The Retail Labour Market

Transcription

1 The Retail Labour Market Report by National Institute of Labour Studies Professor Sue Richardson for Coles Myer Ltd August 2003

2 Contents Executive Summary Introduction Overview of Retail Employment Do Supermarkets Reduce Retail Employment? Productivity in the Retail Industry The Structure of the Retail Industry A Detailed Snapshot of the Retail Industry Changes over Time: Hours Changes over Time: Turnover Changes over Time: Size of Business Employment in the Retail Industry Total Employment Sectors within Retail Employment Female and Male Employment Hours Worked Earnings...44 References...49 List of Tables Table 1: Table 2: Characteristics of the Retail Labour Force, and a Comparison with the Total Australian Labour Force, February Employment, and Wages as a Share of Business Activity, by Type of Food Retail Business 17 Table 3: Output, Labour Input and Productivity in the Retail Sector, to (average annual rates of growth, per cent per year) 20 Table 4: Characteristics of Components of the Retail Industry, Table 5: Retail Group Share in Total Retail Turnover, 1982 and 2003 (per cent) Table 6: Numbers of People Employed in Retailing, 1984,1996 and

3 Figures and Diagrams Diagram 1(b): Structure of All Businesses and Large Businesses, Australia and Retail, Figure 1: Retail Indexes of Output, Hours Worked and Labour Productivity, to (Index = 100) 21 Figure 2: Total Retail Turnover Adjusted for CPI, $m 26 Figure 3: Retail Turnover Trend Estimates, by Industry Group Figure 4: Food Retail Turnover Adjusted for CPI 27 Figure 5a: Figure 5b: Figure 6: Figure 7: Figure 8: Figure 9: Figure 10: Number of Retail Businesses by Employer Size Group, , Number of Retail Businesses by Employer Size Group, m, (reported as deviations from the starting value) 30 Employment in Small Retail Businesses as a Proportion of All Employment in Retail Trade, , Employment in Small Retail Businesses as a Proportion of All Employment in Retail Trade, , Non-employing Retail Businesses as a Proportion of All Retail Businesses, , Number of People Employed in Australian Businesses, m, (reported as deviations from the starting value) 35 Number of Male and Female Employees in the Retail Industry Figure 11(a): Number of Employers in the Retail Industry Figure 11(b): Male and Female Own-Account Workers in the Retail Industry Figure 12: Figure 13: Number of Persons Employed Full-time and Part-time in the Retail Industry as a Proportion of People Employed Full-time and Part-time in All Industries, Hours Worked in the Retail Industry and a Proportion of Hours Worked in All Industries, Figure 14: Real Average Weekly Earnings in Retail, , = Figure 15: Figure 16: Figure 17: Real Average Weekly Earnings, Full-time Adults in Retail, , Base = Average Weekly Earnings for All Persons Employed in the Retail Industry as a Proportion of the Average Weekly Earnings for All Persons Employed in All Industries, Average Full-time Weekly Earnings for Persons Employed in the Retail Industry as a Proportion of the Average Full-time Weekly Earnings for All Persons Employed in All Industries,

4 Executive Summary Retailing is a major economic activity and employs more people than any other Australian industry. This report describes the character of employment in the retail industry, how it compares with employment in other industries, how it has changed over time, and how it differs between large and small retailers. This information is placed in the context of developments in the economy as a whole. Changes in the characteristics of employment over time are related to changes in the structure of the retail industry. Part of the examination includes a direct look at the question of whether growth in market share by the large retailers is detrimental to employment in the retail industry. In May 2003, retailing comprised about 5 per cent of the Australian GDP (compared with 3% for agriculture) and employed 1.5 million people, or 15 per cent of all employment. Like the rest of the economy, the retail industry has the characteristic that most firms are small yet half of the workforce work in large firms. The handful of large businesses accounted for 41 per cent of income. Retail symbolizes much of the change that has been occurring in the Australian workplace over the past 25 years, with a relatively small proportion of employees being full-time males and many working on part-time and casual terms. More than half of retail workers are employed part-time, two-thirds of whom are female. The workforce is also remarkably young, which will pose a challenge for the industry as the cohort of young workers shrinks in size over the coming decades. The distinctive features of retail employment The character of employment in retail is quite different from employment in the rest of the economy. This is especially so if the sale and service of motor vehicles is excluded: car retailing has a relatively masculine, full-time and higher paid workforce. Compared with all employment, employment in retail has:

