EXPORTS AND EFFICIENCY PERFORMANCE OF AUSTRALIAN MANUFACTURING SMALL AND MEDIUM ENTERPRISES (SMES): EVIDENCE FROM BUSINESS LONGITUDINAL DATABASE (BLD) 2005 2007 Dr Viet Le Lecturer in Entrepreneurship, Faculty of Business and Enterprise, Swinburne University of Technology, Melbourne, Australia Postal address: Mail H23, PO Box 218, Hawthorn, Victoria 3122, AUSTRALIA E-mail address: [email protected]

SUMMARY This paper examines efficiency performance of exporting and non-exporting Australian manufacturing SMEs using Business Longitudinal Database released by the Australian Bureau of Statistics for the period 2005-2007. Results from an analysis of 844 manufacturing SMEs find that Australian manufacturing SMEs achieved a high technical efficiency level, especially the exporting SMEs at more than 80 percent. The non-exporting SMEs tended to have lower efficiency level compared to their exporting peers. The results also show an increase in technical efficiency level for both exporting and non-exporting manufacturing SMEs during the period examined. There is evidence of self-selection of more efficient firms into exporting from the increase in exports participation and efficiency level as well as better performance of exporting SMEs. The paper calls for a focus on improving SME efficiency performance as a way to increase their participation in exports and not only the traditional instruments. Keywords: Small and Medium Enterprises (SMEs), Efficiency performance, Manufacturing Firms, Exports, Australia.

Paper prepared for presentation at the International Council for Small Business (ICSB) World Conference, Wellington, New Zealand, 10-13 June 2012.

INTRODUCTION Australian small and medium enterprises (SMEs)1 consistently account for over 99% of the total number of businesses in the past decade. At June 2009, there were 2,044,736 SMEs out of a total of 2,051,085 businesses in operation in Australia, representing a share of 99.7% (ABS, 2010b). The latest figure shows that 45,581 Australian businesses exported goods and services in 2009 (ABS, 2010b). This is a significant growth from about 25,000 exporting businesses in 2001. However, the 2009 figure fell short of the target of the Government to have 50,000 Australian exporters by 2006. Of the total exporters, 43,259 were exporters of merchandise goods and 3,422 were exporters of services (ABS, 2010b). Among the goods exporting businesses in 2008/09, there were 37,327 SMEs representing 83 percent of all goods exporting businesses (ABS, 2010b). In the service sector it was reported that the export participation rate among Australian service firms was only 1.2 percent despite being the second largest service economy in AsiaPacific (Patterson 2004). Sensis Business Index reported that only 17 percent of Australian SMEs exported in 2007 (Sensis,2007). This is the highest level since 2004. It represents a strong growth from only 4 percent reported in 2000 (ABS, 2000). Sensis Business Index reported that 13% of Australian SMEs exported in 2010 (Sensis, 2011). Data from the Business Longitudinal Database (BLD) also indicated a low level of participation in exports by Australian SMEs. According to figure from the BLD, only 11 percent of SMEs were exporters in 2006-2007 (Hansell and Talgaswatta, 2009). The level of export participation in Australian SMEs is thus low compared to OECD countries or European countries (Austrade, 2001; European Commission, 2004). Hansell and Talgaswatta (2009) note previous studies have found that exporting firms and nonexporting firms have different characteristics including productivity. They observe that most studies are from America and Europe with a focus on large businesses. According to them a few Australian studies were available on exporting at the firm level, and those confined to small and medium-sized enterprises (SMEs) are even fewer. Despite the importance of export to Australian economy as outlined in a Austrade’s paper (Austrade, 2000), there are few studies about SMEs exports. The limited number studies about export of Australian SMEs held the low “intention to export” to ‘tyranny of distance’, protectionism until 1980s, and the type of goods exported (mineral and rural commodities) 1

Defined as businesses with less than 200 employees.

1

(Austrade, 2001; Gittins, 20 April 2002; OECD, 1997). All of these factors have been improved greatly due to the increasingly globalised economy and application of internet and e-commerce in the last two decades. Most of the studies about export performance of Australian SMEs look at specific types of firm (such as family business or service firms) and issues (cultural factors, psychic distance, or managerial characteristics) (Brewer, 2007; Gierczycki and Reid, 1998; Graves and Thomas, 2006). A more recent paper used a probit model to analyse traditional factors such as firm age, firm size, foreign ownership, innovation, ICT use and labour productivity, and the probability of exports for Australian SMEs (Hansell and Talgaswatta, 2009). However, this study did not examine the performance of exporting and non-exporting SMEs. Other paper that evaluated the performance of Australian small firms based on BLD focuses on their innovation activities (Gronum and Verreynne, 2011). In the Australian Economic Indicators (ABS, 2010a), the statistics showed that SME exporters have higher total sales, sales per employee, and value-added; pay higher average wages; and possess more assets than SME non-exporters. The SME exporters have also been in operation under current ownership longer and employ more staff than SME non-exporters. This paper has the aim to examine technical efficiency performance of Australian manufacturing SMEs. It has the following objectives: (i) evaluating technical efficiency performance of Australian manufacturing SMEs; (ii) comparing efficiency performance of exporting and nonexporting Australian manufacturing SMEs; and (iii) providing evidence-based recommendations for enterprise managers and policymakers. This research will be the first to examine the technical efficiency of Australian SMEs using the BLD. It analyses two categories of SMEs which are exporting and non-exporting. The paper thus focuses on an important sector of businesses in the Australian economy namely manufacturing sector and understanding their competitiveness in the global market. The research is significant because of several reasons. First, an evaluation of the efficiency level of Australian SMEs from a rich longitudinal dataset will offer a clear picture about performance and competitiveness of these enterprises. Second, an investigation of the relationship between technical efficiency and export could shed lights to understand the low participation of Australian SMEs in export activities. Third, empirical evidence from the research will help develop recommendations regarding this important sector of businesses in Australia.

