DECISION

Note: Data are expressed in original terms. The market sector includes all industries except for Public administration and safety, Education and training and Health care and social assistance.

Source: ABS, Estimates of Industry Multifactor Productivity, 2017–18, Catalogue No. 5260.0.55.002.

[105] Real unit labour costs continue to remain at unusually low levels, 86 falling by a further 1.4 per cent over the year to the December quarter 2018, while nominal unit labour costs increased by 1.6 per cent (Chart 2.7). Real unit labour costs remain below their 10-year average. This implies reduced labour cost pressures over the past decade.

Source: Statistical report, Chart 2.3; ABS, Australian National Accounts: National Income, Expenditure and Product, Dec 2018, Catalogue No. 5206.0.

Business competitiveness and viability

[106] Profits across the whole economy grew by 10.5 per cent over the year to December quarter 2018, which was significantly higher than the preceding year, and above the 5-year and 10-year averages (Table 2.3). This was driven by the 26.3 per cent growth in Mining profits, which accounted for 84.2 per cent of profits growth across all industries. In comparison, profit growth in the non-mining sector was 2.5 per cent over the year, lower than the previous year as well as the 5-year and 10-year averages. Over the past 5 years to the December quarter 2018, total profits growth in the non-mining sector was significantly lower than in Mining, but there is considerable volatility in the yearly figures over both the past 5 years and 10 years. We note that profits have grown in the non-mining sector in every year over the past 10 years at an annual average of 3.9 per cent.

	Mining	Non-mining	All industries
	%	%	%
Dec-08	95.5	–5.0	18.8
Dec-09	–42.6	10.5	–10.1
Dec-10	62.5	1.2	16.4
Dec-11	4.3	1.1	2.2
Dec-12	–27.3	3.3	–7.5
Dec-13	36.4	1.1	10.9
Dec-14	–20.9	1.0	–6.5
Dec-15	–16.8	2.0	–3.4
Dec-16	76.3	10.6	27.0
Dec-17	1.0	5.9	4.2
Dec-18	26.3	2.5	10.5
5 years to Dec-18*	8.2	4.4	5.7
10 years to Dec-18*	3.6	3.9	3.8

Source: Statistical report, Table 3.3; ABS, Business Indicators, Australia, Dec 2018, Catalogue No. 5676.0.

Note: *Annualised growth rates.

[107] The profits share increased by 1.2 percentage points to a 7-year high of 28.5 per cent over the year to the December quarter 2018, while the wages share fell 0.4 percentage points to 52.2 per cent over the same period (Chart 2.8). The wages share has remained relatively stable at around 52 per cent since the middle of 2017.

Source: Statistical report, Chart 3.1; ABS, Australian National Accounts: National Income, Expenditure and Product, Dec 2018, Catalogue No. 5206.0.

[108] There were a number of submissions concerning the significance and causes of the decline in the wages share and the rise in the profits share. The Australian Government observed that the wages share has declined in many countries, falling by 1.0 percentage point between 1995 and 2014 across OECD countries. 87 However, the ACTU pointed out that the decline in the wages share in Australia has been much greater than the OECD average.88

[109] Recent analysis by the RBA discussed the impact of the housing and financial sectors on the long-term decline in the aggregate wages share. 89 The increase in capital income earned in the housing sector was found to be an important reason for the rise in the aggregate capital share due to a higher share being paid in rents.90 This includes both rents paid to landlords and imputed rents to homeowners in the form of an increased value of housing services.91 Since the 1990s, this has been fully explained by an increase in the value of land and dwellings relative to total factor income.

[110] According to the research, the financial sector has undergone significant structural change, with investment in labour-saving technology leading to high productivity growth above the all market industries average. This has translated into a decline in the wages share (and rising capital income/profits), to the extent that if the financial sector is removed from the data, the aggregate wages share has been unchanged since 1990. 92 Across other industries, the labour share has varied with the largest increases typically in the services sector. While the exercise of removing the financial sector from the all market industries average highlights that sector’s distinctive behaviour, it should not be removed from the aggregate data for the purposes of considering the appropriate level of the NMW and modern award minimum wages.

[111] Chart 2.9 shows that, since 2012–13, the business entry rate has increased as the exit rate has declined. This has resulted in a positive net entry rate since 2013–14. The net entry rate in 2017–18 was the highest since 2009–10, due to both a relatively high entry rate and relatively low exit rate.

Source: Statistical report, Chart 3.4; ABS, Counts of Australian Businesses, Including Entries and Exits, various, Catalogue No. 8165.0.

Note: Entry rates are business entries in the financial year as a proportion of total businesses operating at the start of the financial year. Exit rates are total business exits in the financial year as a proportion of total businesses operating at the start of the financial year. Net entry rates are the difference between the entry and exit rates, and represent the percentage growth in the number of businesses over the respective financial year.

[112] Over the year to June 2018, business net entry rates were positive as entry rates were higher than exit rates across all industries except for two, Agriculture, forestry and fishing and Mining, while entry and exit rates were equal in Retail trade (Table 2.4). This compares to the year to June 2017, when exit rates were higher than entry rates in 3 industries. 93

Source: Statistical report, Table 3.5; ABS, Counts of Australian Businesses, Including Entries and Exits, June 2014 to June 2018, Catalogue No. 8165.0.

Note: Entry rates are business entries in the financial year as a proportion of total businesses operating at the start of the financial year. Exit rates are total business exits in the financial year as a proportion of total businesses operating at the start of the financial year. Only data for those businesses that were able to be classified to an industry division are presented. Of all businesses that were actively trading as at June 2014 and June 2018, 1.1 per cent and 0.9 per cent, respectively, were not classified to an industry.

[113] Chart 2.10 shows the survival rates of businesses across 4-yearly intervals, beginning from 2007. The proportion of businesses that were still trading 4 years later was highest in the most recent period (June 2014 to June 2018). In fact, the highest survival rates of over 64 per cent were recorded in the 2 most recent periods and are high in historical terms, but the available statistics are only to June 2018 and may not necessarily reflect current conditions.

