Factors affecting the incidence and intensity of standards certification evidence from exporting firms in Pakistan

Abstract

This article explores the incidence and intensity of certification of Pakistani exporting firms across a range of quality assurance programs. Using firm-level data, the firm's certification status is modelled using Heckman's two-step procedure. The first-stage results using a probit model show that the likelihood of certification is determined by the sector, the firm's awareness of trade standards, the level to which the firm's markets are diversified, external pressure for certification and the firm's primary export market. The intensity of certification is treated as the number of standards a firm has invested in, which can be seen as ordered into distinct categories, such that an ordered probit model can be applied, correcting for sample selection bias. The results show that the number of standards to which a firm is certified is influenced by firm size, a firm's primary export markets, level of awareness of trade standards and level of market diversification. Overall, the results suggest that certification can be enhanced through increased private and institutional incentives, such as reducing transactions costs in the fisheries and agro-processing sectors, as well as increased regulatory enforcements in supply chains and support for raising the technical capacity of the sector.

Acknowledgements

The work of the Pakistan Institute of Development Economics, Islamabad in the data collection and preliminary analysis and of UNIDO in providing access to the data is acknowledged. We would like to thank an anonymous reviewer and the editor of the journal for their valuable suggestions and comments that considerably improved this article.

Notes

¹ The EU and US are Pakistan's largest trading partners accounting for 25 and 20% of Pakistan's exports, respectively. Other major importers include China, Dubai and Saudi Arabia. Major exports are in garment made-ups and bed wear, each of which are valued at US$1.35 billion. Knitwear, readymade garments and cotton yarn also have important export shares (World Bank, 2006, p. 32).

² Since the UK absorbs almost 25% of Pakistan's shrimp exports, while Germany and Japan account for 15%, the importance of food safety management systems cannot be emphasized too strongly (World Bank, 2006).

³ In recent years Pakistan has exported to the EU fish value at about US$45 million per year. The first 4 months of the self imposed suspension, affected US$40 million worth of trade, either being lost or redirected to less remunerative markets (World Bank, 2006, p. 143).

⁴ For analytical convenience, we abstract from inter-temporal issues by assuming that benefits, expected benefits, and costs are discounted back to the period in which the firm must make the certification decision.

⁵ As one reviewer pointed out, Heckman's two-step procedure has been observed to be biased with small samples. Nevertheless, we are of the view that concerns related to sample-selection bias are more concerning than those related to bias arising from a small sample.

⁶ where n is the number of export markets served by firm and S_i is the firm's share of total sales in market i.

⁷ The level of awareness is reflected in the following: firms are more aware of ISO 900 (mean score 4.1), followed by ISO 14000 (mean score 3.2), HACCP (mean score 2.6), traceability (mean score 2.6) SA 8000 (mean score 2.6) OHSAS (mean score 2.1) and EUREPGAP (mean score 2.0).

⁸ ‘Don’t know’ responses were excluded from the analysis. Using ordered mean scores, the most important factors for certification across all standards in textiles and leather were to meet international market requirements, improving sales and firm reputation. In agro-food processing, certification is largely pursued to enhance the reputation of the firm and product quality, while in the fisheries sector certification serves to improve firm reputation and access to better markets.

⁹ Across all standards and in the agro-food processing, fisheries and textiles sectors, certification cost was, on average, the key deterrent to certification. However, there are variations in perceptions across standards, in particular the importance of cost as an impediment to ISO 9000 certification is higher in the agro-processing and fisheries sectors than in the textiles and leather sectors. In agro-food processing, cost and system upgrades were the key factors. In the fisheries sector, the consultancy costs and set up costs were key factors.

¹⁰ Note that estimating the outcome equation requires for identification purposes that at least one variable appears in the likelihood of selection equation (certification/no certification) but not in the outcome equation (intensity of certification) (Sartori, 2003). The exclusion of these variables is justified because we believe that they have little effect on the intensity of certification. Even when included in the outcome equation, these variables are not significant.

¹¹ The World Bank (2006, p. 35–6) shows that, in the textile sector, export processes/administration constitute 4.9% of export costs for smaller firms compared to 1.9% for larger firms.

¹² For example, in the textile sector the cost of initial ISO 9000 certification has dropped from US$ 8300 in 1990 to US$ 2120 in 2004 and of SA 8000 from US$ 8300 in 2001 to US$ 3720 in 2005. Likewise in the leather sector, initial ISO 9000 certification costs have declined from US$ 8330 in 1997 to US$ 5900 in 2004.