2

LITERATURE REVIEW Exporting is widely believed to have a positive impact on the performance of firms due to selfselection of more efficient firms into exporting and learning-by-exporting (Arnold and Hussinger, 2005; Delgado, Farinas, and Ruano, 2002; OECD and APEC, 2007; Wagner, 2007b). While the support for the self-selection hypothesis is confirmed in dozens of studies (Greenaway and Kneller, 2005; Wagner, 2007b), the evidence for the learning-by-exporting hypothesis is mixed (Fernandes and Isgut, 2005a). One of the key studies on the relative productivity of exporters and non-exporters and on causality between exporting and productivity was conducted by Wagner (2007a). Wagner gave a synthesis of findings from 54 empirical studies published between 1995 and 2006 on the effects of international trade on productivity of SMEs. The author investigated the relationship of exports and productivity of SMEs by using firm-level data from 34 countries. Among the countries covered are highly industrialised countries (such as the United States, the United Kingdom, Canada and Germany); countries from Latin America (such as Chile, Colombia and Mexico); Asian countries (China, Korea, Indonesia and Taiwan); transition countries (Estonia and Slovenia); and least developed countries from sub-Saharan Africa. Wagner pointed out exporters are more productive than non-exporting firms of the same size from the same narrowly defined industry. According to findings from the literature on exports and productivity, the author emphasised an important reason for the positive productivity differential between exporters and non-exporters is self-selection of more productive plants on export markets. Furthermore, there is evidence for a market driven selection process in which exporters that have low productivity fail as successful exporters, while only those that are more productive continue to export (Wagner, 2007b). However, some recent studies have indicated that by following an export-oriented strategy, firm may also experience some issues that can have negative impacts on its performance. These can include over-dependency on volatile foreign customers and markets, lack of understanding about overseas markets, and higher transaction costs in export activities (Le, 2010; Mok, Yeung, Han, and Li, 2010). Exporting firms have to bear extra costs, for example market research, adaptation of products to local regulations or transport costs. These extra costs are one reason for a selfselection of the more productive firms on international markets (Wagner, 2011a). In addition, it could take time for the learning from exporting to take place and there could be a time lag before the benefit of this is reflected on efficiency performance. 3

Productivity is defined as the efficiency, with which firms turn inputs (labour, physical capital, energy, materials, managerial know-how) into outputs (goods, services) (Wagner, 2011a, p.4). Productivity is important for the competiveness of firms, regions and countries on local, national and international markets. Productivity is also an important factor of growth and welfare. Therefore, there have been many empirical studies that investigated the relationship between productivity and firms’ international activities (Amendolagine, Capolupo, and Petragallo, 2011; Arnold and Hussinger, 2005; Aw, 2002; Aw and Huang, 1995; Delgado, Farinas, and Ruano, 2008; Fernandes and Isgut, 2005b; Fryges and Wagner, 2008; Grazzi, 2011; Mok et al., 2010; Oh, 2011; Powell and Wagner, 2011; Redding, 2011; Vogel and Wagner, 2011; Wagner, 2007b, 2011a). Table 1 below provides a summary about the above studies.

4

Table 1: Empirical Studies about Exports and Performance of Small and Medium Enterprises (SMEs) No.

Authors

Estimation

Country

period

Topics

Methodology/

Independent/Dependent

investigated

Econometric

Variables

Major Finding

Technique 1

Amendolagine, Capolupo

1995-2003

Italy

Exports and performance in manufacturing firms

Regression analysis; , propensity score matching

N.A

1992-2000

Germany

Exports and performance in manufacturing firms

A two-factor logarithmic CobbDouglas production function Ordinary least squares Matched sample

TFP Export intensity No. employees Sales Innovator R&D expenditure Share of sale (new product) Wage Firm age Sales Employment Book value of the capital stock Expenditures on labour and different types of intermediate inputs Firm age

and

Petragallo (2011)

2

Arnold

and

Hussinger (2005)

3

Aw (2002)

1981-1991

Taiwan

The link between firm size, growth and productivity

Regression analysis , Regression analysis; Ordinary least squares

4

Aw and Hwang

1986

Taiwan

Exports and performance in Taiwanese electronics industry

the Chow test The translog production function Cobb-Douglas production function Cross section

Labour Capital services Value-added

1991-1996

Spanish

Exports and

Descriptive

N.A

(1995)

5

5

Delgado,

Export entrants improve their productivity in the first period after entry although this effect vanishes in the subsequent period. This occurs for both total factor productivity (TFP) and labour productivity growth rates. New exporters earn higher profits than their domestic counterparts do. The average total factor productivity (TFP) growth of exporters is slightly slower than for non-exporting firms. Thus, there are no significant productivity differences between reporting and nonexporting firms. There are no statistically productivity gains from exporting.