Source: Statistical report, Chart 3.5; ABS, Counts of Australian Businesses, Including Entries and Exits, various, Catalogue No. 8165.0.

Note: A surviving business is defined as a business which was actively trading in the first period and continued to be trading in the second period.

[114] The ACTU submitted that the Panel should take into account the legislated changes to the instant asset write off scheme, which increased from $20 000 to $25 000 in January 2019, and to $30 000 from 2 April 2019, as announced in the 2019–20 Budget, with the threshold extended from a turnover of $10 million to $50 million. 94 We do not intend to take this into account in this year’s Review since there is necessarily no data concerning the effect these changes will have on business profitability or survival rates. That is a matter which may arise for further consideration in next year’s Review.

[115] Ai Group submitted that while Australia’s global competitiveness has improved due to a sustained decline in the Australian dollar, Australia continues to have high labour costs relative to other countries, with the minimum wage being the second highest globally in 2017. 95 As we discussed in the previous Review, separate from our consideration of business profitability and survival, productivity, and the effect which minimum wage increases have had upon employment, ‘[i]t is not clear from the submissions how the broader issues of international competitiveness should bear upon our present considerations’.96 We do not consider that Ai Group’s submissions, or that of any other party, in this Review have taken this matter any further.

Small business and surveys of business performance

[116] The general object of the Act directs us to provide a balanced framework for cooperative and productive workplace relations, which promote national economic prosperity and social inclusion for all Australians by, amongst other things, acknowledging the special circumstances of small and medium-sized businesses. 97 In this context the following characteristics of small businesses and their employees may be derived from factual information contained in the submissions of the Australian Government, ACTU and Australian Chamber of Commerce and Industry (ACCI) and in the Statistical report:

•  small businesses accounted for nearly 98 per cent of all businesses, but only around 36 per cent of small businesses were employing businesses; 98

•  small businesses accounted for a greater proportion of employment compared with output in almost every industry, suggesting that they are more labour intensive (and have lower productivity) than larger businesses; 99

•  labour costs for small employing businesses accounted for around 17 per cent of total expenses in 2016–17; 100

•  small businesses accounted for around 34 per cent of all award-reliant employees in May 2018, with 35 per cent of employees in small businesses being award reliant, a higher proportion than for medium-sized businesses (20–49 employees) (32 per cent) and larger businesses (100–999 employees) (17 per cent); 101

•  small businesses had lower survival rates than larger businesses between June 2014 and June 2018; 102

•  average hourly wages for employees in small businesses were 24.6 per cent lower than for employees in larger businesses; 103

•  small businesses operate in every industry, but represent a larger proportion of output and employment in Agriculture, forestry and fishing; Rental, hiring and real estate services; and Construction, and a smaller proportion in Mining; Electricity, gas, water and waste services; and Information media and telecommunications; 104

•  small businesses were more likely to employ a higher proportion of entry-level award-reliant and minimum wage employees than larger businesses, and are likely to be ‘more sensitive to these wage increases; 105

•  average hourly wages for award-reliant employees in small businesses were 19 per cent lower than for award-reliant employees in larger businesses; 106 and

•  small businesses had higher profit margins than for all businesses over the 5 years to 2016–17. 107

[117] The Panel has previously considered relevant business surveys that the Australian Government and RBA believe to ‘be quite reliable predictors of output and employment growth.’ 108 In its May 2019 Statement on Monetary Policy, the RBA found that ‘[b]usiness employment intentions remain above average according to the NAB quarterly survey and the Bank’s liaison program’109 and that ‘[s]urvey measures of expected capital expenditure and business conditions remain at or above average levels, despite easing over the past year’.110

Inflation and wages

[118] Similar to last year, this Review is being conducted during a sustained period of low inflation and low wages growth, which is discussed in the following section.

Inflation

[119] Chart 2.11 shows that inflation remains low and has declined since the middle of 2018 across each measure. The CPI headline rate increased by 1.3 per cent and underlying inflation, measured by the trimmed mean, was 1.6 per cent over the year to the March quarter 2019. 111 The LCI for employee households increased by 1.4 per cent over the year to the March quarter 2019. Accordingly, the measures of inflation are broadly consistent. The Australian Government identified a number of explanations for the relatively low levels of inflation, one of which was slow growth in labour costs.112

Source: Statistical report, Chart 4.1; ABS, Consumer Price Index, Australia, Mar 2019, Catalogue No. 6401.0; ABS, Selected Living Cost Indexes, Australia, Mar 2019, Catalogue No. 6467.0.

Note: CPI and LCI for employee households data are expressed in original terms. Underlying inflation is calculated as the average of the trimmed mean and weighted median.

[120] The inflation data for the March quarter appear to be an aberration. The zero result for headline inflation was only the third time that quarterly headline inflation was either zero or negative in the last 10 years. The result for the trimmed mean was also relatively low and was the lowest quarterly increase since the March quarter 2016. 113 A major reason for low headline inflation in this quarter was a fall in the price of automotive fuel (–8.7 per cent).114 According to the RBA, the slowing in the housing market and government cost-of-living initiatives were ‘important factors’ for the low underlying inflation rate.115

[121] The RBA forecasts underlying inflation (trimmed mean) to increase by 1½ per cent over the year to the June quarter 2019, which would require an increase of 0.5 per cent in the June quarter 2019. 116 The forecast for underlying inflation over the year to the December quarter 2019 is 1¾ per cent and 2 per cent over the year to the June quarter 2020.117 The RBA forecast headline inflation to be 1¾ per cent over the year to the June quarter 2019 and in its May 2019 Statement on Monetary Policy notes:

‘The partial rebound in oil prices in recent months has led to an upward revision to the forecast for headline inflation in the June quarter. The 12 per cent increase in fuel prices in the June quarter to date is expected to add around 0.4 percentage points to headline inflation in the quarter. Headline inflation is expected to reach 2 per cent in the second half of 2019 and increase modestly after that.’ 118

[122] The RBA’s revised forecasts were made after the release of the March quarter inflation data, further evidence that the March quarter outcome is anomalous. While the current headline and underlying inflation rates are lower than at the same time last year, both measures are likely to rise during the period when any adjustment will operate.