5 a much higher proportion of part-time workers, and a higher proportion of women; a higher than expected number of employers, especially female employers; a very young workforce: half of employees aged work in retail but only 11 per cent of those aged over 54; an occupational structure that is quite different from the workforce as a whole, with few managers and administrators and two thirds of all people employed in elementary clerical, sales and service jobs; a disproportionate number of people working part-time, including almost one third of all people who work fewer than 16 hours per week; relatively low wages; a relatively low level of formal education among employees. Does supermarket growth reduce employment? We examine the question of whether an expansion of supermarkets at the expense of smaller stores would reduce total retail employment. To do this, it is necessary to compare supermarket employment with other areas of retail that are similar to supermarkets in the service and products they provide, but differ in size. We find that the employment per unit of sales and per unit of production is at least as high in the large supermarkets/grocery stores as it is in the small greengrocers and liquor stores. While the evidence is not conclusive, it provides no support for the view that the transfer of business from the small stores to the large supermarkets would decrease total employment or labour earnings. Changes in retail over time Over the 24 year period from to , multifactor productivity in the retail sector is estimated to have risen at an average rate of 0.3 per cent per annum. The comparable figure for the whole market sector of the economy is 0.9. Multifactor productivity in the retail sector is estimated to have fallen over the decade , then turned around to rise at a rate of 1.3 per cent for the remainder of the 1990s. Labour inputs, which are measured as the number of hours worked, have grown over

6 the whole period, but at different rates. Growth was especially fast in the period to and continued to be quite strong for the rest of the 1990s. Labour productivity, however, fell in the middle period, while growing strongly at the beginning and end of the period. There is considerable diversity within the retail sector in growth, productivity and employment changes over time. Supermarkets and grocery stores are the largest employers within the retail sector. They saw a steady growth in output but fluctuating changes in employment over the past quarter century, with no growth until after the recession of the early 1990s. After a few years of employment growth, employment has remained steady and the growth in output has been accomplished largely by an increase in output per hour worked (labour productivity), rather than by an increase in employment. Hours of employment have grown fastest (over 60 per cent) in Furniture, houseware and appliances and in Other personal and household goods. They have fallen quite substantially in Department stores and slightly in Clothing and soft goods. The growth in labour hours is not systematically related to growth in output, which causes trends in labour productivity to vary substantially across the sectors. The highest growth in output (50 per cent or more) is in Other personal and household goods, Furniture, houseware and appliances (the two sectors where employment has also grown fastest) and Supermarkets and grocery stores. The selling of food and groceries dominates the retail sector, accounting for 43 per cent of the gross product of the industry. The relatively rapid growth in sales of food and groceries has been in the supermarkets, rather than in the specialist food shops. Supermarkets and grocery stores predominate in terms of output, but small specialised food retailers provide almost as much employment, though at a lower level of annual earnings. The level of annual earnings reflects both hourly rates of pay and number of hours worked. These small food stores are relatively labour intensive, contributing more to employment than their share of industry gross product would suggest.

7 As is shown in more detail in the report, the major changes over the past 18 years in retail employment are: Total employment in retailing grew faster than in the economy as a whole, but a little slower than employment in the private business sector; There has been a sizeable fall in the proportion of retail employment that is in small (less than 20 workers) businesses; Self-employment in retail has fallen from 22 per cent to 14 per cent of all retail employment with most of the fall occurring in the last decade; Part-time employment grew much faster than full-time employment (147% compared with 27%); Employment in Food Retailing grew relatively fast (78%) ; The growth in part-time employment was especially strong (181%), and in fulltime employment especially weak (13%), in Food Retailing. By 2003, there were almost as many people employed in Food Retailing as in Personal and Household Goods. Food Retailing has a high and rising ratio of part-time to full-time staff (1.6 compared with 0.8 for Personal and Household Goods in 2003). Female retail employment in general has not fared well over the past 15 years. The absolute number of women employed full-time has not grown, remaining at about 250,000. The number employed part-time has grown steadily from 225,000 to 425,000 but this has just matched the general growth in part-time employment for women. The combination of a falling share of women s full-time employment accounted for by retail (from 14.5 per cent to 11.8 per cent), and a constant share of women s part-time employment means that the retail share of all hours worked by women has fallen since While total hours worked by women in retail have been growing a little faster than total hours worked by men, the hours growth for women has not matched that occurring in other parts of the economy. Thus there are two trends. Within the retail sector, the share of total hours worked by women has been rising. But for the

8 economy as a whole, the share of total hours worked in the retail sector, and especially the share of women s total hours worked in the retail sector, has been falling. It is clear that over the past 15 or so years in retail there has been a trend towards larger businesses and away from small and owner-operated businesses. While the absolute number of small businesses has not fallen, their share of all businesses has. Employment in small businesses has fallen substantially as a share of all retail employment, to just under half, and this trend has been underway for the whole period since the mid 1980s. Own account workers have also fallen substantially as a proportion of all retail workers. But the fall in self-employment has been quite recent, the downturn starting in the mid-1990s. It is not yet clear that this is part of a longterm trend. Retail employment is not well paid. The average adult full-time retail worker earns about three-quarters of the amount earned by the average Australian adult full-time worker. One reason for the relatively low pay is the high ratio of low paying occupations, such as sales, to high paying occupations, such as professional and managerial, in the retail sector. Relative pay for full-time women retail workers is higher than it is for men and has fluctuated less. As for men, however, it has fallen somewhat since the mid-1980s. In brief, the composition of employment in the retail sector has been changing. The self-employed are diminishing as a component of total retail employment and becoming more masculine; employees are becoming more feminine; there has been a rapid expansion of the role of part-time staff and increasingly people are employed by large rather than small businesses.