1. Firms with higher initial levels of productivity are more likely to survive and grow in size 2. The growth in productivity in several of Taiwan’s major industries is positively related to firm size, suggesting that in these industries micro firms that survive and grow bigger have higher growth in TFP. There are significant differences in productivity levels between exporters and non-exporters. The contribution of these differences to output differences between the two groups of producers varies 3-20%, depending on the electronic product and model specification. Firms that eventually enter the export

Farinas

and

1981-1991

Colombia

Exports and profitability in manufacturing enterprises

Descriptive analysis; Regression analysis; Ordinary least squares

Output Labour Intermediate inputs Skill intensity Wage premium Capital Vintage Production experience Plant age Export experience

1995-2005

Germany

Exports and profitability in manufacturing enterprises

Descriptive analysis; Regression analysis; generalized propensity score methodology The KolmogorovSmirnov statistics

Number of employees Wage per employee Labour productivity Firm size

1989-2004

Italy

Trade and profitability

Size (Total sales) Productivity (Labour & TFP) Gross Margins

2002

China

Exports and technical efficiency in manufacturing enterprises

Descriptive analysis; nonparametric comparison of distributions; regression analysis Regression analysis Non-parametric data envelopment analysis Stochastic frontier analysis

market had higher than 7% of the median productivity than non-exporters in the period prior to their entry There are no significant differences between the productivity growth distribution of entering exporters and the distribution of continuing exporters in the period after entry and similarly for entering exporters and non-exporters. There is strong evidence of learning-byexporting for young Colombian manufacturing plants. Total factor productivity increases 4-5%for each additional year a plant has exported, after controlling for the effect of current exports on total factor productivity. Learning-by-exporting is more important for young than for old plants and in industries that deliver a larger percentage of their exports to high-income countries. There is a causal effect of firms’ export activities on labour productivity growth. At an export-sales ratio of 19%, a firm’s export activities have a causal effect 3% increase in the firm’s labour productivity growth rate. However, exporting improves labour productivity growth only within a sub-interval of the range of firms’ exportsales ratios. No evidence for profitability differential between exporters and non-exporters over all; positive relation for some sectors, negative for others.

Number of employees Net value of fixed assets Technical efficiency Exports Firm size Tangibility

Clothing firms with a high degree of sales in the domestic market or with a high level of export orientation experience a higher level of technical efficiency than those firms trying to conquer both the local and the overseas markets

Ruano (2008)

6

Fernandes

and

Isgut (2005b)

7

Fryges

and

Wagner (2008)

8

Grazzi (2011)

9

Mok (2010)

6

et

al.

performance in manufacturing firms

analysis; Regression analysis; The KolmogorovSmirnov statistics Nonparametric estimation

10

Oh (2011)

11

Powell

and

1993-2003

Korea

Productivity growth, efficiency change and technical change

N.A

Germany

Exports and productivity In manufacturing industries

Wagner (2011)

12

Redding (2011)

N.A

N.A

Reviews the recent theoretical literature on heterogeneous firms and trade

13

Vogel

2003-2005

Germany

Exports and profitability in business service enterprises

and

Wagner (2011)

14

Wagner (2007c)

N.A

N.A

Exports and profitability in enterprises

15

Wagner (2011b)

N.A

N.A

Exports and profitability in enterprises

7

Descriptive analysis; Regression analysis A second-stage regression analysis. Descriptive analysis; New unconditional quantile estimation technique Quantile estimation Ordinary least squares The Melitz Model Integrated Equilibrium Trade and Market Size Gravity Descriptive analysis; Regression analysis; generalized propensity score methodology Survey of the Evidence from Firm-level Data A synthesis and evaluation of the literature

Capital/labour ratio TFP Sales Capital input Labour/Number of worker Total fixed assets TFP Total sales per employee

N.A

Productivity (turnover per employee) Firm size (number of employed persons) Export status (dummy) N.A

N.A

A competitive market condition, R&D activities, export activities and innovativeness increased the rate of productivity growth Neither low productivity exporters nor high productivity non-exporters are a rare species There is a rather large difference in productivity at the bottom of the distribution. The exporters are more productive throughout the entire distribution With positive fixed exporting costs and for sufficiently large values of fixed and variable trade costs, only some firms export. These exporting firms are larger and more productive than firms that only serve the domestic market Services exporters are less profitable compared to non‐exporters, though difference is small. Evidence for self‐selection of less profitable services firms into exports. No positive causal effect of exports on profits. Exporters are more productive than nonexporting firms of the same size from the same narrowly defined industry Exporting firms are more productive than otherwise identical firms that sell on the national market only.