[123] In the period during which any increases in the NMW and modern award minimum wages we determine will operate it is likely that underlying inflation will be between 1¾ and 2 per cent.

Wages

[124] Over the last year, wages growth has picked up only slightly and remains low, with the WPI having increased by 2.3 per cent in the 12 months to the March quarter 2019. 119 This is slightly above the corresponding figure for 2018 (2.1 per cent), and the 5-year average (2.1 per cent), but below the 10-year average (2.7 per cent). In the December quarter 2018, the average annualised wage increase (AAWI) for federal enterprise agreements approved in the quarter was 3.0 per cent. Average weekly ordinary-time earnings (AWOTE) increased by a relatively low 2.4 per cent. However, due to the decline in the inflation rate these low rates of increase in nominal wages have translated into a higher rate of increase in real wages. We discuss this further in Chapter 3. Consistent with our expectation in the 2017–18 Review decision,120 wages have grown more slowly than forecast in the 2018–19 Budget.

[125] The reasons for the persistent low rate of growth in nominal wages remain ‘somewhat puzzling’. 121 Research by Commission staff undertaken for this Review (Research Report 1/2019—Developments in wages growth) provides some background and detail on the various wage measures and summarises attempts by researchers to determine the main causes.122

[126] In its summary of research on the reasons for low wages growth, the report mainly focused on Australian studies or those covering advanced economies. Many of these studies examined whether short-term cyclical factors that are generally associated with wages growth, such as inflation, productivity and labour market spare capacity, are relevant causes. However, while some papers did find these factors to be important, it was concluded that they did not provide a full explanation for low wages growth. The focus of some other studies covered in the research report was on longer-term structural factors such as job security, advances in technology and shifts in bargaining power. Again, none of these factors could provide definitive reasons for low wages growth. 123

[127] Research by Bishop (2018) (updating previous research) on quarterly WPI data found that the average frequency of wage changes had increased between 2016 and 2018, accounting for all of the increase in aggregate wage growth, although the average size of wage increases remained similar to that in 2016. 124 The analysis found an increase in the proportion of jobs receiving wage increases of between 3 and 4 per cent and suggested that part of the reason was due to Review decisions to increase the NMW and modern award minimum wages by 3.3 per cent in 2017 and 3.5 per cent in 2018.

[128] Further analysis of quarterly WPI data was undertaken by the Australian Bureau of Statistics (ABS) (Page 2018), focusing on growth across methods of setting pay using microdata from the WPI which are not publicly available. This analysis provides some further information as to why recent increases awarded in Reviews have not translated to higher growth in the WPI. 125

[129] Page (2018) found that, since 2009, wage increases in collective agreements had made the largest contribution to wage growth, followed by wage increases in individual arrangements and modern award minimum wages. In 2017, the most recent year analysed, the contribution from modern award minimum wages was found to be 10.7 per cent, which was the largest contribution since 2006, and up from 6.9 per cent in 2014. This compared with a contribution of almost 50 per cent from collective agreements and around 40 per cent from individual arrangements. 126 This helps to explain why recent Review decisions appear not to have had a significant effect on the WPI, although it does not explain why they have had little apparent aggregate effect even in the most award-reliant industries.

[130] Chart 2.12 shows that the cumulative growth in C14 and C10 over the past decade has matched the growth in the WPI, while growth in AWOTE has been similar to that of AAWI and each have exceeded the growth in the C14 and C10 rates. Over the past two years, growth in the NMW and modern award minimum wages has exceeded the growth in previous Reviews.

Source: Statistical report, Chart 5.1; ABS, Average Weekly Earnings, Australia, Nov 2018, Catalogue No. 6302.0; ABS, Wage Price Index, Australia, Mar 2019, Catalogue No. 6345.0; Department of Job and Small Business, Trends in Federal Enterprise Bargaining, September quarter 2018, <http://employment.gov.au/trends-federal-enterprise-bargaining>; Metal, Engineering and Associated Industries Award 1998; Manufacturing and Associated Industries and Occupations Award 2010.

Note: AWOTE estimates refer to full-time adult employees. The AAWI index is calculated by first deriving a quarterly rate from the AAWI per employee for agreements approved in the quarter for all sectors. The C14 and the C10 are minimum award rates set under the Manufacturing and Associated Industries and Occupations Award 2010 and the former Metal, Engineering and Associated Industries Award 1998. AWOTE data are published half-yearly for May and November, hence, a quarterly series has been derived. AWOTE data are expressed in original terms.

[131] The RBA expects wages growth to increase gradually, but to continue to be constrained by, among other things, ‘changes in competitive dynamics such as globalisation’. 127

[132] A number of submissions raised issues concerning the relationship between the consumer wage, producer wage and labour productivity. The Australian Government submitted that growth in the real minimum wage has outpaced labour productivity growth in recent years. 128 The Australian Government also contended that the real producer wage has grown at a similar rate to the period before the mining boom, while the real consumer wage has been ‘broadly flat since 2011’ and would be expected to grow less than labour productivity during the adjustment from the mining boom.129 The Australian Government also submitted that the divergence between labour productivity and wages growth over the preceding cycle (2003–04 to 2011–12) was linked to the mining boom, but also that labour productivity and wages growth have been more closely aligned over the current incomplete cycle.130

[133] The ACTU argued that the growth in low and median wages has ‘decoupled’ from growth in productivity: 131

‘The OECD reports that Australia is one of several countries which “have been grappling not only with slow productivity growth but have also experienced a slowdown in real average wage growth relative to productivity growth, which has been reflected in a falling share of wages in GDP. At the same time, growth in low and median wages has been lagging behind average wage growth, contributing to rising wage inequality. Together, these developments have resulted in the decoupling of growth in low and median wages from growth in productivity.”’ 132