9 1. Introduction 1 Retailing is a major economic activity and a major employer. In May 2003, it comprised about 5 per cent of the Australian GDP (compared with 3% for agriculture) and employed 1.49 million people, or 15.6 per cent of all employment 2. Indeed, it is the biggest employer of all industries in Australia. Its output (ie, value added) rose in real terms at 2.7 per cent per annum between and , to $31.1 billion. 3 It has a structure much like that of the rest of the private sector, with small businesses (employing fewer than 20 people) comprising 97 per cent of all businesses but only 38 per cent of total income. The fewer than one per cent of large businesses which employ 200 people or more accounted for 41 per cent of income. It symbolizes much of the change that has been occurring in the Australian workplace over the past 25 years, with a relatively small proportion of employees being full-time males and many working on part-time and casual terms. More than half of retail workers are employed part-time, two-thirds of whom are female. How retail employment is changing, and whether it represents an exemplar of the future of work in general is a question of great social interest. The past 25 years has seen a large withdrawal of men from the labour force and a large growth in the employment of married women. It has also seen a substantial substitution of part-time and casual jobs for full-time and continuing jobs. These changes have important implications for the way that families work, whether they are formed at all and their capacity to provide secure and adequate incomes from working. Employment in retailing has given many people their first work experience. It has been less good at providing stable employment with reasonably predictable weekly earnings and some prospects for career progression. 1 Members of the NILS team, particularly Lauren Miller Lewis and Skye Jacobi, have made substantial contributions to this report. 2 ABS Labour Force, catalogue number , Table 4. 3 Value added is the value of sales minus the value of purchased inputs: it is the same as wages and salaries, plus depreciation, interest and profits. The source for these data is the ABS National Accounts. Note that retail as defined by the ABS excludes the provision of services that are sold direct to consumers, such as hairdressing, real estate sales, house cleaning and trades. See Merrilees and Miller, 1996.

10 This report describes the character of employment in the retail industry, how this has changed over time, and how it differs between large and small retailers. Where possible, this information is placed in the context of developments in the economy as a whole. Changes over time are related to changes in the structure of the retail industry. Part of the examination includes a direct look at the question of whether growth in market share by the large retailers is detrimental to employment in the retail industry. In answering this question, we believe it is necessary to consider the meaning of detrimental. For example, if a given number of low paid, casual, deadend jobs was replaced by a smaller number of better paid, more secure jobs with prospects of advancement, that might be considered beneficial rather than detrimental. The quality of jobs and the skills development of those who occupy them matters, as well as the number of jobs. Employment can be categorized in a number of ways. These include by type of employment contract and by the types of people who are employed. We examine both perspectives. The report focuses on retailing in general, excluding car sales and service where possible. To the extent possible with the available data, the report also looks separately at food retailing and small and large businesses. Note that the usual measure of retailing, and the one being used here, includes only the sale of goods direct to the consumer, not of services. 2. Overview of Retail Employment We start with a picture of the structure of the retail sector, and a comparison with the structure of Australian businesses as a whole. The retail sector described in the tree diagram below is based on the definition of retail as used by the Australian Bureau of Statistics, from where the data are drawn. The ABS defines retailing as comprising businesses mainly engaged in the resale of new or used goods to final consumers for personal or household consumption or in selected repair activities such as repair of

11 household equipment or motor vehicles. (ABS, 1993:157) The data include motor vehicles sales and service, which constitute about 20 per cent of total retailing. The structure of businesses and of employment in the Australian economy is set out in the tree diagram, Diagram 1 below. Comparable data for the retail industry are included in each box, in italics. The percentage that retail represents of the relevant category is provided in parentheses. Thus, in there were 168,700 retail businesses, employing 1,216,200 people. These represent 13.2 per cent and 14.2 per cent (respectively) of all businesses and of all employment 4. Diagram 1 (a) gives additional data on small business, while Diagram 1 (b) focuses on large businesses. Further down Diagram 1 (a) we see that there were 93,000 private sector nonagricultural small retail businesses that employed staff, in which 501,100 people were employed. This represented 17.2 per cent of such small employing businesses and 19.7 per cent of people employed in such small employing businesses. We can learn a number of things from this diagram. The retail sector, at 14 per cent of total employment, is a big employer. The average size of retail businesses (in terms of employment) is similar to that for all businesses. This is indicated by the fact that the share of retail in the number of businesses (13.2%) is almost the same as its share in the number of persons employed (14.2%). Because all retail firms are in the private sector, and none are in agriculture, forestry or fishing, then retail is a larger share of non-agricultural private businesses (14.5%) and employment (17.6%). It does not have an unusually large proportion of small businesses, since its share of nonagricultural small businesses (14.5%) is the same as its share of all non-agricultural businesses (14.5%). It has a somewhat lower than average number of non-employing businesses and a larger number of small employing businesses. Among the small employing businesses, it has a much higher proportion than other industries of people who are working in their own business (23.6% of all such people) rather than working as employees.

