An Analysis of Firms’ Adoption of International Standards: One
Size Fits All?
Professor John Hudson* and Assoc. Professor Marta Orviska**
ABSTRACT
We analyse the take-up by firms of internationally recognized standards such as ISO 9000 and ISO
14000. Based on an analysis of 11,668 firms in countries in Asia and Eastern Europe, we conclude
that the probability of standard certification increases with firm size, is greatest in large cities and in
manufacturing industries. There are other differences, including between countries. Given these
differences, we argue that single generic standards for all firms may not be optimal and that there is a
case for the simultaneous publication of differentiated standards targeted at different user
characteristics.
*Department of Economics, University of Bath, Bath, BA2 7AY, United Kingdom; email:
[email protected]. [corresponding author].
**Faculty of Economics, Matej Bel University, Tajovskeho 10, 975 90 Banska Bystrica, Slovak
Republic; email: [email protected]
The authors gratefully acknowledge valuable comments and suggestions on an earlier version of the
paper. This paper was presented at the 16th EURAS Annual Standardization conference in Kaunas,
Lithuania, June 2011.
Key words: ISO 9000; ISO 14000; firms; standards; product quality
JEL: L15, L25, O19.
Running title: Firms’ Adoption of International Standards
1
An Analysis of Firms’ Adoption of International Standards: One Size Fits All?
1 Introduction
Standards are an important feature of modern life. They have been fundamental both in the development of the
European Union and in the development of complex modern technologies such as the mobile phone. There are
many examples of standards. Two of the better known are the ISO 9000 and ISO 14000 families of standards
which are examples of internationally recognized quality standards pertaining to management. Standards tend to
be arrived at by ‘consensus’ and it is a general rule that a single, generic standard emerges which all firms are
then able to benefit from. However, standards are often, if not always, the result of a compromise between
various interested parties. But being they are the result of compromise, are they equally suited to all firms, large
as well as small, service firms as well as manufacturing ones and firms in different countries? Is the single,
generic standard designed for all users appropriate, or would a differentiated standard targeted at different types
of firms be better? This is the key issue we are seeking to analyse.
There are examples of differentiated standards, for instance standards specifically developed for small firms
which are different to those used by large firms. But these are often developed several years after the core
standard has appeared and then not been taken up by small firms. An example is ISO/IEC 29110 This relies on
existing standards, such as ISO/IEC 12207 (software engineering life cycle standard). The motivation for this
was that small firms were not willing to use large standards, such as 12207, due to their complexity and lack of
support (Varkoi, 2010). Varkoi also notes that several countries translate the standards and supporting material
to better serve their local industry. Standards often require skilled workers to implement and many countries
have a shortage of skilled workers and again this may dictate the need for differentiated standards.
The issue is important, as all standards, including quality ones, offer benefits to firms, regions and countries.
This can happen in a number of ways. Firstly, quality standards may actually improve efficiency, leading to cost
savings and potential quality enhancements Whether or not standards such as ISO9000 actually do lead to such
efficiency gains is the subject of some dispute in the literature. But even if they do not, there are still potential
gains to the firm in signalling quality. This is consistent with a literature, e.g. on advertising, that the mere
expenditure of a substantial amount of money on signalling quality is something only a high quality firm would
do (Milgrom and Roberts, 1986). Thirdly, the adoption of a standard may be necessary to gain entry to certain
markets. Finally the adoption of certain types of standards may be useful in improving relationships with the
2
local community, shareholders, workers (if linked with safety) and other stakeholders. Hence differential access
to standards may disadvantage some firms.
Our analysis will be based on firms in both manufacturing and services, as well as other sectors such as
construction which do not fit neatly in either category. As far as we are aware, our study is the only large scale
analysis of firms’ use of standards across a substantial number of countries. Previously too, most studies have
focused on manufacturing (Pekovic, 2010). The countries we have chosen are ‘transition ones’ in Europe and
Asia. Some are well advanced in the transition process and indeed are members of the EU, others are less well
advanced. The regression results will show systematic differences in standard certification by characteristics
such as firm size, location and ownership, as well as between sectors and countries. This suggests the need for
greater efforts by ISO and individual countries in facilitating standard certification amongst certain types of
firms. In addition, the fact that take up is substantially less amongst some firms than others suggests there is a
case for differentiated standards. Thus it may not be optimal to develop single homogenous, generic standards
for all firms, i.e. one size may not fit all.
The paper proceeds as follows. In the next section, we discuss the ISO 9000 and ISO 14000 families of
standards. There then follows a discussion of the factors which both theory and the literature suggest impact on
the firm’s standard certification decision. Section 4 presents the data and section 5 the regression results. Finally
we conclude the paper.
2. The ISO Family of Standards
The International Organization for Standardization (ISO) was established in 1947, at a time when several
worldwide organisations were being established including the IMF and the World Bank. Its remit, in tune with
the spirit of the times, was to develop worldwide standards to facilitate international trade and improve
international communication and collaboration. ISO now produces many standards, indeed at the end of 2011
over 19,000, e.g. ISO 22000 and ISO/IEC 27001 which relate to food safety and information security
respectively. However, probably its two most prominent products are the ISO 9000 and ISO 14000 families of
standards. There have been many studies done on the impact of these on various aspects of a firm’s
performance. There has also been a number of studies examining the diffusion of these standards at the level of
a country. There has however been less done examining the reasons why individual firms seek such
certification.
3
In line with ISO’s initial remit, international standards may serve to reduce information costs and, thereby,
increase trade (Swann et al., 1996). Moreover, international standard certification is sometimes a regulatory
requirement, particularly when trading in the EU. It has also been argued that international standards can reduce
any competitive disadvantage associated with other sources of reputation, such as country of origin. A World
Bank (1999:288) report states: “To expand their trade, countries also need good standards, measurement,
testing, and quality control systems. These constitute the infrastructure for technical activity, and their
significance grows as traded products and services increasingly have to conform to world standards and
regulation. If consumers cannot readily distinguish between products or services produced by different firms,
poor quality by one producer in a market can damage all others...........Obtaining certification for meeting quality
standards is especially important for countries with a reputation for poor products ” (our emphasis). This same
report highlighted the use of ISO 9000 as an international standard. In this respect standards may be particularly
relevant for the developing and transition countries in our analysis.
2.1 ISO 9000
ISO 9000 can trace its ancestry back to military procurement standards around the Second World War.