[134] In contrast, ACCI argued that weak labour productivity growth over the past 7 years has likely contributed to weak wages growth 133 and, in response to a question on notice regarding research by the RBA on the labour and capital shares of income, submitted that the divergence between consumer wages and labour productivity ‘appears to be a temporary phenomenon’ and that some of the stagnation in consumer wages over recent years is due to the unwinding of the mining boom.134 Australian Business Industrial and the NSW Business Chamber Ltd (ABI and NSWBC) contended that productivity growth has been lower than wages growth over recent years.135

[135] Chart 2.13 shows movements in the consumer wage (the average real wage deflated by the prices facing consumers—i.e. the wage as an income) and the producer wage (average real wage deflated by the prices facing firms—i.e. the wage as a cost) compared to growth in labour productivity over the last decade, and also tracks movements in the ‘consumer minimum wage’ and the ‘producer minimum wage’. It shows that after a period from about 2011 to 2016, when the producer wage grew faster than labour productivity, there has been a greater alignment between them in the last few years. It also shows that the ‘producer minimum wage’ fell behind labour productivity for the first half of the decade before coming into a rough alignment from about 2015.

[136] The volatility of the terms of trade, including its consequences for the exchange rate, have caused an unusual divergence between the prices that face producers for the goods that they sell and the prices paid by consumers for the goods that they buy. This divergence is clearly apparent in the years 2009–2016, when the consumer wage has moved quite differently from the producer wage. For much of this period, both series rose faster than labour productivity, giving a particular benefit to wage earners and increasing labour costs for firms. Conversely the real minimum wage rose by less than labour productivity for most of the same period, both as an income to wage earners and as a cost to firms. Over the past 2 years, the series have resumed a more normal relation with each other, and the rate of increase in both measures of the real value of the minimum wage has caught up with the increases in labour productivity. They are broadly aligned also with both measures of the average wage.

Source: Statistical report, Chart 5.4; ABS, Australian National Accounts: National Income, Expenditure and Product, Dec 2018, Catalogue No. 5206.0; Manufacturing and Associated Industries and Occupations Award 2010.

Note: The real consumer wage is AENA per hour deflated by the household consumption deflator; the real producer wage is AENA per hour deflated by the GDP deflator; labour productivity is GDP per hour worked; the real minimum wage is the C14 rate deflated by either the GDP deflator or the household consumption deflator.

Labour market

[137] Consistent with the approach taken in recent Review decisions, 136 we assess the state of the labour market using data concerning employment, hours worked, workforce participation and other relevant indicators.

Employment and hours worked

[138] Growth in employment and hours worked is shown in Chart 2.14. Over the year to April 2019, the number of employed persons increased by 2.5 per cent. Although lower than at the time of the last Review, employment growth is above its 5-year average (2.2 per cent). 137 Growth in hours worked rebounded to 2.8 per cent over the year to April 2019, above its 5-year average (2.0 per cent).

Source: Statistical report, Chart 6.3; ABS, Labour Force, Australia, Apr 2019, Catalogue No. 6202.0.

Note: Data are expressed in trend terms.

[139] Chart 2.15 shows the changes across full-time and part-time employment and by gender over the year to April 2019. Total employment increased by 310 500 persons over the year, with 84 per cent of the increase attributable to full-time employment, more than 10 percentage points above the proportion reported in the previous Review. 138 Total employment growth was higher among females (161 700 persons or 2.8 per cent) than males (148 800 persons or 2.2 per cent). Conversely, almost three-quarters of the growth in part-time employment came from males, with part-time employment increasing by 3.0 per cent for males compared with 0.5 per cent for females.

Source: Statistical report, Chart 6.5; ABS, Labour Force, Australia, Apr 2019, Catalogue No. 6202.0.

Note: All data are expressed in trend terms.

[140] Employment increased across the majority (13) of the 19 industries over the year to February 2019 (Table 2.5). The highest increases were in Public administration and safety (14.6 per cent) and in Mining (10.3 per cent).

[141] Employment grew in all award-reliant industries over the year to the February quarter 2019 except for Retail trade, which fell by 0.9 per cent. The yearly employment growth figures for Retail trade have proved volatile in recent years, with employment having fallen in the 12 months to February 2017 and then risen strongly the following year before falling again. As discussed later in the context of award-reliant industries, we consider the likely explanation for this to be disruption caused to the industry by technological change, international competition and the ‘growing online marketplace’, 139 and intense competitive pressures within the industry itself. Other services (6.6 per cent) and Accommodation and food services (2.4 per cent) experienced employment growth above the all industries average (2.3 per cent), while growth in Health care and social assistance (1.6 per cent) and Administrative and support services (1.4 per cent) was below the average.

Source: Statistical report, Table 6.3; ABS, Labour Force, Detailed, Quarterly, Feb 2019, Catalogue No. 6291.0.55.003.

Note: All data are expressed in trend terms.

Unemployment and underemployment

[142] As employment grew over the year to April 2019, the unemployment rate declined by 0.4 percentage points to 5.1 per cent and is below its five-year average of 5.7 per cent. 140 The underemployment rate decreased to a lesser degree, falling 0.2 percentage points to 8.3 per cent over the same period to also be below its five-year average (8.5 per cent). While the unemployment rate has declined since 2014, the underemployment rate only began to decline from 2017.

Source: Statistical report, Chart 6.1; ABS, Labour Force, Australia, Apr 2019, Catalogue No. 6202.0.

Note: Data are expressed in trend terms.

[143] The submissions made in this Review and in recent previous Reviews have raised issues concerning the extent to which the underemployment rate (as distinct from the unemployment rate) is demonstrative of spare capacity in the labour market, and whether the underemployment rate should be considered the primary indicator in this respect. In particular, Ai Group submitted that:

‘[I]n the past five or so years, the Australian labour market appears to have passed a tipping point where it [is] now the case that underemployment rather than unemployment is the more relevant indicator of labour market outcomes.’ 141

[144] These submissions have been advanced against a background in which the underemployment rate as measured by the ABS has risen relative to the unemployment rate.