METHODS Productivity and efficiency represents the economic aspects of firm performance. Growth in efficiency and productivity is the most important aspect of growth as it focuses on the quality of growth. For this reason theoretical and empirical works on firm performance focus on measuring enterprise productivity and efficiency (Storey, 1990). This study focuses on technical efficiency which refers to the ratio between actual output and the maximum output the firm could produce with the set of inputs and technology it uses (Coelli, Rao, and Battese, 2005). According to Farrell (1957), efficiency measure contains an efficient production frontier which is the output that a perfectly efficient firm could obtain from any given combination of inputs. The performance of a productive unit will be measured against that efficient frontier (Farrell, 1957). If there is a gap between the efficient frontier and the actual performance of the firm that firm is inefficient. A model to measure of technical efficiency in the ith firm can be defined as: TE

Yi Yi *

(1)

where: Yi: Actual output Y*i: Maximum possible output The above equation is the basic model used for measuring technical efficiency according to Kalirajan and Shand (1999:152). The actual output is observable in this equation. However, maximum possible output is not observable and must be estimated. A ratio of one in the above equation would mean that the firm is technically efficient and operates on the production frontier. A number of techniques have been developed to estimate this frontier. Several authors broadly classified them into two main groups: parametric and non-parametric (Coelli et al., 2005; Kalirajan and Shand, 1999; Kumbhakar and Lovell, 2003; Murillo-Zamorano, 2004). The parametric method uses an econometric technique by specifying a stochastic production function which assumes that the error term is composed of two elements. One 8

is the typical statistical noise which represents randomness. The other represents technical efficiency which is commonly assumed in the literature to follow a one-sided distribution (Alvarez and Crespi, 2003; Murillo-Zamorano, 2004). In the context of this study the stochastic frontier production function approach is most relevant. The first reason is the ability of the stochastic frontier approach to consider both factors beyond the control of the firm and firm-specific factors, and hence it is closer to reality. The second reason is the separation of the random variation of the frontier across firms, the effects of measurement error and other random shocks from the effect of inefficiency. The stochastic frontier production model was developed independently and simultaneously by Aigner, Lovell and Schmidt (ALS) (1977), Meeusen and Van den Broeck (MB) (1977), and Battese and Corra (1977). In this model there is a composed error term which captures the effects of exogenous shocks beyond the control of the analysed units in addition to incorporating technical inefficiency. Errors in measurement of outputs and observations are also taken into consideration in this model (Kumbhakar and Lovell, 2003; Murillo-Zamorano, 2004). The generalised functional form in the Cobb-Douglas case of the stochastic production function can be specified as:

Yi

xi

(Vi U i )

,

i = 1, …,N,

(2)

where

Yi xi

is the production (or the logarithm of production) of the i-th firm; is a k

1 vector of (or transformation of) the input quantities of the i-th

firm; is a vector of unknown parameters;

9

Vi

are random variables which are assumed to be independently and identically distributed (iid) as N(0,

Ui

2 v

),2

which are non-negative random variables that are assumed to account for technical inefficiency in production and are often assumed to be iid. N (0,

2 u

)

. It is assumed to be half-normal, exponential and truncated from

below at zero.3 However, the Translog model has been widely accepted as a relatively flexible functional form because it is in second order log-linear form. It does not impose assumptions about constant elasticities of production nor elasticities of substitution between inputs. It thus allows the data to indicate the actual curvature of the function, rather than imposing a priori assumptions (Pascoe, Kirkley, Gréboval, and Morrison-Paul, 2003). This functional form allows variability in the elasticity of substitution among factors of production and is more flexible in permitting substitution effects among inputs. Thus, it is claimed to be a relatively dependable approximation to reality (Giulkey, Lovell, and Sickles, 1983 as cited by Sun, 2006:9). However, it has some drawbacks, including a potentially high correlation between cross-term variables due to the fact that more parameters have to be estimated. In addition, if a sufficient data series is not available, it could result in degrees of freedom problems (Pascoe et al., 2003). However, the drawbacks will be expected to be minimum when the sample is large and the correlation among variables is not too high, which is the case for the data used in this study. In their original forms, the Cobb-Douglas function and the Translog function are not linear. By transforming the variables into logarithms, both models will become linear and can be estimated in a linear regression framework (Coelli et al., 2005:212). The translog specification can be expressed as follows:

2

This means that the errors are independently and identically distributed normal random variables with zero means and variances σ2. 3 Ui reflects one-sided deviations of actual output from the maximum level of production due to technical inefficiency. If a firm is fully technically efficient, Ui=0, otherwise it will be greater than zero. Thus, it is also called a one-sided error component.

10

ln Yi

0

1

ln Ki

2

2

4

(ln Ki )

7

ln Ki ln Li

5

ln Li

(ln Li ) 8

3

ln M i

2 6

(ln M i ) 2

ln K i ln M i

9

ln Li ln M i Vi U i

Where: Yi = Output of firm i Ki = Capital input of firm i Li = Labour input of firm i Mi = Intermediate input for firm i Vi = Random error in which vi

N(0,

2

Ui = Technical Inefficiency in which ui

v)

N( i,

2

u)