12 The picture for large businesses mirrors that of small retail businesses. Large retail businesses are 14 per cent of all non-agricultural private sector large businesses the same share as small retail businesses. Like small businesses, their share of employment exceeds their share of businesses. This implies that on average each large retail business employs more people than the typical large business. We next present a much more detailed look at the characteristics of the retail labour force, and compare these with the national picture. When reading the right hand column of the table below, it is helpful in most instances to keep in mind that 15 per cent of all employed people were employed in retail, in Then we can see, for example, that slightly over one third were employed in food retailing, and that retailing employed much more than its share of workers aged (50% of such workers, compared with 15% of workers overall). 4 This is the latest date for which data in this form can be obtained. The difference between these estimates of the size of retail and those given earlier are accounted for by minor differences in data source and definition as used by the ABS.

13 Diagram 1 (a):structure of Australian Business and the Retail Sector, ALL BUSINESSES (a) Total public and private sector 1,281,700 businesses Retail 168,700 (13,2%) 8,564,500 persons employed Retail 1,216,200 (14.2%) PRIVATE SECTOR PUBLIC SECTOR Agriculture, forestry and fishing(b) 112,100 businesses (c) Retail 0 355,000 persons employed(c) Non-agricultural sector 1,164,100 businesses Retail 168, 700 (14.5%) 6,901,900 persons employed Retail 1,216,200 (17.6%) Public trading and General government organisations 5,500 organisations Retail 0 1,307,600 persons employed PRIVATE SECTOR SMALL BUSINESSES Agriculture, forestry and fishing(d) 111,200 businesses Retail 0 304,000 persons employed Non-agricultural small businesses(e) 1,122,000 businesses Retail 162,800 (14.5%) 3,259,100 persons employed Retail 595,900 (18.3%) Non-employing businesses 582,100 businesses Retail 69,800 (12.0%) 713,200 own account workers Retail 94,800 (13.3%) Employing businesses 539,900 businesses Retail 93,000 (17.2%) 2,545,900 persons employed Retail 501,100 (19.7%) Employers 276,500 persons employed in their own business Retail 65,200 (23.6%) Employees (wage & salary earners) 2,269,400 persons employed Retail 435,900 (19.2%) Notes: (a) Generally, the number of businesses (management units) and persons employed have been obtained by averaging the estimates for the middle months of each quarter for the financial year. (b) Includes ANZSIC Subdivisions 01 Agriculture, 02 Services to Agriculture; Hunting and Trapping, 03 Forestry and Logging and 04 Commercial fishing. (c) Estimates are based on data from two different sources; ANZSIC Subdivision 01 data are drawn from the Agricultural Finance survey, while ANZSIC Subdivisions 02, 03 and 04 estimates are drawn from the Economic Activity Survey. Excludes management units in ANZSIC Subdivision with an estimated annual value of agricultural operations (EVAO) of less than $22,500. Employment estimates exclude unpaid family helpers. (d) Agricultural small businesses include those management units coded to ANZSIC Subdivision 01 with an EVAO of more than $22,500 but less than $400,000, and those management units coded to ANZSIC Subdivisions 02, 03 and 04 which employ less than 20 persons. (e) Small business (except in agriculture) are defined as those management units which employ less than 20 persons. Source: Australian Bureau of Statistics (ABS) (2001), Small Business in Australia, Cat. no, p.6, 12.