[145] These submissions use a headcount measure of underemployment—that is, the number of persons who are employed part time but say they want to work more hours. As we pointed out in the 2016–17 Review decision, 142 that measure suffers from two difficulties. First, it does not take into account the number of additional hours such persons may want to work, which will usually be less than the number of hours of work sought by an unemployed person. Second, a person does not have to be actively searching for more hours of employment in order to be counted as underemployed. By contrast, to be counted as unemployed, a person must be actively seeking work.

[146] The RBA has developed an alternative volume-based measure of underutilisation which measures the sum of hours of work sought by unemployed people and additional hours of work actively sought by underemployed workers, as a share of total hours worked and actively sought. 143 This creates a more reliable indicator of spare capacity in the labour market and leads us to conclude, as the RBA has,144 that the unemployment rate is the primary indicator of labour market spare capacity.145 The addition of underemployment to unemployment does not alter the understanding of changes in the degree of spare capacity in the labour market. However there is no doubt that the underemployment rate is indicative of a degree of spare capacity, and is likely to be contributing to the suppression of wages growth.

[147] Despite the growth in employment and the decline in unemployment and underemployment, nominal wage growth as we have earlier observed remains historically low. This tends to support the proposition that the rate of full employment in Australia has fallen in recent years, meaning that at 5 per cent unemployment there may remain more spare capacity in the labour market than may previously have been the case with unemployment at that level. 146

Youth unemployment

[148] The youth unemployment rate fell 0.2 percentage points over the year to April 2019 to be 11.8 per cent, below its five-year average of 12.6 per cent. 147 In the 2017–18 Review decision, we stated that ‘[o]ver the year to April 2018, the youth unemployment rate fell by 0.3 percentage points to 12.6 per cent…’.148 However the figure of 12.6 per cent for April 2018 was subsequently revised downwards to 12.0 percent.149 Short-term fluctuations in the data may be obscuring longer-term improvements in the youth labour market that are broadly consistent with the level of improvement in the general labour market. The RBA has highlighted that youth unemployment has ‘historically been higher than for the rest of the population’.150

[149] The Panel has previously noted that youth unemployment is generally more sensitive to demand or supply shocks affecting aggregate unemployment. 151 This is a sentiment that has been reiterated by the RBA recently:

‘The youth labour market is distinct from the overall labour market because younger workers are most likely to be new entrants transitioning into the workforce with minimal skills and experience. Accordingly, they are more exposed to cyclical swings in the economy than other parts of the labour market. They may also be more exposed to structural changes in the labour market.’ 152

[150] Labour dynamics for younger people are also distinct and characterised by lower labour force participation, and higher levels of part-time work (associated with higher shares of participation in full-time study since the 1980s). 153

Long-term unemployment

[151] Long-term unemployment fell by 10.1 per cent over the year to April 2019, and while lower compared with last year’s Review, 154 remains relatively high as a proportion of all unemployed persons (22.7 per cent) (Table 2.6).

Source: Statistical report, Table 6.10; ABS, Labour Force, Australia, Detailed—Electronic Delivery, Apr 2019, Catalogue No. 6291.0.55.001.

Note: Data are trend estimates. The long-term unemployed refers to the number of persons unemployed for 52 weeks or more. The long-term unemployment ratio refers to the number of long-term unemployed persons expressed as a percentage of the total unemployed population. The percentage change is calculated in relation to the corresponding month in the previous year.

Workforce participation

[152] Labour force participation was strong across 2018, remaining steady at a record high of 65.6 per cent for most of the year, and increasing to 65.7 per cent in April 2019. 155 Over the year to April 2019, the employment to population ratio among the adult working-age population (aged 20–64 years) increased by 0.8 percentage points. In December 2018 the employment to working-age population ratio reached a new high of 77.5 per cent.156

[153] Chart 2.17 shows the participation rate and employment to population ratio in original terms, and adjusted for the ageing of the population. The chart shows that, while both have increased, if there were no change in the age distribution of the working age population since April 2014, the participation rate and employment to population ratio would be 1.0 percentage point higher in April 2019. This is the same impact as reported in the last Review. 157

Source: Statistical report, Chart 6.2; ABS, Labour Force, Australia, Detailed—Electronic Delivery, Apr 2019, Catalogue No. 6291.0.55.001.

Note: Data are expressed in original terms with only the April figure of each year shown. The age-adjusted participation rate is calculated using 10-year age groups, with the proportion of the population held fixed at January 2014 levels. It includes all persons aged 15 and over.

[154] For the purpose of understanding the strength of the labour market, the age-adjusted participation rate and employment to population ratio measures are preferred over those measures without an age adjustment because they capture the changes in the age composition of the population. As the chart shows, both employment and participation have risen since 2017, which is a very positive sign of strength in the labour market.

[155] Rather than do a full adjustment for the different propensities of each age group to be in the labour force or employed, a simplified adjustment to take account of the aging of the population has been used in previous decisions of the Panel in the form of the adult working-age population (20–64 years). Data using this simplified measure are available in the Statistical report. 158 They show that over the year to April 2019:

•  the female working-age participation rate increased by 0.8 percentage points;

•  the male working-age participation rate increased by 0.3 percentage points;

•  the proportion of persons aged 20–64 years working full-time rose by 0.6 percentage points;

•  the proportion of persons aged 20–64 years working part-time increased by 0.2 percentage points; and

•  the proportion of persons aged 20–64 years who are employed increased by 0.8 percentage points.

[156] Chart 2.18 shows the employment to working-age population (this time defined as 15–64 years) ratio in Australia compared to the OECD average and G7 nations. The ratio in Australia is significantly higher than the OECD average. It is also significantly higher than the United States of America (US), which currently has a lower unemployment rate than Australia and has lower minimum wages. 159

Source: Statistical report, Chart 6.8; OECD (2019), Employment rate, <https://data.oecd.org/emp/employment-rate.htm#indicator-chart>.