This empirical research will analyse Business Longitudinal Database (BLD) released by the Australian Bureau of Statistics in 2009. The dataset includes a sample of 2,732 SMEs covering three periods 2004-2005, 2005-2006, and 2006-2007. The data contains information about basic financial indicators from Business Activity Statement (BAS). It identifies firms that are exporters of goods and services and those who are non-exporters. This allows an analysis of the difference in technical efficiency performance between exporters and non-exporters. According to the statistics from ABS (ABS, 2010b), SMEs represented 86% of goods exporters by number and contributed less than 5 percent of the total value of goods exports. The Wholesale trade and Manufacturing industries had the highest value of the SME goods exports in 2008-09 (ABS, 2010b). Table 2 shows the number of employing SMEs from the 2006-07 BLD according to industry and whether or not they are exporters. At the aggregate level 11.3 percent of businesses in the sample are exporters. The three sectors with highest share of exporters are Information Media and Telecommunications (27.5%), Wholesale Trade (26.5%), and 11

Manufacturing (24.4%). Among these sectors Manufacturing has the largest number of observations in the sample and for these reasons the paper will focus on manufacturing sector. Data from the BLD also revealed that almost half of the manufacturing SMEs in the sample involved in exports in 2005. Export participation rate increased to 53 percent in 2006 and 59 percent in 2007 (Table 3). Table 2 - Number and Proportion of 2006 - 07 BLD Units by Export Status Non-Exporting SMEs Exporting SMEs Total No. % No. % No. Industry Division (ANZSIC 2006) Agriculture, Forestry and Fisheries Mining Manufacturing Construction Wholesale Trade Retail Trade Accommodation and Food Services Transport, Postal and Warehousing Information Media and Telecommunications Rental, Hiring and Real Estate Services Professional, Scientific and Technical Services Administrative and Support Services Arts and Recreation Services Other Services Total

1179 184 659 331 405 265 375 377 145 174 252 198 221 331 5096

95.4 88.9 75.6 96.2 73.5 89.5 98.4 97.4 72.5 97.2 85.7 91.7 93.3 95.4 88.7

57 23 213 13 146 31 6 10 55 5 42 18 16 16 651

4.6 11.1 24.4 3.8 26.5 10.5 1.6 2.6 27.5 2.8 14.3 8.3 6.8 4.6 11.3

1236 207 872 344 551 296 381 387 200 179 294 216 237 347 5747

Source: Hansell and Talgaswatta (2009)

Table 3: Exporting and Non-exporting Manufacturing SMEs from BLD Year

Exporting SMEs

Non-Exporting SMEs

Total

Firms

% of Total

Firms

% of Total

2005

127

49%

134

51%

261

2006

149

53%

133

47%

282

2007

179

59%

122

41%

301

Source: Author’s calculation from BLD

The study uses input-output variables from the BAS in the BLD including: (i) total sales; (ii) capital purchase; (iii) non-capital purchase; and (iv) total salary, wages and other payments. These variables represent output (Y), capital (K), material and energy (M), and labour (L), respectively. Export sales data from the BAS was used to distinguish between 12

exporting and non-exporting firms. There are a total of 844 observations in the survey sample of the BLD in 2005, 2006 and 2007 eligible for analysis. A cross-sectional analysis was carried out for exporting and non-exporting firms for the three periods and the results were presented and discussed in the next section.

13

RESULTS AND DISCUSSIONS Maximum likelihood estimation (MLE) was carried out using the frontier model in STATA for the BLD in the Remote Access Data Laboratory (RADL) with access granted by ABS. The restrictive nature of RADL made it difficult to attempt estimations with other software packages written for efficiency analysis with more flexibility such as FRONTIER 4.1. Results from estimations were shown in Tables 4, 5 and 6 below. Since the first order coefficients estimated in the Translog stochastic production function are not informative and do not convey direct economic meaning (Hill and Kalirajan, 1993) they are not discussed here.4 The MLE also provides estimates of the variance parameters sigma-squared (σ2). It determines whether there is technical inefficiency or not. If σ2 is equal to zero, all firms are fully efficient. If σ2 is larger than zero, then all firms are not fully efficient. The values of σ2 are 0.07 in 2005, 0.05 in 2006 and 0.12 in 2007, indicating that all firms in the sample are not fully efficient. As shown in Tables 4, 5 and 6, the mean technical efficiency for exporting SMEs in the manufacturing sector is estimated at 84 percent and 88 percent, in 2005 and 2006, respectively. Thus, there was an increase of mean technical efficiency for these firms during the period studied. These results indicate they can increase their output by 16 percent in 2005 and by 12 percent in 2006 with the same level of input. Estimation results for exporting SMEs were not available under MLE for the 2007 data because a discontinuous region was encountered. RADL does not allow access to raw data so it is not possible to understand the reason for this. There is evidence of self-selection of more efficient firms into exporting as the increase in technical efficiency level over the examined period also corresponds with the increase of export participation as shown in Table 3. Average technical efficiency level of non-exporting firms is 60 percent and 69 percent in 2005 and 2006, respectively. Efficiency level found for Australian non-exporting manufacturing SMEs is comparable to the range reported reported for manufacturing firms in other countries at 60 to 70 percent (Tybout, 2000). However, SMEs in the

4

To understand how much output will increase when an input increases, while the other inputs are held constant elasticities of substitution among the inputs are often calculated. However, this is not the focus of the paper.