14 Diagram 1(b): Structure of All Businesses and Large Businesses, Australia and Retail, ALL BUSINESSES (a) Total public and private sector 1,281,700 businesses Retail 168,700 (13.2%) 8,564,500 persons employed Retail 1,216,200 (14.2%) PRIVATE SECTOR PUBLIC SECTOR Agriculture, forestry and fishing(b) 112,100 businesses (c) Retail 0 355,000 persons employed(c) Non-agricultural sector 1,164,100 businesses Retail 168, 700 (14.5%) 6,901,900 persons employed Retail 1,216,200 (17.6%) Public trading and General government organisations 5,500 organisations Retail 0 1,307,600 persons employed PRIVATE SECTOR LARGE BUSINESSES Agriculture, forestry and fishing(d) 900 businesses Retail 0 51,000 persons employed Non-agricultural large businesses(e) 42,100 businesses Retail 5,900 (14.0%) 3,642,800 persons employed Retail 620,300 (17.0%) Notes: (a) Generally, the number of businesses (management units) and persons employed have been obtained by averaging the estimates for the middle months of each quarter for the financial year. (b) Includes ANZSIC subdivisions 01-Agriculture, 02-Services to agriculture, hunting and trapping, 03-Forestry and logging and 04-Commercial fishing. (c) Estimates are based on data from two different sources; ANZSIC subdivision 01 data are drawn from the Agriculture Finance Survey, while ANZSIC subdivisions 02, 03 and 04 estimates are drawn from the Economic Activity Survey. Excludes management units in ANZSIC subdivision 01 with an estimated annual value of agricultural operations (EVAO) of less than $22,500. Employment estimates exclude unpaid family helpers. (d) Agricultural small businesses include those management units coded to ANZSIC subdivisions 01 with an VAO of more than $22,500 but less than $400,000 and those management units coded to ANZSIC subdivisions 02, 03 and 04 which employ less than 20 persons. (e) Large businesses (except in agriculture) are defined as those management units which employ more than 20 persons, except for manufacturing where large businesses are those which employ more than 100 persons. Source: Australian Bureau of Statistics (ABS) (2001), Small Business in Australia, Cat. no, p.6, 12.

15 Table 1: Characteristics of the Retail Labour Force, and a Comparison with the Total Australian Labour Force, February 2003 Retail employment % of equivalent for whole workforce Total employment in retailing* 1, Employment in Food retailing Employment in Personal and household goods Total retail employment excluding motor vehicles Full-time males Full-time females All full-time Part-time males Part-time females All part-time All males All males in retailing excluding motor vehicles Married females All females All females in retailing excluding motor vehicles Male employers Female employers Male own account workers Female own account workers Male employees Female employees Aged Aged Aged Aged Aged Aged Born in Australia 1, Born outside Australia Managers and administrators Professionals Associate professionals Trades etc Advanced clerical and service Intermediate clerical and service Intermediate prod. and transport Elementary clerical, sales, service Labourers and related Average weekly hours worked Work 1-15 hours Work hours Work hours Works hours Works more than 48 hours Laid off/retrenched, now unemployed Involuntarily lost job, now unemployed Voluntarily quit job, now unemployed Total previously employed, now unemployed Note: * Retailing includes sale and repair of vehicles unless indicated otherwise. Source: ABS, Labour Force, Australia, February 2003, catalogue number 6203.

16 From Table 1 we see that retail employment, compared with total employment for the whole economy: When broadly defined, comprises 15.4% of total employment; When retailing excludes the sale and repair of cars, comprises 12.7% of total employment; Has a much higher proportion of part-time workers (about one quarter of all parttime employees), and a higher proportion of women; Has a higher than expected number of employers, especially female employers: they comprise one quarter of all female employers; Has a very young workforce: half of employees aged work in retail but only 11 per cent of those aged over 54; Has an occupational structure that is quite different from the workforce as a whole, with few managers and administrators and two thirds of all people employed in elementary clerical, sales and service jobs; Has a disproportionate number of people working part-time, including almost one third of all people who work fewer than 16 hours per week; Has a disproportionate number of former workers who are now unemployed a relatively high proportion of whom quit their jobs voluntarily. While not shown in the table, it is also true that employees in retail have less formal education than the Australian workforce as a whole. For example, in per cent of retail employees had post-school qualifications compared with 40 per cent for all workers. Put another way, 9 per cent of workers with post-school education and 21 per cent of workers without post-school qualifications worked in retail. Between 1984 and 1997, the proportion of all workers without post-school qualifications who worked in retail rose, from 18 per cent to 21 per cent. Retail employed a higher proportion of such workers than any other sector (the next largest employer being manufacturing, with 14 per cent) 5. The major characteristics of retail employment are not static. An examination of how they have changed over the past 15 years shows that, between 1984 and 2003: Total employment in retail (excluding motor vehicle sales and service) has grown from 700,000 to 1,204,000, or by 72 per cent; 5 See Barnes, P, Johnson, R, Kulys, A and Hook, S, 1999, Productivity and the Structure of Employment, Productivity Commission staff research paper, Ausinfo, Canberra, p.27