[157] The broad-based growth in employment, particularly full-time employment, the lower unemployment rate, and the historically high working-age employment to population ratio and participation rate are all indicative of a strong labour market. No party identified any data which demonstrated adverse employment effects arising from the previous 2 Review decisions, each of which resulted in real wage increases for award- and NMW-reliant employees. 160

Labour market transitions

[158] The Australian Government again presented updated data from the Household, Income and Labour Dynamics in Australia (HILDA) survey on the duration in low-paid employment, for 2017 (the latest release). First, the Australian Government stated that ‘[o]ver a third (36 per cent) of people who enter the workforce do so by taking a low-paid job … with 41 per cent of workers aged under 25, and 41 per cent of workers with year 12 qualifications and below’. 161

[159] The updated data support the Panel’s conclusion in previous Reviews that about half of low-paid workers 162 spend less than a year in low-paid work before moving to higher-paid work. The remainder either spend more than one year in low-paid work, or move from a low-paid job into unemployment or leave the labour force (Tables 2.7 and 2.8).163

Source: Australian Government submission, 15 March 2019 at p. 62, Table 7.1; HILDA Survey, release 17 (December 2018), balanced panels waves 1 to 17 with longitudinal weights.

Note: Data are based on flows into low-paid work, not the number of people in low-paid work at a point in time. Numbers are mutually exclusive. Low-paid is defined as earning less than two-thirds of the median hourly wage.

Source: Australian Government submission, 15 March 2019 at p. 62, Table 7.2; HILDA Survey, release 17 (December 2018), wave 17.

Note: Those remaining in low pay for 5 years or more are not shown due to a small sample size.

[160] Relying on the data presented in the 2017–18 Review, 164 ACCI submitted that ‘the Panel should take into account that only a very small proportion of the population rely on the minimum wage and those that do use it as a stepping-stone to higher paid work’.165

[161] In previous Reviews, the Panel has agreed that while low-paid employment is a pathway for many into higher-paid work, there are also many instances where this is not the case. 166 The Panel has previously stated, ‘[w]e cannot be indifferent to the standard of living of low-paid workers just because many do not stay in that situation for long periods.’167 Nothing has been put to us in the current Review which would cause us to revisit this position.

Award-reliant industries

[162] Based on the new data from the Survey of Employee Earnings and Hours (EEH) for 2018, the 5 most award-reliant industries are presented in Table 2.9 with a number of key indicators. We note that Health care and social assistance is now the 4^th most award-reliant industry, overtaking Retail trade. 168 The Panel continues to use these data to assist with an understanding of the overall conditions of these industries169 and they are also of utility in assessing the effects of previous Review decisions.

Source: Statistical report, Table 7.2; ABS, Australian National Accounts: National Income, Expenditure and Product, Dec 2018, Catalogue No. 5206.0; ABS, Business Indicators, Australia, Dec 2018, Catalogue No. 5676.0; ABS, Counts of Australian Businesses, including Entries and Exits, Jun 2014 to Jun 2018, Catalogue No. 8165.0; ABS, Employee Earnings and Hours, Australia, May 2018, Catalogue No. 6306.0; ABS, Labour Force, Australia, Detailed, Quarterly, Feb 2019, Catalogue No. 6291.0.55.003; ABS, Wage Price Index, Australia, Mar 2019, Catalogue No. 6345.0; Department of Jobs and Small Business, Trends in Federal Enterprise Bargaining, December quarter 2018, <http://employment.gov.au/trends-federal-enterprise-bargaining>.

Note: (a) All industries excludes Agriculture, forestry and fishing; (b) All industries excludes Education and training, Health care and social assistance and some subdivisions of Finance and insurance services. n/a = not available. The award-reliant industries selected are the five industries with the highest proportion of employees reliant on award rates of pay according to the EEH 2018 survey. The WPI and actual hours worked data are expressed in original terms. Employment data are expressed in trend terms. Entry rates are business entries in the financial year as a proportion of total businesses operating at the start of the financial year. Exit rates are total business exits in the financial year as a proportion of total businesses operating at the start of the financial year.

[163] In summary, Table 2.9 shows generally positive outcomes for these industries with the exception of Retail trade. The main conclusions we draw, while remaining aware that some of the indicators are quite volatile when disaggregated to this industry level, are the following:

•  Output growth in 3 of the 5 award-reliant industries outpaced the total industry average growth, particularly Health care and social assistance.

•  Annual growth in gross operating profits was almost twice the all industries average in Accommodation and food services and Other services, and below average but still positive in Administrative and support services and Retail trade.

•  Business entry rates exceeded exit rates for all of these industries except for Retail trade, (where entry and exit rates were the same) as they have for the whole economy.

•  Annual growth in the WPI for Health care and social assistance and Other services was above the all industries average. WPI growth for Accommodation and food services was at the average, while WPI growth for Administrative and support services and Retail trade was below the average.

•  Wage growth under new collective agreements in 2 of the 5 award-reliant industries was above the all industries average, with Retail trade experiencing the strongest growth at 3.3 per cent.

•  Employment growth was mixed with growth in employment and hours worked stronger in Other services and Accommodation and food services (both above average growth), while growth was below average but still positive in Health care and social assistance, but in Retail trade both employment and hours worked fell over the year.

[164] With the exception of wages growth in approved collective agreements, Retail trade stands out as having weakened over the year, with all remaining indicators below the all industries average. A number of submissions expressed concerns around the challenges retailers are facing, particularly increased competition (driven by technology and international competitors) putting downward pressure on prices. 170 The Australian Government submitted that the fall in Retail trade employment reflects ‘heightened competitive pressure’, with an ongoing shift towards online alternatives.171 These submissions appear to provide the most plausible explanation for the poorer outcomes in the industry, although it must be borne in mind that employment still remains above the level recorded two years ago.