14

manufacturing sector appear to have achieved full technical efficiency in 2007 at least among the non-exporting SMEs. Similar to the exporting SMEs, there was a rise in technical efficiency of non-exporting SMEs in the 2005-2007 period. Compared with technical efficiency level of exporting SMEs, the non-exporting firms achieved a much lower technical efficiency level, at least in the first two periods when a comparison is possible. The difference in technical efficiency between exporting and non-exporting SMEs is quite significant at 14 and 19 percent in 2005 and 2007. Although it is not possible to conclude about the learning-by-exporting hypothesis, there is additional evidence for the self-selection of more efficient firms into exporting. Table 4: Estimation Results for Manufacturing SMEs 2006 2005

K (Captial) M (Non-capital) L (labour)

Exporting SMEs

Non-exporting SMEs

Coeff

S.E

Z

P>|z|

Coeff

S.E

Z

P>|z|

-0.5034458

0.2708786

-1.86

0.063

0.1229583

0.2707441

0.45

0.65

1.956028

0.4090677

4.78

0

0.0756213

0.396042

0.19

0.849

0.5064225

0.3223972

1.57

0.116

0.375049

0.2225648

1.69

0.092

2

-0.004152

0.0060471

-0.69

0.492

-0.0084189

0.0134793

-0.62

0.532

2

M

-0.0492432

0.0159492

-3.09

0.002

0.0286895

0.0151713

1.89

0.059

L2

0.0062523

0.0107947

0.58

0.562

-0.0158149

0.0055336

-2.86

0.004

L*M

-0.0252458

0.0131719

-1.92

0.055

0.0006934

0.0133858

0.05

0.959

K*M

0.0526921

0.0229793

2.29

0.022

-0.0053806

0.0230617

-0.23

0.816

K*L

-0.0114308

0.0139338

-0.82

0.412

0.0108879

0.0108404

1

0.315

K

Constant

-6.366972

2.483535

-2.56

0.01

4.918866

3.26611

1.51

0.132

σ2

0.0699514

0.0158854

0.0388167

0.1010862

0.8345841

0.1536084

0.533517

1.135651

Mean technical efficiency

0.8442877

0.0928712

0.6036868

0.1389179

log likelihood

38.953339

-133.13898

127

134

No of observations

Source: Author’s calculation from BLD

15

Table 5: Estimation Results for Manufacturing SMEs 2006 2006

Exporting SMEs

Non-exporting SMEs

Coeff

S.E

Z

P>|z|

Coeff

S.E

Z

P>|z|

K (Captial)

-0.4056887

0.1770194

-2.29

0.022

-0.5418886

0.1894399

-2.86

0.004

M (Non-capital)

-0.1467092

0.3316214

-0.44

0.658

0.4528358

0.2563331

1.77

0.077

L (labour)