Eventually these led to BS 5750, the first commercial quality management standard published by the British
Standards Institute in 1979. This was adopted, with some changes, as ISO 9000 in 1987. There was a minor
update to the standard in 1994, a more major one in 2000, and a further one still in 2008, By the end of
December 2009, at least 1,064,785 ISO 9001 (2000 and 2008) certificates had been issued in 178 countries and
economies. In 2009, the growth rate over 2008 was 8 %1. The ISO 9000 family of standards represents an
international consensus on good management practice. It is a ‘generic management standard’. Theoretically at
least, the same standard can be applied to any organisation, large or small, whatever its product or sector of
activity, regardless of whether it is a business enterprise, a public administration or a government department. It
is primarily concerned with quality management, i.e. concerning what the organisation does to ensure that its
products conform to customers’ requirements. While it is not a guarantee of quality per se, ISO certification
means that an independent auditor has checked the process influencing quality. According to Beck and
Walgenbach (2005) the introduction of a quality management system based on ISO 9000 increases
management’s ability to control processes through their formalization.
1
Figures obtained from http://www.iso.org/iso/survey2009.pdf
4
ISO 9000 facilitates global trade (Peach 1995). Its use, particularly in Europe, has been promoted by the EU’s
incorporation of ISO 9000 as part of the Communities’ Global Approach to Testing and Certification which
revised policy on standards as a component of the EC 1992 economic integration initiative. ISO 9000
certification is generally given by an independent registrar and lasts for three years. Registrars are generally
themselves accredited by national standards authorities. But there are perceived differences in the criteria used
in different countries In addition in some countries there is often a conflict of interest with registrars acting as
consultants. Partly, because of this there is no general acceptance of ISO 9000 certification made in one country
by another and this is the subject of multilateral or bilateral agreements (Hudson and Jones, 2003).
The arguments that have been put forward for firms adopting ISO 9000 relate to quality effectiveness,
increased efficiency (Gotzamani and Tsiotras, 2002) and increased operational control (Bhuiyan and Alam,
2005). The process of certification addresses the issue of setting and implementing a management system that
produces consistent levels of quality (Voehl et al., 1994). Implementation results in improved communication
between employees, cost savings, reduced paper work, greater competitive advantage, more organized design
and output and access to global markets (Peach, 1995). Tzelepis et al. (2006), on the basis of an econometric
analysis of Greek firms, conclude that ISO 9001 increases productive efficiency. The World Bank (1999, Box
2.1) cites evidence that amongst 93 major Brazilian enterprises surveyed in 1994, 55 percent increased
productivity as a result of ISO 9000 certification, 35 per cent improved the standardization of processes, 31
percent increased employee participation in quality control and more than 20 percent reported an increase in
employee satisfaction. Many others have found benefits in terms of product quality, customer satisfaction, cost
reduction, productivity and operating advantages (Gotzamani and Tsiotras, 2002). Reducing informational
asymmetries is often cited as an advantage. Nicolau and Sellers (2010) find that ISO 9000 has a particularly
high impact on the market value variation of hotel chains, which they attribute to a reduction in information
asymmetry, Terlaak and King (2006) conclude that ISO 9000 certification facilities faster growth in US firms,
which they link to a reduction in informational asymmetries in supply chains. Caro and Garcia (2009) find that
ISO 9000 certification improves consumers' perceptions with regard to quality, satisfaction, and corporate
image. In terms of the characteristics of firms who obtain certification, Lafuente et al. (2010) find ownership
structure, especially whether a multinational firm is the largest shareholder, to be important. This is consistent
with the benefits listed above. Prakash et al. (2006) found that manufacturing firms tended to get more from
these benefits in efficiency gains than service firms, with there being little difference between the two in terms
of the costs of obtaining ISO9000 registration. Others, have gone further in questioning the extent of these
5
benefits (Quazi et al., 2002). In addition too there is the potential for increased bureaucracy (Boiral, 2003) which
may have an adverse impact on efficiency.
2.2 ISO 14000
ISO 14000 built on the success of ISO 9000, and is again a ‘generic management standard’ applicable to any
organisation. In many respects ISO 14000 is similar to ISO 9000, but signals environmental quality rather than
quality per se. Standardising multiple national environmental management system (EMS) standards is of
specific benefit to companies operating across several countries, who had previously been faced with many,
potentially incompatible, systems (Chan and Wong, 2006). By the end of December 2009, 223,149 ISO
14001:2004 certificates had been issued in 159 countries and economies. The reasons for obtaining ISO 9000
certification are similar to those relating to ISO14000 certification. But there are also differences. Thus it has
been suggested that ISO 14001 certification is a way of enhancing relationships with communities,
environmentally conscious investors and consumer groups. (Stenzel, 2000). Poksinska et al. (2003), with respect
to Sweden, conclude that benefits partially stem from marketing advantages and improved relations with
stakeholders and partially from greater awareness of environmental considerations within the firm. Hemenway
and Hale (1995) argue that ISO14000 can benefit companies through reduced energy costs and by improving
the company’s image with the general public and other stakeholders. Chan and Wong (2006), in the context of
Hong Kong, argue that the factors leading hotels to adopt certification include corporate governance,
particularly when part of a group, and the need to be seen to comply with legislation. Curkovic et al. (2005)
concludes firm size, the proportion of sales going to consumers, rather than up the supply chain, foreign
ownership and export focus are important in determining ISO14000 certification. Finally, Massoud et al. (2010)
argue that small to medium-size enterprises (SMEs) may face serious constraints in setting up and maintaining
an EMS.
3. Why Firms Adopt International Standards
We focus on the case where certification is not a statutory, or near statutory, requirement to enter certain
markets. Firms will seek certification if it is in their financial interests to do so, i.e. if the increase in revenue
exceeds any net increase in costs (ΔC). The change in net costs relates to the costs of certification (CS) less any
savings, possibly negative, the firm may make as a consequence of having better quality control techniques in
place:
6
ΔC = CS + (dC/dE)ΔE
(1)
dC/dE represents the reduction (or increase) in costs as a consequence of increased efficiency and ΔE the
change in efficiency as a consequence of standard certification. The increase in revenue comes about through
the impact of increased efficiency on the ability to raise price or increase quantity sold. Thus these gains are
likely to be greatest for large volume producers or high price producers, or some combination of the two. The
impact of certification on efficiency is potentially twofold. Firstly, there is the possibility that actual product
quality may rise. In this case, the impact of quality certification is likely to be greater the more complex the
organisation. Complexity can be expected to increase with size, be greater for group firms, may differ between
industries and may also be greater in manufacturing, where raw materials need to be brought in and final
products shipped out, than in services. Secondly there is the possibility that quality is more effectively signalled
to customers and potential customers.