[165] The Panel asked retail employer organisations to explain the low industry WPI increase of 2 per cent (over the year to the December quarter 2018) compared with the 3.5 per cent increase in modern award minimum wages in 2018. 172 The National Retail Association (NRA) submitted that the scope of what is considered to be the retail industry varies as between the wide definition used by the ABS for the purposes of the WPI and the narrower coverage of the General Retail Industry Award 2010, with significant portions of the Retail trade covered by other modern awards.173 However, we do not consider that this explains anything, since modern award minimum wages in all awards were increased by the same percentage amount as a result of the 2017–18 Review decision. The Australian Retailers Association (ARA) and ACCI submitted that the majority of employees in this industry (around 70 per cent) have their pay set by other methods that may not have increased by as much.174 This would appear to assume a degree of absorption of modern award minimum wage increases where it is possible to do so. The NRA explicitly submitted that absorption of the increases in the case of employees on above-award salary arrangements was occurring.175

[166] Absorption of modern award minimum wage increases into above-award remuneration is no doubt part of the explanation for the low WPI growth rate in the Retail trade industry, but it would be expected that if this occurred over a period of time, the level of award reliance in the industry would rise. However, it has fallen. We cannot ultimately identify a satisfactory explanation, and it may be that non-compliance is a contributing factor. We note that new enterprise agreements in the December quarter of 2018 have provided for a significantly higher level of wage increases (3.3 per cent) than the WPI for the industry. We observe that this may well have been influenced by a number of new enterprise agreements being approved applying to major employers in the Retail sector, after a period of hiatus. This may also be an indirect consequence of the increase to base modern award rates in last year’s Review decision, and it should be the case that these higher increases will lift the WPI for the industry by the time of next year’s Review.

[167] No detriment resulting from the wage increases awarded in the last 2 Review decisions can be identified in the data for the 5 most award-reliant industries. No party provided any relevant data or evidence to the contrary. Apart from the specific circumstances which appear to be applicable to Retail trade, these industries are travelling reasonably well in terms of profit growth, business survival and employment growth. Wage growth in all these industries remains below 3 per cent despite last year’s increase. The data tends to confirm our view as to the scope to adjust minimum wages without occasioning adverse labour market consequences.

Apprenticeships and traineeships

[168] As they have in previous years, ACCI, ACTU, Australian Council of Social Service (ACOSS), ARA, and Housing Industry Association (HIA) all outlined the ongoing reductions in commencement rates and rising cancellation/withdrawal rates for apprenticeships and traineeships. 176 HIA submitted that this trend is particularly concerning in the context of residential construction, citing the ongoing skill mismatches between available tradespersons, and business requirements.177 ACCI and ARA again attributed declines in commencements to demand-side factors (those that have reduced employer demand for apprentices), specifically identifying increases in modern award minimum wages and the cumulative effect of consecutive decisions.178

[169] Conversely, ACOSS again attributed ongoing declines in commencements mainly to supply-side factors (those that have reduced the number of people wanting to undertake apprenticeships), such as age-related cultural issues, lower wage rates, and the long duration of apprenticeships. 179 In addition, ACOSS submitted that private-sector ‘free riding’, associated with firms poaching tradespeople rather than investing in dedicated training, has also contributed to the decline.180 HIA described the balance between managing supply and demand-side factors and stated that while ‘higher rates of pay can act as an incentive to undertake and remain in apprenticeships’, they need to be tempered against ‘the cost of employing such labour.’181 The ACTU submitted that it anticipates further declines in commencements and completions and that ‘[s]uch a decline is not able to be addressed through some different treatment of apprentice or trainee wages’.182

[170] In the previous Review, 183 the Panel considered and reiterated the findings of the Commission’s Research Report 3/2017—Factors affecting apprenticeships and traineeships in February 2017,184 which investigated both supply and demand-side effects and found that:

‘…the decline in government subsidies clearly contributed to the decline in commencement rates, whilst the decision made by the Full Bench in the 2013 Modern Awards Review (Apprentices decision) to increase apprentice wages may have played a role, but it seems that any effect appears minor.’ 185

[171] In the absence of further evidence advanced in this Review, the Panel retains its conclusion that the removal of government subsidies clearly contributed to a decline in commencement rates and that any effect of the Apprentices decision was only minor. 186

[172] We do not consider that the available data supports the proposition that the cumulative effect of recent Review decisions has caused a reduction in the number of apprenticeships and traineeships. Table 2.10 shows the most recent data (to the September quarter 2018) published by the National Centre for Vocational Education Research as to apprenticeships and traineeships numbers. We would expect that any demand-side effect of excessive regulated minimum wage levels would most clearly be evident in the number of commencements. However two things are clear. First, a trend towards reduced commencement rates is evident over the whole period analysed, pre-dating the two most recent Review decisions (2016–17 and 2017–18) which awarded higher wage rate increases than in a number of previous years. Second, the rate of reduction in commencements significantly slowed rather than accelerated from September 2017 to September 2018. The number of persons in-training has also remained fairly stable over that period. Completions have continued to drop significantly, but this points to problems other than a demand-side problem with excessive wages, and may rather be indicative of a supply-side problem with insufficient wages, as the ACTU suggests.

Source: Research reference list; National Centre for Vocational and Education Research (NCVER) (2019), Apprentices and trainees 2018: September quarter – Australia, NCVER, Adelaide.

Economic outlook

[173] The most recent economic forecasts have reduced estimates of growth over the next 2 years, but solid growth is still forecast.

Global forecasts

[174] The International Monetary Fund (IMF) global growth forecasts, presented in Table 2.11, project GDP growth in Australia to be lower in 2019 than in 2018, but to rise again in 2020. These patterns are similar for world GDP. While the projections for GDP growth in Australia are below world GDP growth, they are higher than growth in advanced economies.

Source: Statistical report, Table 13.2; IMF (2019), World Economic Outlook, April.

Note: Year-on-year percentage changes shown. World and domestic economy growth rates are calculated using GDP weights based on purchasing power parity (PPP).

[175] The 2019–20 Budget stated that there was some moderation in global growth in the second half of 2018. 187 Table 2.12 presents Australian Government forecasts of world GDP growth and for Australia’s major trading partners from the 2019–20 Budget. The Budget states that the Australian economy is expected to continue to benefit from growth in major trading partners, particularly in the Asian region.188 Over the forecast period, Australia’s major trading partners are expected to grow at 4 per cent, outpacing global GDP growth forecasts of 3½ per cent over the forecast period.189

[176] A number of international factors are anticipated to pose downside risks to global growth, with economic and geopolitical uncertainty weighing on global confidence. 190 Risks related to US-China trade tensions, lower levels of global trade and uncertainty around the trade outlook have weighed on confidence, new export orders, and investment intentions.191 Other risks include tightening global monetary policy, the pace of natural disaster recovery in Japan, Italy’s financial sector, and Brexit (although Australia’s trade is focussed more on Asia than on Europe).192

Source: Statistical report, Table 13.1; Australian Government, Budget Paper No. 1: Budget Strategy and Outlook 2019–20, Canberra, p. 2-7.