1.722881

0.4100966

4.2

0

0.3120966

0.1874757

1.66

0.096

2

-0.0162494

0.0072316

-2.25

0.025

0.011659

0.0087278

1.34

0.182

2

M

0.081286

0.0196056

4.15

0

0.0126318

0.0094328

1.34

0.181

2

K

L

-0.0121655

0.0268371

-0.45

0.65

-0.0152004

0.0034748

-4.37

0

L*M

-0.116932

0.0325352

-3.59

0

-0.0057128

0.0097504

-0.59

0.558

K*M

-0.0033783

0.0159167

-0.21

0.832

0.0045992

0.0126839

0.36

0.717

K*L

0.0584876

0.0245833

2.38

0.017

0.0217364

0.0077338

2.81

0.005

Constant

0.4233777

1.351578

0.31

0.754

6.281955

2.343575

2.68

0.007

σ

0.0280556

0.0657066

0.3973095

0.0785826

0.2432904

0.5513286

0.6893406

0.1302328

2

0.0468811

0.009605

Mean technical efficiency

0.8892146

0.0459062

log likelihood

47.416823

-83.373746

149

133

No of observations

Source: Author’s calculation from BLD

Table 6: Estimation Results for Manufacturing SMEs 2007 2007 No export

Exporting SMEs Coeff

S.E

Non-exporting SMEs Z

P>|z|

K (Captial) M (Non-capital) L (labour) 2

K

2

M

Estimation results not available under MLE

Coeff

S.E

Z

P>|z|

0.2496593

0.1661907

1.5

0.133

-0.1575443

0.2836054

-0.56

0.579

0.7569434

0.2036306

3.72

0

0.0042837

0.0073407

0.58

0.56

0.0226485

0.0104943

2.16

0.031

-0.0355226

0.0049066

-7.24

0

L*M

0.0301278

0.0108538

2.78

0.006

K*M

-0.0182831

0.0140101

-1.3

0.192

K*L

-0.0022399

0.0050984

-0.44

0.66

L2

Constant σ2 Mean technical efficiency log likelihood No of observations

2.865249

2.256391

1.27

0.204

0.1152774

0.014978

0.0859212

0.1446337

0.997506

0.0000105

-41.317428 179

122

Source: Author’s calculation from BLD

16

CONCLUSIONS AND IMPLICATIONS Using the Business Longitudinal Database, this paper has evaluated efficiency performance of Australian manufacturing SMEs in the period from 2005-2007. It examined the difference between the performance of exporting and non-exporting firms in an attempt to understand the reason for the participation of SMEs in exporting. Estimation results from Stochastic Frontier Model indicate that Australian manufacturing SMEs achieved a high technical efficiency level, especially the exporting SMEs at more than 80 percent. The non-exporting SMEs tend to have lower efficiency level compared to their exporting peers. Their performance is at the similar range recorded for manufacturing firms in developing countries. The results also show an increase in technical efficiency level for both exporting and non-exporting manufacturing SMEs during the period examined. There is evidence of self-selection of more efficient firms into exporting from the increase in exports participation and efficiency level as well as better performance of exporting SMEs. However, there is not enough evidence to conclude about learning-by-exporting for these manufacturing SMEs. Empirical evidence from this paper suggests that in order to increase export participation of Australian SMEs it is important to improve efficiency performance of the SMEs. Traditional instruments such as trade fairs, marketing studies, and prospective missions are useful in reducing entry costs and facilitating entry of new firms, but they may not be enough to sustain firm competitiveness in foreign markets (Alvarez, 2007). By increasing efficiency SMEs will be competitive in the global market. Thus, priority supports to Australian manufacturing SMEs should be given to improvement of efficiency. Areas of support include human resources, skills for employees and technology for the SMEs and not just the traditional instruments listed above. As this paper focuses only on manufacturing SMEs, future studies can examine the performance of other industries in the BLD. A release of the BLD for the recent period 2008-2010 in December 2011, together with the BLD 2005-2007, makes it possible to carry out a longitudinal analysis due to the availability of date for a longer time period. An attempt has been made in the current analysis to breakdown firms by size categories. However, results were not released by ABS due to confidentiality restriction in small

17

samples especially in medium firm (20-200 employees) category. Future research can address this with an analysis carried out for the aggregate sample.

REFERENCES ABS. (2000). Business Longitudinal Survey 1994/95-97/98. ABS. (2010a). Australian Economic Indicators, Cat No. 1350.0. ABS. (2010b). Number and Characteristics of Australian Exporters, 2008-2009. Aigner, D. J., Lovell, C. A. K., and Schmidt, P. (1977). Formulation and Estimation of Stochastic Frontier Production Function Models. Journal of Econometrics 6(1), 21-37. Alvarez, R. (2007). Explaining Export Success: Firm Characteristics and Spillover Effects. World Development, 35(3), 377-393. Alvarez, R., and Crespi, G. (2003). Determinants of Technical Efficiency in Small Firms. Small Business Economics, 20(3), 233-244. Amendolagine, V., Capolupo, R., and Petragallo, N. (2011). Export Status and Productivity Performance: Evidence from Matched Italian Firms. Aussenwirtschaft, 66(2), 151-180. Arnold, J. M., and Hussinger, K. (2005). Export Behavior and Firm Productivity in German Manufacturing: A Firm-Level Analysis. [Article]. Review of World Economics, 141(2), 219-243. Austrade. (2000). Why Australia Needs Exports: The Economic Case for Exporting: A Discussion Paper of the Australian Trade Commission and the Centre for Applied Economic Research (CAER) at the University of New South Wales by Tim Harcourt. Austrade. (2001). Knowing and Growing the Exporter Community. Sydney: Australian Trade Commission. Aw, B.-Y. (2002). Productivity Dynamics of Small and Medium Enterprises in Taiwan. Small Business Economics, 18(1-3), 69-84. Aw, B. Y., and Huang, A. R. (1995). Productivity and the Export Market: A Firm-Level Analysis. Journal of Development Economics, 47(2), 313-332. Battese, G. E., and Corra, G. (1977). Estimation of a Production Frontier Model with Application to the Pastoral Zone of Eastern Australia. Australian Journal of Agricultural Economics, 21(3), 167-179. Brewer, P. (2007). Psychic Distance and Australian Export Market Selection. Australian Journal of Management, 32(1). Coelli, T., Rao, D. S. P., and Battese, G. E. (2005). An Introduction to Efficiency and Productivity Analysis (2nd ed.). New York: Springer.

18

Delgado, M. A., Farinas, J. C., and Ruano, S. (2002). Firm Productivity and Export Markets: A Non-Parametric Approach. Journal of International Economics, 57(2), 397-422. Delgado, M. A., Farinas, J. C., and Ruano, S. (2008). Firm Productivity and Export Markets: A Non-parametric Approach. In D. Greenaway, H. Gorg & R. Kneller (Eds.), Globalization and Productivity. Volume 2. (Vol. 20). Cheltenham, U.K. and Northampton, Mass.: Elgar: Elgar Reference Collection. Globalization of the World Economy series., 511-536 European Commission. (2004). Internationalisation of SMEs. Luxembourg: EIM. Farrell, M. J. (1957). The Measurement of Productive Efficiency. Journal of the Royal Statistical Society. Series A (General), 120(3), 253-290. Fernandes, A. M., and Isgut, A. (2005). Learning-by-Doing, Learning-by-Exporting, and Productivity: Evidence from Colombia” World Bank Policy Research Working Paper No. 3544. Washington D.C.: World Bank. Fryges, H., and Wagner, J. (2008). Exports and Productivity Growth: First Evidence from a Continuous Treatment Approach. [Article]. Review of World Economics, 144(4), 695-722. Gierczycki, L., and Reid, D. M. (1998). Internationalisation of Australian SMEs: A Review of the Literature as a Precursor to A Study of the Cultural Factors in the Internationalisation Process. Paper presented at the International Council for Small Business Conference, Singapore. Gittins, R. (20 April 2002). Why 24 out of 25 Australian Firms Don't Export. Sydney Morning Herald. Graves, C., and Thomas, J. (2006). Internationalization of Australian Family Businesses: A Managerial Capabilities Perspective. Family Business Review, XIX(3), 207-224. Grazzi, M. (2011). Export and Firm Performance: Evidence on Productivity and Profitability of Italian Companies. Journal of Industry, Competition and Trade(Preprints), 1-32. Greenaway, D., and Kneller, R. (2005). Firm Heterogeneity, Exporting and Foreign Direct Investment: A Survey, Globalisation, Productivity and Technology Research Paper Series No. 2005/32. Nottingham: University of Nottingham. Gronum, S., and Verreynne, M.-L. (2011). How Much Innovation is Enough? The Case of Australian Small Firms. Paper presented at the the 8th AGSE International Entrepreneurship Research Exchange, Swinburne University of Technology, Melbourne, Australia, 1-4 February 2011 Hansell, D., and Talgaswatta, T. (2009). Exporting Among Australian Small to MediumSized Enterprises: An Exploratory Analysis of the Business Longitudinal Database. Paper presented at the Australian Conference of Economists 2009, Adelaide, 28-30 September