We illustrate the latter point now using a Bayesian adjustment approach, as in Hudson and Jones (2001). The
expected, which we assume is the perceived, level of quality of the firm’s product given it adopts the standard is
q* = qSσ2/(σ2 + σ2S) + qσ2S/(σ2 + σ2S)
(2)
where qS is the level of quality signalled by the standard, q that perceived without standard certification.
Perceived quality with standard certification (q*) is a weighted average of q and q S. The weights relate to σ2S,
the variance of the standard quality signal, and σ2, the variance of perceptions without the standard. The increase
in perceived quality (Δq*) will be a function of the gap between qS and q, and also of σ2S relative to σ2:
Δq* = (qS – q)σ2/(σ2 + σ2S)
(3)
The confidence that people have in their expectations when the firm is standard certified is inversely related to
the variance of those perceptions (2*) which equals σ2σ2S/(σ2 + σ2S). Thus the gain in confidence, or decline in
the variance, following standard certification is:
7
Δ2* = σ2[σ2/(σ2 + σ2S)]
(4)
The smaller is σ2S relative to σ2, the greater the gain in confidence. Hence, the standard certified firm gains from
(i) signalling higher quality and (ii) reduced uncertainty, and both can impact positively on revenue. To the
extent that perceptions of quality have become associated with the income per capita of the country of origin,
developing and transition countries find it especially difficult to signal quality with respect to consumer goods
generally (Jones and Hudson, 1996). Many of the other signals such as brand name and word of mouth relate to
established products and are also unavailable to them. Hence for firms in these countries who are trying to
compete in export markets, international quality standards may be of particular importance.
The signalling argument is relevant to firms higher up the supply chain as well as consumers. This is
particularly the case if standard certification reduces the need for purchasing firms to undertake their own
quality control checks on supplier firms (Tirole, 1988)2. Hence, Anderson et al. (1999) suggest, and provide
empirical support for, the hypothesis that firms supplying to other firms are particularly likely to face demands
for ISO certification. However, firms which already have a reputation for high quality may not need to register
(Kreps and Wilson, 1982).
There are possible reasons for adopting quality certification which do not fit neatly into an immediate impact
on profits. These include, for ISO 14000 in particular, concerns over the health and safety of the workforce, a
better relationship with shareholders and a better relationship with local people and local government. The latter
may facilitate planning applications. In addition, firms in industries regulated by the EU are more likely to
demand from their suppliers ISO 9000 certification (Anderson et al., 1999) and also ISO 14000 certification.
Following on from this discussion, and also building on the literature review, we will include the following
explanatory variables in our analysis of firms who have standard certification:
Domestic, or export, focus: ISO certification is least likely amongst firms who trade locally. It is easier for
customers, consumers and other firms to observe the firm if it is local. In addition firms that export abroad may
also be exposed to new ideas and for this reason may learn about and adopt management standards (Montiel and
Husted, 2009). Apart from exporting firms, this is also relevant for firms trading within their country, but
outside their region. Reflecting this we will also include factors which facilitate communication.
2
For example. when three US car manufacturers adopted ISO 9000 procedures it replaced three separate programs that
required suppliers to submit different sets of documentation and allow periodic audits by the car firms. ISO 9000 registration
then became the only supplier requirement.
8
Cost Pressures: We will also include information on cost pressures as suggested by Pekovic (2010). As we shall
see, this will supplement our analysis of the impact of domestic focus.
Managerial experience: To the extent that ISO certification increases knowledge on best practices, the need for
certification will decrease with managerial experience. Experienced managers have already acquired significant
knowledge and may feel, perhaps not always correctly, there is less to gain from ISO certification. However, it
needs to be emphasized that managers with long years of experience are also successful managers. Some
younger managers may eventually prove unsuccessful and exit the industry.
Size: ISO certification is particularly likely for large firms whose certification costs can be spread over a large
revenue stream and who potentially have more to gain from standardising complex administrative structures
(Dobbin and Sutton, 1998). Smaller firms too, face problems in accessing the expertise necessary to implement
the relevant procedures. In larger firms this is often found ‘in house’ (Grolleau et al, 2007).
Ownership: ISO certification is more likely for foreign and group firms, than other firms. They have access to
greater resources, greater knowledge (Gourlay and Pentecost, 2002) and perhaps face greater internal pressure to
seek certification (Pekovic, 2010). Particularly for group firms, with a potentially complex management
structure spread over several sites, it may be a way of ensuring managerial efficiency in the face of what is in
effect a principal agent problem.
Location: ISO certification may be more likely in larger cities, which are likely to have certification agents and
also where the effects of other firms’ certification can be more readily observed.
If significant, domestic focus and managerial experience would be consistent with a standard which does not
adversely differentiate between firms, as these reflect differential benefits between firms. But some of the
remaining variables, if significant, would suggest that some firms are disadvantaged in terms of accessing
standards. For example, if it is found that the probability of standard certification is low for small firms, it would
suggest that the high costs of certification relative to sales may be a deterrent factor and that small firms would
benefit from a simpler, less costly standard. Similar arguments can be made if, e.g., there are systematic
differences in take up between the service and manufacturing sectors or between countries. In all these case,
there would then be a basis for us to question whether a single generic standard is really optimal.
4. Data
9
We analyse the data from the World Bank’s 2009 Enterprise Surveys 3, of a sample of firms in Central and
Eastern Europe and Asia. The sample was chosen as it represents a range of transition countries, some well
established in international markets and others much less so. It is based on 11,668 firms in the countries shown
in Figure 1. The surveys are targeted at establishments with five or more full time employees. Geographical
distribution is defined to reflect the distribution of non-agricultural economic activity and for most countries this
implies including the major urban centres. This data base is being increasingly used by, e.g., Gorodnichenko et
al. (2010) with respect to innovation and Beck et al. (2004) in analysing access to credit across a range of
countries, including transition ones.
Insert Figure 1 about here
The specific variable we will analyse is based on the responses to the question “Does this establishment have
an internationally-recognized quality certification?”. The interviewers were instructed that “If there is need for
clarification, some examples are: ISO 9000, 9002 or 14000”. The raw data is shown in Figure 1 and shows
substantial differences between countries. There is a tendency, which is consistent with expectations, for
standard certification to be high in EU countries or aspiring ones such as Turkey. But there are some EU
countries such as Poland and the Baltic states, where certification is quite low. The question clearly leans
towards an interpretation that we are considering adoption of ISO standards and in particular the ISO
9000/14000 sets of standards. But is this interpretation justified? Figure 2 shows the proportion of firms in the
survey who have adopted a standard plotted against the ratio of the total number of ISO 9000 and ISO 14000
series standards to Gross National Income (GNI, the unit being $100m) in the different countries4. There is a
close relationship which justifies us in assuming that to a considerable extent the standards referred to in the
survey are ISO standards.