17 This is faster than the rate of growth of all employment (58 per cent); Employment in personal and household goods has grown less rapidly than in retail as a whole and in food retailing; There has been a rise in the proportion of all employment that is accounted for by retail employment and by food retail employment, but no rise in the proportion employed in personal and household goods; The number of people employed part-time has more than doubled for food and for total retailing and has grown more slowly for personal and household goods; In consequence, part-time employment in food and in total retailing has grown, as a share of all part-time employment, especially for the former: it has fallen for personal and household goods retailing; There has been no growth in full-time employment in personal and household goods retailing and little in food or in total retailing; In consequence, full-time employment in total and in the main categories of retail (excluding motor) has fallen as a proportion of all full-time employment; Total hours worked in retail have grown by over 10 million, but they have fallen as a proportion of all hours worked, from 13.1% to 12.5%. 3. Do Supermarkets Reduce Retail Employment? In this section we seek to establish whether the expansion of supermarkets at the expense of small businesses is likely to reduce overall retail employment. It would do so if the labour content of a unit of sales were higher if the goods were sold by a small business than if the same goods were sold by a supermarket. By labour content we mean the amount of worker time together with the wage paid for that time. This is an empirical question. In order to answer it we draw on data from Retail Industry Census (the latest one available). Among other things, the census contains information on: Turnover Cost of goods sold Wages and other costs of employing labour Interest, depreciation, rent, insurance and other expenses Number of employed persons Industry Gross Product.

18 It is not easy to establish whether small businesses are more labour intensive in their operations than are larger firms. In principle, the share of labour costs in value added (the difference between revenue and the cost of purchased inputs) is the best measure of how labour intensive is a firm or industry. But many factors can affect labour s share of value added at a given moment of observation. In order to identify the effects of firm size, we need to be able to compare firms that differ in size but otherwise are the same. For example, they need to be selling the same sort of product, and located in similar places. In a retail store, two things are being sold. One is the product. The other is the service that surrounds the sale. Some products, such as spectacles, inherently require more sales service than do others, such as milk. Some products, such as fresh bread and takeaway food, are prepared on the premises and the labour involved in preparation is indistinguishable from the labour involved in the selling. If small firms specialise in the sale of products that require above average levels of service or instore preparation, then the labour content of their value added will be higher. In this case the high labour content is caused by the nature of the product sold and not by the fact that the retail businesses are small. In the table below, we examine two measures of the labour intensity of the retail activity. The retail businesses are confined to those that sell food and groceries, in order to make them as comparable as possible in their activities. The table distinguishes supermarkets and grocery stores on the one hand and a range of specialized food and drink stores on the other. It is not possible to separate supermarkets from other grocery stores. Table 2 shows that, compared with the value of goods purchased and other expenses, the labour content of what is sold is particularly high for bread and cakes and high also for takeaway food and meat, fish and poultry. It is lowest for liquor and low also for supermarkets and groceries and greengrocers. Thirty per cent of total turnover goes to paying wages for bread and cake shops, whereas only 7 per cent of turnover goes to paying wages in the sale of liquor. This large difference in labour content occurs despite the fact that the average size of business is about the same in the two types of retail outlet. It is reasonable to conclude that the nature of what is sold is the key factor driving the share of turnover accruing to wages, rather than the size of the

19 business. Where a product is partly or wholly produced on the premises, the labour content of what is sold is naturally higher than when a business provides only a retail service. In terms of the type of retail product sold, perhaps the best match with supermarkets is greengrocers and liquor stores. Clearly, the average greengrocer/liquor store (with 8-10 workers) is much smaller than the average supermarket/grocer (with 62), so this gives us an opportunity to see if there is a big difference in labour content according to size. If labour content is measured as labour costs as a fraction of turnover, the answer is no: labour content is about the same for all three types of retail outlet, with liquor stores having the lowest fraction. An alternative measure of labour intensity is payments to labour as a share of industry value added. Value added is defined by the ABS as Turnover less purchases and selected other expenses plus (or minus) the change in the value of stocks. (ABS, 1994:46). The share of labour in value added indicates how labour intensive, as distinct from capital or shop space intensive, are the arrangements used for providing the retail service.

20 Table 2: Employment, and Wages as a Share of Business Activity, by Type of Food Retail Business Labours share in Type of business Supermarkets & groceries Meat, poultry fish, Fruit and vegetables Total Income Liquor 7 Bread cake Takeaway food and Industry value added Number of employees per business Wages per employed person $ 62 16, , , , , ,128 Source: derived from ABS Retail Industry Census, , cat. no Table 2 shows that the share of value added paid to labour was highest for supermarkets and grocery stores, and lowest for liquor stores. Indeed, at 59 per cent, supermarkets and grocery stores have the second highest labour share of any of the components of retail that are included in the table. It is important to note that labour s share includes payments to working proprietors and partners, as well as to employees. It includes not just wages, but also other labour costs such as superannuation and workers compensation premiums. The two measures of labour intensity that we have used both suggest that (the large) supermarkets and grocery stores have at least as high,, if not higher, labour content in their sales than do the comparable (small) segments of retailing covered by green grocers and liquor stores. The figure on which we have been concentrating is overall payments to labour. It does not directly compare the number of people employed because it does not distinguish part and full time employment or high and low paid employees. The final column of Table 2 shows that the annual earnings of people employed in supermarkets and