19

Hill, H., and Kalirajan, K. P. (1993). Small Enterprise and Firm-Level Technical Efficiency in the Indonesian Garment Industry. Applied Economics, 25(9), 11371144. Kalirajan, K. P., and Shand, R. T. (1999). Frontier Production Functions and Technical Efficiency Measures. Journal of Economic Surveys, 13(2), 149-172. Kumbhakar, S., and Lovell, C. A. K. (2003). Stochastic Frontier Analysis (2nd ed.). Cambridge U.K.: Cambridge University Press. Le, V. (2010). Technical Efficiency Performance of Vietnamese Manufacturing Small and Medium Enterprises. Unpublished Doctor of Philosophy thesis, School of Economics, Faculty of Commerce, University of Wollongong, Wollongong. Meeusen, W., and Van den Broeck, J. (1977). Efficiency Estimation from Cobb-Douglas Production Functions with Composed Error. International Economic Review, 18(2), 435-444. Mok, V., Yeung, G., Han, Z., and Li, Z. (2010). Export Orientation and Technical Efficiency: Clothing Firms in China. Managerial and Decision Economics, 31(7), 453-463. Murillo-Zamorano, L. R. (2004). Economic Efficiency and Frontier Techniques. Journal of Economic Surveys, 18(1), 33-77. OECD. (1997). Globalisation and Small and Medium Enterprises (Vol. 1 Synthesis Report). Paris: OECD. OECD and APEC. (2007). Removing Barriers to SME Access to International Markets (No. APEC#207-SM-04.2). Singapore: APEC Secretariat. Oh, D.-h. (2011). Productivity growth, efficiency change and technical progress of the Korean manufacturing industry. [Article]. Journal of the Asia Pacific Economy, 16(1), 50-70. Pascoe, S., Kirkley, J. E., Gréboval, D., and Morrison-Paul, C. J. (2003). Measuring and Assessing Capacity in Fisheries: Issues and Methods. Rome: FAO. Powell, D., and Wagner, J. (2011). The Exporter Productivity Premium along the Productivity Distribution: Evidence from Unconditional Quantile Regression with Firm Fixed Effects. RAND Corporation Publications Department, Working Papers: 837. Redding, S. J. (2011). Theories of Heterogeneous Firms and Trade. Annual Review of Economics, 3(1), 77-105. Sensis. (2011). Sensis Business Index - Small and Medium Enterprises. Melbourne: Telstra Corporation. Storey, D. J. (1990). Firm Performance and Size. In Z. J. Acs & D. B. Audretsch (Eds.), The Economics of Small Firms: A European Challenge. Dordrecht and Boston: Kluwer Academic Publishers, 43-50 Sun, S. (2006). Technical Efficiency and Its Determinants in Gansu, West China. Canberra: Crawford School of Economics and Government, Australian National University. 20

Tybout, J. R. (2000). Manufacturing Firms in Developing Countries: How Well Do They Do, and Why? Journal of Economic Literature, 38(1), 11-44. Vogel, A., and Wagner, J. (2011). Robust estimates of exporter productivity premia in German business services enterprises. Economic and Business Review, 13(1-2), 726. Wagner, J. (2007a). Entry, Exit and Productivity Empirical Results for German Manufacturing Industries. Jena Economics Paper, No. 2007 - 064 Retrieved 5 March, 2008, from http://zs.thulb.unijena.de/servlets/MCRFileNodeServlet/jportal_derivate_00047027/wp_2007_064. pdf Wagner, J. (2007b). Exports and Productivity: A Survey of the Evidence from Firm-level Data. The World Economy, 30(1), 60-82. Wagner, J. (2011a). International Trade and Firm Performance: A Survey of Empirical Studies since 2006: Institute for the Study of Labor (IZA), Discussion Paper, Vol. 5916, Max Planck Society for the Advancement of the Sciences - Max Planck Institute for Economics.

21