Insert Figure 2 about here
3
This is also known as the Business Environment and Enterprise Performance Survey (BEEPS). More information is
available and the data accessible at http://www.enterprisesurveys.org/.
4
The ISO number of standards relate to the end of 2008 (Source: ISO, http://www.iso.org/iso/survey2008.pdf). The GNI
figure is for 2009 and has been converted to international dollars using purchasing power parity rates (Source World
Bank). GNI is being used here as a proxy for the number of firms in the country.
10
Table 1 shows the correlations between standard certification and firm characteristics. Certification is
positively correlated with being a large firm, a city location, managerial experience, being in the manufacturing
sector, foreign and group firms and using the Internet to communicate to clients. It is negatively correlated with
being a small, or a young, firm, located in rural areas or small towns, being in the service sector and the extent
of focus on the domestic market. The average figures emphasize just how common standard certification is
amongst foreign, and in particular group, firms. There is no evidence of the expected impact of managerial
experience. Instead certification seems to be highest for the most experienced of managers. The domestic focus
variable is more in line with expectations. Firms with a 100% domestic focus appear least likely to have
standard certification, but the highest certification is for firms with some, but not a total, export focus5. The data
relating to sectors shows certification is lowest in the service sectors and tends to be more common in the
manufacturing sectors. Hotels and restaurants, and also firms in the retail sector, are characterized by
particularly low certification. This is consistent with a signalling hypothesis, as the consumer can directly see
and experience the quality of these establishments, and for hotels there is often an alternative, * based, system of
quality rating.
Insert Table 1 about here
5. Regression results
Insert Table 2 about here
Table 2 shows the regression results. Because the dependent variable is binomial, the technique of probit, within
STATA, was used. We have used the robust or sandwich estimator of the standard errors 6. This estimator is
robust to some types of misspecification so long as the observations are independent. The pseudo R 2s are as
reported by STATA and are quite high for this type of analysis. The results in 2.1 show that the probability of
5
Export focus is, of course, a mirror image of domestic focus.
6
An alternative is to use a cluster-robust estimator. But with a small number of clusters, e.g. less than 50, or very unbalanced
cluster sizes, this can present problems (Nichols and Schaffer, 2007). In our case the first criteria mitigates against using
countries as the base for the cluster and the second precludes the use of regions.
11
standard certification increases with the age of the firm, is less for small firms compared to medium sized ones,
and for the latter compared to large firms. Similarly, firms in rural areas are less likely to have standard
certification than firms in medium sized towns, which in turn are less likely to be certified than those in larger
cities. Both foreign and group firms are more likely to be standard certified. Finally we note that firms which
use the Internet to connect with clients are much more likely to use standards. This could be because it
represents a progressive and entrepreneurial attitude. To check this we re-estimated the regression, as shown in
2.2, replacing individual Internet firm usage with that of the average usage for the firms in the region 7. It
remains significant with the same sign, indicating that the quality of the Internet connection does indeed impact
upon standard adoption. One interpretation of this is that the Internet facilitates the acquisition of information
including that related to standards. But we will delay discussion of this until having considered further
regressions. In none of the regressions was the quality of regional transport, derived in a similar way to regional
Internet usage, significant. There are also significant differences between countries, and as suggested by Figure
1, these are substantial. They are not shown in the regressions, but in regression 2.1, firms of a given set of
characteristics were least likely to be standard certified in Albania, Uzbekistan and the Ukraine, and most likely
in the Czech Republic, Hungary and Romania. The combined significance of these variables suggests significant
differences in standard certification in a manner which may plausibly be linked with differences in the costs of
accessing or the benefits of acquiring the standard and which do not simply reflect differences in managerial
experience or a need to signal quality to distant markets. Instead they reflect differences arising from the way
the standard has been designed and the way the net benefits are systematically greater for some firms than
others.
Turning to the other variables, with respect to company form, partnerships and sole proprietors are less likely
to use standards than other firms. Standard certification declines with managerial experience, but at a decreasing
rate8. It begins to turn up again with managerial experience of about 26 years. Most firms have managerial
experience less than that, but the turning up of the curve may be real, rather than an artefact of using a quadratic
form, as suggested in Table 1. There are several possible explanations as to why this could be so, including a
7
When calculating this average for the i’th firm, we excluded the i’th firm from the calculation. It thus represents average
email usage for all other firms in the same region as the i’th one.
8
This apparently contradicts the positive correlation reported in Table 1, but this is multiple regression where we are
considering the impact of managerial experience given all the other variables.
12
loosening of control amongst older managers, which they seek to compensate for by implementing quality
certification. Clearly, more research is needed on this.
In 2.3 we introduce two variables reflecting cost pressures, as suggested by Pekovic (2010). The results
indicate that if this pressure comes from domestic competitors, firms are less likely to adopt standards, but if it
comes from foreign firms then they are then more likely to do so. They do not respond to pressure from
customers. This suggests that domestic competition is focused on price, and international competition more on
quality. Regression 2.4 adds to this the domestic focus of the firm, which is very significant. The curve linking
the probability of standard certification with domestic focus is an inverted U shaped one. The maximum
probability of being standard certified is for firms with some export focus, but not a total one. The raw data
supports this conclusion as can be seen from Table 1. Consequently we replaced the two domestic focus
variables with two dummy variables, the first operative if the firm had domestic focus less than 25% and the
second greater than 75%. Both were negatively significant at the 1% level of significance in an additional
regression not reported in the Table9. The coefficients were -0.34 and -0.43 respectively, confirming that the
greatest probability of certification is for firms with some export focus, rather than a dominant one.
These regressions may be affected by problems of endogeneity. Firms who become standard certified may
then be more likely to have a significant export focus and as a consequence face cost pressures from foreign
firms. We therefore instrumented these variables within a two stage estimation process. The instruments are the
regional and industry (at the SIC 2 digit level of classification) averages of (i) firms’ domestic focus and (ii)
their views on the impact of foreign and domestic competition on their costs. For the i’th firm these averages
relate to all other firms in the i’th firm’s region or sector. There were 150 regions. These instruments are all
significant in the three regressions at the 1% level and hence they appear suitably strong instruments. The
predicted values from the reduced form regressions were then included in the regression as shown in column
2.5. Both domestic focus and foreign competition were significant at the 1% level, but not domestic competition
and this was dropped from the regression. A nonlinear impact of domestic focus is still present. The signs of the
other variables are largely unchanged, but with reduced significance 10.