21 grocery stores tend to be higher than those of other food retail segments. (They much higher than for takeaway food, where payments are well below average.) They are slightly higher than payments to workers in greengrocers, and less than in liquor stores. On this evidence, the size of the wage bill is a reasonable indicator of the quantity of employment. Wages and salaries in supermarkets are the same or higher as a share of turnover and of value added as they are in greengrocers and liquor stores. This suggests that the employment per unit of sales and of production is at least as high in the large supermarkets/grocery stores as it is in the small greengrocers and liquor stores. The transfer of business from the small stores to the large supermarkets would not, on this evidence, decrease total employment or labour earnings. Thus there is nothing in Table 2 to support the view that the expansion of supermarkets at the expense of smaller food retail outlets would reduce overall employment. 4. Productivity in the Retail Industry The changes summarized above have occurred in the context of an economy that has had several major developments over the period since the mid-1980s. Overall, the growth in output and productivity for the Australian economy has been robust, compared with other developed countries. The 1980s saw a strong growth in employment and output, but quite low productivity growth: this has been ascribed in part to the success of the Accord in keeping wage growth low and in encouraging employment. Broadly, the reverse occurred in the 1990s, with strong growth in productivity and less success in increasing employment. In between, there was a severe recession in , which caused low productivity, output and employment for a few years. Since this recession, macroeconomic conditions in Australia have seen sustained and quite high levels of growth in output, if not so much in employment. In the retail sector, output is measured by the Australian Bureau of Statistics (ABS) as the difference between the value of sales and the cost of the goods sold (value added). This measure assumes that the average level of service associated with the sale of goods does not change over time. Productivity is measured as the ratio of the value of

22 output to the value of the inputs. Labour productivity is the value of output divided by the number of labour hours. Capital productivity is the value of output divided by the value of services from capital equipment. Multifactor productivity is the value of output divided by the value of both the labour and capital inputs. Both the value of output and the value of capital services are difficult to measure in ways that make them comparable over time. Hence some caution should be exercised in interpreting trend figures. Over the 24 year period from to , multifactor productivity in the retail sector is estimated to have risen at an average rate of 0.3 per cent per annum. The comparable figure for the whole market sector of the economy is Multifactor productivity in the retail sector is estimated to have fallen over the decade , then turned around to rise at a rate of 1.3 per cent for the remainder of the 1990s. Multifactor productivity rises, by definition, when output rises faster than inputs. Table 3 below shows the behaviour of the key productivity measures, for three intervals over the period The growth in output varied considerably over the three periods reflected in Table 3. At 1.5 per cent per annum during to , it was less than half the growth that was achieved for the remainder of the decade (3.9 per cent pa). Productivity growth tends to decline when output growth slows and vice-versa, and this is apparent in the table. It is notable that when output growth was slowing, input (particularly labour input) growth continued at a rapid rate. Labour inputs, which are measured as the number of hours worked, have grown over the whole period, but at different rates. Growth was especially fast in the period to and continued to be quite strong for the rest of the 1990s. Labour productivity, however, fell in the middle period, while growing strongly in the earlier and later periods. 6 This and subsequent productivity figures are taken from Johnson et al, 2000.

23 Table 3: Output, Labour Input and Productivity in the Retail Sector, to (average annual rates of growth, per cent per year) to to to Output Total inputs Labour Labour productivity Multifactor productivity Source: Estimates based on unpublished ABS data, Gretton and Fisher (1997). Figure 1, which is taken from Johnston et al (2000), shows the output, employment and labour productivity trends in sub-sets of the retail industry, for the period from to , using unpublished ABS data. It is interesting to see the variety of experience across the different parts of retail. The data are able to exclude motor vehicles, which represent about a quarter of retail industry output and 18 per cent of employment. Hours of employment have grown fastest (over 60 per cent) in Furniture, houseware and appliances and in Other personal and household goods. They have fallen quite substantially in Department stores and slightly in Clothing and soft goods. The growth in labour hours is not systematically related to growth in output, which is why trends in labour productivity vary so much across the sectors. The highest growth in output (50 per cent or more) is in Other personal and household goods, Furniture, houseware and appliances (the two sectors where employment has also grown fastest) and Supermarkets and grocery stores. Interestingly, labour productivity has grown fastest in Department stores, which is a sector that has seen slow growth in output. It obtained the productivity increase (of 50 per cent) by substantially reducing the number of hours employed. The next best productivity performer, Supermarkets and grocery stores, had a substantial rise in output and obtained its productivity growth by meeting this growing demand with only a modest growth in hours worked.

