLUSIONS
This analysis of trust in retail banking demonstrates that customers do have a
moderately high level of trust in their bank, particularly when compared to
mobile phone providers, employers and supermarkets. Levels of trust do vary
across banks with ICBC and Bank of China scoring relatively highly while ABC
and CCB have relatively low scores. Trust tends to increase with both age of
customer and length of relationship with bank and female customers display
higher levels of trust than males. Customers with a broader relationship – ie
holding more than 3 products display higher levels of trust than customers with
fewer products, highlighting the importance of trust in enabling banks to cross
sell to their customer base. Perceptions of bank integrity and benevolence
(having the customers best interests at heart) are the most important drivers of
consumer trust, suggesting thank if banks are looking to enhance levels of
consumer trust, a focus on enhancing reputation in these areas would be the
most effective strategy.

18
Appendix:
Measures of Trust and Trustworthiness

Strongly Agree
Disagree
Strongly

Neutral
My main BANK …..

I trust my bank to do what it says it will do 1 2 3 4 5

I trust my bank to have my best interests at 1 2 3 4 5
heart
My bank is very reliable 1 2 3 4 5
My bank is always honest with me 1 2 3 4 5
My bank is concerned about my best interests 1 2 3 4 5
Overall I feel I can trust my bank 1 2 3 4 5
My bank makes every effort to address my 1 2 3 4 5
needs
My bank has a reputation for being reliable 1 2 3 4 5
My bank has a reputation for being honest 1 2 3 4 5
My bank has a reputation for being dependable 1 2 3 4 5
My bank has a reputation for looking after its 1 2 3 4 5
customers
My bank has a reputation for having its 1 2 3 4 5
customers interests at heart
Overall I feel my bank is trustworthy 1 2 3 4 5

Table A2: Measures of the Drivers of Trustworthiness

Strongly Agree
Disagree
Strongly

Neutral

My main BANK …..

Does whatever it takes to make me happy 1 2 3 4 5

Keeps its word 1 2 3 4 5
Acts in the best interests of its customers 1 2 3 4 5
Shows high integrity 1 2 3 4 5

19
Is honest 1 2 3 4 5
Conducts transactions fairly 1 2 3 4 5
Has the information it needs to conduct its 1 2 3 4 5
business
Is consistent in what it does 1 2 3 4 5
Can be relied upon to give honest advice 1 2 3 4 5
Shows respect for the customer 1 2 3 4 5
Treats customers fairly 1 2 3 4 5
Has the same concerns as me 1 2 3 4 5
Is receptive to my needs 1 2 3 4 5
Competently handles all my requests 1 2 3 4 5
Is efficient 1 2 3 4 5
Communicates clearly 1 2 3 4 5
Is responsive when contacted 1 2 3 4 5
Informs me immediately of any problems 1 2 3 4 5
Has the same values as me 1 2 3 4 5
Informs me immediately of new developments 1 2 3 4 5
Acts as I would 1 2 3 4 5
Is knowledgeable 1 2 3 4 5
Communicates regularly 1 2 3 4 5

20