9
10
The fit of the regression was not as good as that in 2.4
The locational variables are insignificant, but are significant in all the reduced form equations. Hence their impact is
through the instrumented variables.
13
The final two regressions repeat regression (2.4), but just on firms in the construction and service sectors (2.6)
and on manufacturing firms (2.7). The results are different. The two managerial experience variables were only
significant in the manufacturing regression. However, a single dummy variable operative for firms where the
manager had relatively little experience 11 was positively significant in the services and construction equation,
and this is the variable shown in the regression. This once more suggests that low managerial experience is
associated with a greater probability of certification, but for services there is no evidence that firms with high
managerial experience also see a high probability of certification. Neither were the spatial variables relating to
rural and town based location significant in this regression. However, the regional Internet usage variable was
very significant. In the regression relating to the manufacturing sector, the significance was the other way round.
This tentatively suggests these two sets of variables are correlated and indeed if we omit the Internet variable
from the services sector regression then the locational variables become significant 12. The impact of the
domestic competition variable is also only significant in the manufacturing equation, with pressure from
customers on costs leading to a decline in standard certification in the services’ regression. In the services’
sector over 95% of sales are for the domestic market, whilst in the manufacturing sector the figure was only
80%. Thus this once more suggests that domestic cost pressures, whether from firms or customers, reduce the
likelihood of standard certification.
6. Conclusions and policy Implications
The results confirm that there are substantial differences in standard certification between firms of different
characteristics. In part these reflect the differing value of quality standards to, e.g. exporting and non-exporting
firms. But this is not always the case. Firms outside large cities, and particularly in rural areas, appear
disadvantaged in terms of accessing ISO certification. The raw data from Table 1 suggests the former are about
60% more likely to be standard certified than other firms and large firms are almost four times more likely than
small firms. The regression results in table 2 confirm that this is not simply because of other differences in
characteristics. These are large differences and to the extent that it is because there are differential transaction
11
That is 5 years or less.
12
If we do the reverse for the manufacturing regression, then the sign on electronic communication becomes positive as
expected, although it fails to be significant at the 5% level.
14
costs in acquiring ISO certification, it is an issue which should be of concern to the ISO 13 and national and
regional governments. If ISO 9000 and also ISO 14000 improve managerial efficiency as the literature suggests
may be the case, then it may be to the firms’ disadvantage, and to the country’s, that they are not standard
certified to a similar extent as larger firms.
These differences in quality standard certification both within and between countries support the hypothesis
that the same generic standard is not optimally suited to all firms, all sectors and all countries. Thus the
argument can be made that standards such as ISO 9000 and ISO 14000 should be defined differently for firms in
manufacturing as opposed to those in service sectors or for small as opposed to large firms. Even if there is no
such case on technical grounds, the fact that there are different take-up rates may in itself justify different
standards. The optimal standard is a trade-off between certification costs and benefits. Larger firms, for
example, will want more ambitious standards than smaller ones. The advantage of a single, uniform, generic
standard is that it strengthens the signal and increases the degree of interoperability. Having two different
standards, e.g. one for large and one for smaller firms, might lead to some confusion and might weaken the
signal. But this advantage comes at a cost, particularly if the gap between what different firms want is large. In
the case of the standards analysed in this paper, this may have contributed to the uneven take up across firm
size, location, organizational structures and sectors. This is possibly also the case for other standards, e.g.
technical ones as published by the IEC. These again are often the outcome of compromise in a consensus
seeking process, with different countries favouring different standards on geographical and climate
considerations, as well as their skill base. In this case what is optimal for the richer countries of the temperate
north may not be so suited to the needs of predominantly poorer countries in the south. This then sometimes
leads to standards being adapted by different countries. Differentiated standards may therefore actually lead to
less heterogeneity, with perhaps two variants being sufficient to satisfy all countries, without need for further
adaptation. The whole issue of the optimal number of standard variants is one deserving of further analysis.
There are other policies open to increase standard certification. Brown et al. (1998) suggest that sharing
expert time with other SMEs and/or involving students who are following quality programs are potential
solutions to the problem of low registration amongst SMEs. This, and greater use of the Internet in the
13
There is the possibility that firms in rural areas have a greater local focus than those in large towns and cities. However,
the latter are still more likely to have standard certification when restricted to manufacturing, than firms in smaller towns
and rural areas. This suggests a transaction costs explanation. Although to be definitive on this issue we need information
on the full costs of ISO certification in differing locations.
15
certification process, may also help firms outside the large towns and cities. From a different perspective,
Hudson and Jones (2003) have argued that an effective form of aid to developing and transition countries would
be to subsidize ISO 9000 certification. There is an even stronger case for subsidizing ISO 14000 certification
due to the public good element of improved environmental performance.
More generally, our results are of importance in helping us to understand firms' growth and to formulate
policies to facilities this, especially in the tradeables sector. Economic integration appears to be an effective way
of speeding up economic growth. For example, the enlargement in market potential is capable of explaining
between 15% and 40%, of the economic growth of the new member countries after joining the EU (Clemente et
al., 2009). Clemente et al. also emphasise that peripheral countries face a greater challenge in order to
appropriate such benefits, and again international quality standards can facilitate their integration. The benefits
of free trade are well known, if under some attack in recent years following the economic crisis (Salvatore,
2009). Perhaps the greatest benefits in the immediate future, for both the EU and globally, lie in opening up the
services sector. The EU has been trying to do this for some time (European Commission, 2011). De Bruijn et al
(2008) estimate that the 2004 Services Directive could increase the volume of trade in ‘other commercial
services’ by as much as 62%. However, simply reducing the barriers to trade in services will be only partially
successful if not accompanied by a means for unknown, often small, firms to both signal and achieve adequate
quality. Quality standards such as ISO9000 can facilitate this, but to do so, as we emphasised earlier, a
differentiated standard may be needed targeted more at both the service sector and small firms, who are
particularly important in this sector. There are still large gains still to be made in manufacturing. Hassan et al
(2010) show that progressive reductions in trade restrictions in Bangladesh have played a role in improving
manufacturing firms’ total factor productivity growth. They emphasise that in order to maximize domestic
growth opportunities, domestic capacity building should be encouraged, highlighting education and the labour
market. But quality standards can again potentially play a role in both improving and signalling quality. The
opening up trade, in part stimulated by international quality management standards, may also have spin off
effects on other firms who are forced to compete against new firms from outside their region and even their
country14.
Finally, these results to an extent confirm previous research, although seldom has this research been on as
large a group of firms and countries as we have used. But our analysis has also added to previous research in
14
Our results, particularly with respect to the impact of cost pressures, lend support to this possibility.