24 Figure 1: Retail Indexes of Output, Hours Worked and Labour Productivity, to (Index = 100) Source: Johnston, A., Porter, D., Cobbold, T. and Dolamore, R. (2000), Productivity in Australia s Wholesale and Retail Trade, Productivity Commission.

25 5. The Structure of the Retail Industry The retail industry covers the sale of a wide range of products, from flowers to boats to fast food. The nature of businesses and of employment differs with the type of product being sold. Butchers differ from department stores which in turn differ from fast food outlets. In the analysis of changes in the structure of the retail industry, we must rely on data from the Australian Bureau of Statistics. The ABS sub-divides the retail industry into three main components. These components are Food, Personal and household goods and Motor vehicle retailing and services. In what follows, Motor vehicle retailing and services are excluded from the data, so that we refer only to Food and Personal and household goods. Further subdivisions are made within each of these components, but most data are published only at the more aggregated level. 5.1 A Detailed Snapshot of the Retail Industry A census of the retail industry conducted by the ABS in (the last that has been conducted) provides a snapshot look at the level of employment, number of management units, wage bill and contribution to industry gross product by detailed components of the retail industry. Table 4 below provides a summary of this information. The selling of food and groceries dominates the retail sector, accounting for 43 per cent of the gross product of the industry. Supermarkets and grocery stores predominate in terms of output, but small specialised food retailers provide almost as much employment, though at a lower level of annual earnings. The level of annual earnings reflects both hourly rates of pay and number of hours worked. These small food stores are relatively labour intensive, contributing more to employment than their share of industry gross product would suggest.

26 Table 4: Characteristics of Components of the Retail Industry, Retailing Description Persons employed per management unit* % Persons employed Labour costs per employee $ % Industry operating profit before tax Supermarket & grocery stores 62** 26 18, Specialised food , Department stores 3,032** 11 18,000 5 Clothing and soft goods ,400 8 Furniture, houseware & , appliances Recreational goods ,500 5 Other personal & household , goods Household equipment repair services ,700 1 Note: *A management unit is in nearly all cases the legal entity that owns the business (company, partnership, sole trader, etc.). In a large diversified business it may be a division of that company. ** These estimates have high standard errors and should be used with caution. Note that employed persons includes the self-employed, whereas employee does not. Source: Australian Bureau of Statistics, Retail Industry , Cat. No Most individual retailers employ 7 to 11 people. The big exception is the Department stores, where each company (called a management unit by the ABS) employs on average 3,032 people. This number reflects both a relatively large number of workers per store, and a large number of stores per company. Despite their large size, Department stores employed (in ) only 11 per cent of all retail industry workers, and generated 5 per cent of industry pre-tax profit. They are no more nor less labour intensive than other areas of retail, paying out 11 per cent of total labour costs, while employing 11 per cent of the total workforce. (This last figure is not shown in the table). Supermarkets and grocery stores also differ from the predominant pattern of small businesses, with an average number of people employed of 62. These stores employ over one quarter of all retail industry workers and pay a commensurate share of the total wage and salary bill. Average annual labour costs per worker differ quite a lot across the retail industry. The sectors that have the highest per worker costs, probably reflecting a higher ratio of full to part-time employment, as well as somewhat higher hourly wages, are Furniture, houseware and appliances and Household equipment repair. The sector with clearly the lowest annual earnings is Specialised food. A more detailed look

27 within Specialised food shows that the low labour costs per worker of this sector is generated almost entirely in Takeaway food, which had an average per worker cost of only $10,600. Low hourly wages, paid to youth, as well as a high level of part-time employment undoubtedly both contribute to this outcome. 5.2 Changes over Time: Hours In Figure 1 we showed how each of the sectors of the retail industry have evolved between and , in terms of output, hours worked and productivity. We have already discussed the productivity aspects of these changes. Here we focus on the changes in hours worked. Substantial growth in hours worked is apparent for three sectors, namely Furniture, houseware and appliances, Other personal and household goods and Specialised food. The sector with the fastest growing employment was also the highest paying, namely Furniture, houseware and appliances. It seems to have been largely unaffected by technological change, as productivity per worker hour has fallen over the period. The next fastest growing area of employment, Other personal and household goods, was about average in terms of annual earnings. It experienced a rapid growth in output as well as in employment, and modest labour productivity growth. The largest employer of all, Supermarkets and grocery stores, saw a steady growth in output but fluctuating changes in employment, with no growth until after the recession of the early 1990s. After a few years of employment growth, employment has remained steady and the growth in output has been accomplished by an increase output per hour worked (labour productivity). 5.3 Changes over Time: Turnover Developments in the retail industry can be viewed through a variety of lenses. These include turnover, employment, value added, total wages and total profits. Here we focus on turnover as a measure of size. Figure 2 shows how the value of turnover in retail (excluding motor vehicles) has grown over the period The information is expressed in constant value