16
several ways. In addition, to the qualification of the impact of an export or domestic focus, the results with
respect to managerial experience and spatial factors are new to the literature. Also new to the literature is the
impact of Internet usage. This was a regional, rather than a firm specific, variable and suggests that the quality
of the Internet infrastructure can have an impact on a firm acquiring standard certification. There are two
possibilities, firstly the Internet facilitates knowledge acquisition and reduces the transaction costs involved with
standard certification. Secondly it facilitates access to long distance markets where signalling quality is more
important than in local markets. Finally, with respect to the impact of domestic focus, it is worth noting that it is
a significant factor despite the inclusion of a variable reflecting competitive pressures from foreign firms. Hence
the significance of domestic focus is due to reasons other than simply competing against foreign firms.
However, the impact is not unambiguous, firms with a very large export market are less likely to have standard
certification than firms with smaller, but still sizeable, export sales as a proportion of their total market. It may
be that the former are well established in export markets and have less need for standard certification to signal
quality. This is consistent with the argument by Kreps and Wilson (1982) that firms which already have a
reputation for high quality in a market may not need to acquire one by means such as standard certification.
References
Anderson, S.W., Daly, J D. & Johnson, M.F. (1999). Which firms seek ISO 9000 certification: Regulatory
compliance or competitive advantage?, Production and Operations Management, 8, 28-43.
Beck, N. & Walgenbach, P. (2005). Technical efficiency or adaptation to institutionalized expectations?, The
adoption of ISO 9000 standards in the German mechanical engineering industry. Organization Studies, 26, 841866.
Beck, T., Demirguc-Kunt, A. & Maksimovic, V. (2004). Bank competition and access to finance: International
evidence, Journal of Money Credit and Banking, 36, 627-648
Bhuiyan, N. & Alam, N. (2005). An investigation into issues related to the latest version of ISO 9000, Total
Quality Management and Business Excellence, 16, 199-213.
Boiral, O. (2003). ISO 9000, outside the iron cage, Organization Science, 14, 720-737.
Brown, A., van der Wiele, T. & Loughton, K. (1998). Smaller enterprises’ experiences with ISO 9000,
International Journal of Quality Reliability Management, 15, 273-285.
17
Caro, L.M. & Garcia, J.A.M. (2009). Does ISO 9000 certification affect consumer perceptions of the
service provider?, Managing Service Quality, 19, 140-161.
Chan, E.S.W. & Wong, S.C.K. (2006). Motivations for ISO 14001 in the hotel industry, Tourism Management,
27, 481-492.
Clemente, J., Pueyo, F. & Sanz, F. (2008). Market potential, European Union and growth, Journal of Policy
Modeling, 31, 719-730.
Curkovic, S., Sroufe, R. & Melnyk, S. (2005). Identifying the factors which affect the decision to attain ISO
14000, Energy, 30, 1387-1407.
De Bruijn, R. Kox, H. & Lejour, A. (2008). Economic benefits of an integrated European market for services,
Journal of Policy Modeling, 30, 301-319.
Dobbin, F. & Sutton, J.R. (1998). The strength of a weak state: The rights revolution and the rise of human
resources management divisions, American Journal of Sociology, 104, 441-476.
European Commission (2011) A Strategic Vision for European Standards: Moving Forward to Enhance and
Accelerate the Sustainable Growth of the European Economy by 2020, Brussels, European Commission.
Gorodnichenko, Y., Svejnar, J. & Terrell, K. (2010). Globalization and innovation in emerging markets,
American Economic Journal: Macroeconomics, 2, 194-226.
Gotzamani, K.D. & Tsiotras, G.D. (2002). The true motives behind ISO 9000 certification, The International
Journal of Quality Reliability Management, 19, 151-169.
Gourlay, A. & Pentecost, E. (2002). The determinants of technology diffusion: Evidence from the UK financial
sector, Manchester School, 70, 185-203.
Grolleau, G., Mzoughi, N. & Pekovic, S. (2007). Chemical firms’ registration for the responsible care program
and the ISO 14001 standard: A comparative approach, Economic Bulletin, 12, 1-13.
Hassan, M.K., Isik, I. & Mamun, A. (2010). Trade liberalisation and industry performance in Bangladesh,
Journal of Policy Modeling, 32, 399-417.
Hemenway, C. G. & Hale, G. J. (1995). Are you ready for ISO 14000, Quality, 34, 26–28.
Hudson, J. & Jones, P. (2001). Measuring the efficiency of stochastic signals of product quality, Information
Economics and Policy, 13, 35-49.
Hudson, J. & Jones, P. (2003). International trade in ‘quality goods’: Signalling problems for developing
countries, Journal of International Development, 15, 999-1013.
18
Jones, P. & Hudson, J. (1996). Standardization and the costs of assessing quality, European Journal of Political
Economy, 12, 355-361.
Kreps, D. & Wilson, R. (1982). Reputation and imperfect competition, Journal of Economic Theory, 27, 253279.
Lafuente, E. Bayo-Moriones, A. & Garcia-Cestona, M. (2010). ISO-9000 certification and ownership structure:
Effects upon firm performance, British Journal of Management, 21, 649-665
Massoud, M.A., Fayad, R, El-Fadel, M. & Kamleh, R. (2010). Drivers, barriers and incentives to implementing
environmental management systems in the food industry: A case of Lebanon, Journal of Cleaner Production,
18, 200-209
Milgrom, P. & Roberts, J. (1986). Price and advertising signals of product quality. Journal of Political
Economy, 94, 796-821.
Montiel, I. B. & Husted. B. (2009). The adoption of certified environmental management
programs in Mexico: First movers as institutional entrepreneurs, Journal of Business Ethics, 88, 349-363.
Nichols, A. & Schaffer, M. (2007). Clustered errors in Stata, United Kingdom Stata Users' Group Meetings,
http://www.stata.com/meeting/13uk/nichols_crse.pdf
Nicolau, J.L. & Sellers, R. (2010). The quality of quality awards: Diminishing information asymmetries in a
hotel chain. Journal of Business Research, 63, 832-839.
Peach, R.W. (1995). The ISO 9000 handbook, second ed. Irwin, Homewood, IL.
Pekovic, S. (2010). The Determinants of ISO 9000 certification: A comparison of the manufacturing and service
sectors, Journal of Economic Issues, 44, 895-914.
Poksinska, B., Dahlgaard, J., Jörgen, J. & Eklund, A.E. (2003). Implementing ISO 14000 in Sweden: Motives,
benefits and comparisons with ISO 9000, International Journal of Quality Reliability Management, 20, 585606.
Prakash J. Singh, M.F. & Smith, A. (2006). ISO 9000 series of standards: comparison of manufacturing and
service organisations, International Journal of Quality Reliability Management, 23, 122-142.
Quazi, H.A., Hong, C.W. & Meng, C.T. (2002). Impact of ISO 9000 certification on quality management
practices: A comparative study, Total Quality Management, 13, 53-67.
Salvatore, D. (2009). The challenges to the liberal trading system, Journal of Policy Modeling, 31, 593-9.
Stenzel, P. L. (2000). Can the ISO 14000 series environmental management standards provide a viable
alternative to government regulation?, American Business Law Journal, 37, 237-299.
19
Swann, P., Temple, P. & Shurmer, M. (1996). Standards and trade performance: The UK experience, Economic
Journal, 106, 1297-1313.
Terlaak, A. & King, A.A. (2006). The effect of certification with the ISO 9000 Quality Management Standard:
A signaling approach, Journal of Economic Behaviour and Organization, 60, 579-602.
Tirole, J. (1988). The Theory of Industrial Organization. Cambridge, Mass, MIT Press.
Tzelepis, D., Tsekouras, K., Skuras, D. & Dimara, E. (2006). The effects of ISO 9001 on firms' productive
efficiency, International Journal of Operations Production Management, 26, 1146-1165.
Varkoi, T. (2010). Process assessment in very small entities: An ISO/IEC 29110 based method, Paper presented
at 2010 Seventh International Conference on the Quality of Information and Communications Technology
Voehl, F., Jackson, P. & Ashton, D. (1994). ISO 9000: An Implementation Guide for Small to Mid-Sized
Businesses, Delray Beach, FL., St. Lucie Press.
World Bank (1999). World Development Report: Knowledge for Development. Oxford, UK, Oxford University
Press.
Appendix: Definition of variables:
Endogenous variables
Standard
Binary variable, coded 1 if firm has, or is acquiring, an internationally-recognized
quality certification (such as ISO 9000, 9002 or 14000).
Exogenous variables (binary, unless otherwise stated)
Age:
The length of time, in years, the establishment has been operating in the country.
Small firm
Coded 1 if the number of full time employees is less than 20.
Medium sized firm Coded 1 if the number of full time employees is between 20 and 100,
Sole proprietor
Coded 1 if the firm has the legal status of a sole proprietor.
Partnership
Coded 1 if the firm is a partnership
Private Ltd Co
Coded 1 if the firm is a private limited company
Publicly listed Co. Coded 1 if the firm is a publicly listed company
Group
Coded 1 if the firm is part of a larger firm.
Foreign
Coded 1 if the share of the company held by foreign individuals or companies > 49%.
Manager’s
The number of years of experience working in the sector for the firm’s top manager.
experience
Rural
Coded 1 if the firm’s location has less than 50,000 people
Town
Coded 1 if the firm’s location has between 50,000 and 1 million people (and is not a
capital city)
Cost Pressures
Three variables relating to the pressures on costs from domestic competitors,
foreign competitors and customers respectively, coded 1 if these were ‘fairly’ or ‘very’
important
Domestic focus
The % of the firm's sales which were national sales and is of course inversely related to
export focus.
Electronic
Binary variable, coded 1 if the firm used the Internet to communicate with clients
communication
or suppliers. (The regional average of this is used as an indicator of regional Internet
infrastructure),
Transport
The extent to which transport presented an obstacle to the firm, responses ranged from no
obstacle (coded 1) to very severe obstacle (coded 5).
20
For the last two variables regional averages were calculated. For firm i they relate to the average response in
the region for all firms other than firm i. In addition there are industry/sector and country variables as defined
in the tables.
Table 1: Summary statistics of standard usage
Correlation Average (%)
Correlation Average (%)
All firms
28.1
Small firm
-0.27**
13.0
Electronic communication
0.23**
33.8
Medium firm
0.018*
29.6
Foreign firm
0.15**
50.9
Large firm
0.28**
49.8
Group firm
0.13**
64.1
Rural
-0.03**
26.0
Service & construction
-0.15**
22.0
Town
-0.06**
24.4
Manufacturing
0.15**
35.4
City
0.08**
40.1
Manager's experience
0.063**
Young
-0.06**
20.9
Experience <5 years
-0.00398
26.7
Domestic focus
-0.20**
Experience >=5 & <25 years
-0.0382**
26.7
D. Focus =100%
-0.29**
20.1
Experience >25 years
0.036**
31.5
D. Focus>=25% & <75%
0.18**
56.1
Retail
-0.160**
16.1
D. Focus<25%
0.10**
45.0
Hotels & restaurants
-0.045**
17.6
Transport sector
-0.004
27.4
Garments (Manufacturing)
-0.050**
19.2
Construction sector
0.023*
31.5
Wholesale
-0.014
25.9
Food sector
0.056** 36.6
Other services
-0.010
25.6
Textiles sector
0.040** 38.3
Other manufacturing
0.140**
38.7
Notes: **/* denotes significance at the 1% and 5% levels respectively The average shows the proportion of
firms with standard certification in that group. The correlation relates to that between the variable and standard
certification.
Table 2: Regression results
2.1
All firms
Electronic
Communication
Regional
Transport
Manager's
experience
Manager's
experience2
Log firm age
Small firm
Medium
Sized Firm
Rural
Town
Foreign
Group
Partnership
Private Ltd Co.
0.406**
(8.96)
-0.0434
(0.38)
-0.0167**
(3.87)
0.032**
(3.50)
0.0616**
(2.75)
-0.937**
(22.69)
-0.4566**
(12.86)
-0.1678**
(3.93)
-0.1156**
(3.20)
0.3145**
(5.90)
0.3156**
(6.86)
-0.2251*
(2.18)
0.0465
2.2
2.3
All firms All firms
0.185**
(2.67)
-0.0522
(0.46)
-0.0162**
(3.79)
0.0302**
(3.33)
0.0639**
(2.87)
-1.004**
(24.75)
-0.4877**
(13.81)
-0.1741**
(3.87)
-0.1132**
(2.95)
0.3359**
(6.26)
0.3202**
(6.93)
-0.2313*
(2.25)
0.0386
0.1908**
(2.72)
-0.0678
(0.59)
-0.0162**
(3.78)
0.0301**
(3.31)
0.0602**
(2.70)
-0.9843**
(24.15)
-0.4731**
(13.29)
-0.1674**
(3.71)
-0.1048**
(2.72)
0.3089**
(5.72)
0.3191**
(6.91)
-0.2311*
(2.24)
0.0421
2.4
2.5
All firms All firms
0.1899**
(2.72)
-0.0916
(0.80)
-0.0168**
(3.88)
0.0314**
(3.42)
0.0446*
(1.98)
-0.9264**
(22.27)
-0.4444**
(12.32)
-0.1691**
(3.70)
-0.1047**
(2.70)
0.3019**
(5.50)
0.3297**
(7.11)
-0.2361*
(2.28)
0.0246
2.6
Services &
construction
0.2881** 0.4203**
(3.97)
(4.33)
-0.3338** 0.0359
(2.85)
(0.23)
-0.0181** 0.1517*
(4.14)
(2.40)a
0.0344**
(3.64)
0.0025
0.0192
(0.10)
(0.58)
-0.8106** -0.7967**
(13.65)
(14.16)
-0.4056** -0.3807**
(8.90)
(7.29)
0.0104
-0.0864
(0.18)
(1.41)
0.0725
-0.046
(1.51)
(0.82)
0.1715** 0.3546**
(2.60)
(4.49)
0.2698** 0.3746**
(5.82)
(5.93)
-0.2341* -0.3272*
(2.24)
(2.29)
0.0617
-0.1026
2.7
Manufacturing
-0.082
(0.77)
-0.2539
(1.52)
-0.0224**
(3.66)
0.0438**
(3.43)
0.0562
(1.81)
-1.101**
(17.07)
-0.502**
(9.76)
-0.2642**
(3.72)
-0.1588**
(2.85)
0.2528**
(3.30)
0.2917**
(4.21)
-0.125
(0.84)
0.1893*
21
Publicly
Listed Co.
Sole
proprietor
Domestic
focus
Domestic
focus2
Cost Pressures
Domestic
competition
Foreign
competition
Customers
(0.80)
0.0989
(1.44)
-0.1855**
(2.59)
(0.67)
0.0754
(1.11)
-0.2271**
(3.19)
(0.73)
0.074
(1.09)
-0.2262**
(3.16)
(0.42)
0.0505
(0.73)
-0.2161**
(3.01)
0.0274**
(10.20)
-0.0273**
(11.74)
-0.0762*
(2.28)
0.2694**
(8.34)
-0.053
(1.54)
-0.0607
(1.78)
0.2082**
(6.24)
-0.0456
(1.32)
(1.05)
0.0246
(0.36)
-0.2535**
(3.53)
0.0406**
(3.62)
-0.0018**
(2.78)
(1.27)
-0.0901
(0.93)
-0.3112**
(3.17)
0.0262**
(4.87)
-0.0266**
(6.01)
(2.25)
0.2271*
(2.27)
-0.0913
(0.85)
0.0267**
(8.31)
-0.0261**
(9.20)
1.119**
(8.49)
-1.070**
(8.14)
0.002
(0.04)
0.1997**
(4.16)
-0.1379**
(2.84)
-0.1294**
(2.70)
0.2138**
(4.53)
0.0422
(0.85)
Sectors
Food
0.0307
0.0166
0.048
0.0994*
0.2544**
0.0863
(0.63)
(0.34)
(0.99)
(2.02)
(4.40)
(1.67)
Textiles
-0.4451** -0.4534** -0.4778** -0.5327** -0.4632**
-0.5659**
(5.39)
(5.51)
(5.76)
(6.27)
(4.69)
(6.39)
Garments
-0.5686** -0.584** -0.6026** -0.5666** -0.5872**
-0.5839**
(7.84)
(8.14)
(8.34)
(7.55)
(7.43)
(7.46)
Construction
-0.2087** -0.2258** -0.1726** -0.0509
0.2405** -0.1599
(4.03)
(4.39)
(3.31)
(0.95)
(2.91)
(1.03)
Other
-0.2104* -0.22**
-0.1917* -0.0914
-0.0352
-0.2077
services
(2.46)
(2.58)
(2.22)
(1.05)
(0.35)
(1.22)
Wholesale
-0.3958** -0.3814** -0.3644** -0.2923** -0.3111** -0.4243**
(6.93)
(6.67)
(6.37)
(5.02)
(4.43)
(2.70)
Retail
-0.6066** -0.6203** -0.591** -0.4828** -0.4031** -0.6067**
(14.42)
(14.83)
(14.05)
(11.07)
(6.14)
(3.99)
Hotels &
-0.5111** -0.5434** -0.4964** -0.4141** -0.0065
-0.5387**
restaurants
(5.96)
(6.42)
(5.86)
(4.89)
(0.06)
(3.19)
Transport
-0.3823** -0.3895** -0.3557** -0.3353** -0.0122
-0.4218**
(5.61)
(5.72)
(5.20)
(4.84)
(0.15)
(2.60)
Observations
10998
10996
10996
10996
10995
6031
4965
Log Likelihood
-5209
-5248
-5214
-5132
-5197
-2579
-2500
X2
1966
1965
2021
2178
2640
976
1091
Pseudo R2
0.201
0.195
0.2
0.212
0.202
0.187
0.224
% correct
77.80%
77.60%
77.70%
78.20%
78.10%
80.80%
75.10%
Notes: Regressions estimated by probit; (.) denotes t statistics, */** significance at the 5% and 1% levels
respectively. Standard errors have been corrected for heteroscedasticty. In 2.5 foreign competition and domestic
focus have been instrumented. Variables defined in data appendix, country fixed effects included. In 2.1
electronic communication relates to whether the firm uses this, in all other regressions it is the regional average
as described in the data appendix. All regional variables are logged. Χ2 represents the likelihood ratio test
statistic. The coefficients on the two squared terms have been multiplied by 100 for presentational purposes. %
correct, denotes the % of correct predictions. aIn 2.6 the managerial experience variable is a dummy one,
operative for firms with top managers having five years or less experience.
22
Density
Figure 1: Firms with an international standard
Albania
Belarus
Georgia
Tajikistan
Turkey
Ukraine
Uzbekistan
Russia
Poland
Romania
Serbia
Kazakhstan
Moldova
Bosnia & Herz
Azerbaijan
Fyr Macedonia
Armenia
Kyrgyz Republic
Mongolia
Estonia
Kosovo
Czech Republic
Hungary
Latvia
Lithuania
Slovak Republic
Slovenia
Bulgaria
Croatia
Montenegro
Explanation
Histogram shows the proportions without a standard and then with
.5
Figure 2: The relationship between the survey standards and ISO standards
.4
Turkey
.3
R2 of fitted line = 0.56
.1
.2
Bulgaria
0
2
4
6
ISO ratio to GNI
Survey standards
Fitted values
The survey proportion is the proportion of firms in the survey with a standard;
The ISO ratio is the ratio of the number of ISO9000/14000 standards to GNI
23