Note: World growth rates are calculated using GDP weights based on PPP, while growth rates for major trading partners are calculated using goods export trade weights. These growth rates are estimates in 2018 rather than outcomes.

Australian forecasts

[177] The 2019–20 Budget states that the ‘fundamentals of the Australian economy remain sound’,¹⁹³ however, growth forecasts for 2018–19 and 2019–20 have been revised down since the 2018–19 Budget. Growth in real GDP for 2018–19 is now forecast to be 2¼ per cent, with the economy expected to continue to grow at around its estimated potential rate of 2¾ per cent over 2019–20 and 2020–21. Growth is expected to be supported by contributions from household consumption, non-mining business investment, public final demand, and exports. Across the forward estimates, mining investment is expected to make positive contributions to growth (the first time in around 7 years), 194 while dwelling investment is forecast to detract from growth.195

[178] Growth in household consumption ‘is expected to pick up over the forecast period,’ supported by ongoing employment growth and increasing wage growth. 196 However, while consumption growth is expected to increase by 2¼ per cent in 2018–19 and then to 3 per cent by 2020–21, these forecasts have been revised downward since the 2018–19 Budget, reflecting weaker consumption growth in the latter part of 2018. The 2019–20 Budget states:

‘A more subdued outlook for household income, or a further tightening in credit conditions could constrain household spending amid high levels of household debt. By contrast, faster-than-expected wage or employment growth could lead to household consumption being stronger than forecast.’ 197

[179] Business investment is expected to fluctuate throughout the forecast period, largely due to volatility in mining investment. Non-mining investment is expected to be solid through the forecast period, 198 while mining investment is expected to detract from growth in 2018–19, before making a positive contribution to growth across the forward estimates.

[180] The unemployment rate is expected to remain steady over the forecast period at 5 per cent while employment growth is forecast to gradually slow, reflecting sustained labour market tightening since 2014. Labour force participation is expected to remain near record-high levels of 65½ per cent over the forecast period.

[181] Prices and wages are expected to gradually pick-up as output increases and labour market spare capacity reduces. 199 Wages growth is forecast to increase only by 2½ per cent in 2018–19, before rising to 2¾ per cent in 2019-20 and 3¼ per cent by 2020–21.200 Consumer prices are also expected to gradually grow over the forecast period, only 1½ per cent in 2018–19, picking up to 2½ per cent by 2020–21.

[182] Net exports are forecast to grow modestly through 2018–19 and 2019–20 and are predicted to detract from growth in 2020–21 as export growth slows due to mining. 201 The Budget notes that rural exports are forecast to detract from GDP growth in 2018–19 due to drought conditions, but are expected to rebound and contribute to growth in 2019–20 based on an assumed return to average seasonal conditions.202

[183] Domestic downside risks include the extent and impact of declining house prices on investment and consumption, subdued income growth, tightening credit conditions, commodity prices, and uncertainty around the recovery of rural exports. 203

[184] Table 2.13 presents Australian Treasury forecasts of the domestic economy.

Source: Statistical report, Table 13.3; Australian Government, Budget Paper No. 1: Budget Strategy and Outlook 2019–20, Canberra, p. 2-4.

Note: The forecasts for the domestic economy are based on several technical assumptions. The exchange rate is assumed to remain around its recent average level — a trade-weighted index of around 61 and a US$ exchange rate of around 71 US cents. Interest rates are assumed to move broadly in line with market expectations. World oil prices (Malaysian Tapis) are assumed to remain around US$67 per barrel. Population growth is assumed to average around 1.7 per cent per year over the forecast period.

(a) Percentage change on preceding year unless otherwise indicated.

(b) Calculated using original data unless otherwise indicated.

(c) Excluding second-hand asset sales from the public sector to the private sector.

(d) Percentage point contribution to growth in GDP.

(e) Through-the-year growth rate to the June quarter.

(f) Seasonally adjusted, through-the-year growth rate to the June quarter.

(g) Seasonally adjusted rate for the June quarter.

(h) The forecasts are underpinned by price assumptions for key commodities: Iron ore spot price falling over the year to reach US$55 per tonne free-on-board (FOB) by the end of the March quarter 2020; metallurgical coal spot price falling over the year to reach US$150 per tonne FOB by the end of the March quarter 2020; and the thermal coal spot price remaining at US$91 per tonne FOB.

[185] In the May 2019 Statement on Monetary Policy, the RBA revised downwards its growth forecasts for 2019 as a result of the lower rate of growth in the last half of 2018. Its current forecasts for growth, unemployment and inflation are contained in Table 2.14.

Source: RBA (2019), Statement on Monetary Policy, May, p. 71.

Note: Percentage changes are for the year-ended. *Average rate in the quarter. Technical assumptions include A$ at US$0.70, Trade Weighted Index at 60, Brent crude oil price at US$68 per barrel. Shaded regions are historical data.

[186] The RBA considers that subdued growth in household income and the adjustment to the housing market are affecting consumer spending and residential construction, and this has led to a downward revised outlook for household consumption spending and dwelling activity. Stronger growth in exports and work on new mining investment projects are anticipated to support growth. 204 Monetary policy is expected to remain accommodative for some time, and this has caused easing in market interest rates internationally and domestically.205 Leading indicators of labour demand suggest that employment will grow at the same rate as the working-age population over the next 6 months and then pick up somewhat as growth increases, resulting in the unemployment rate remaining unchanged until 2021.206 A moderate increase in wages growth is forecast,207 but there remains considerable uncertainty about household income growth and its effect on household consumption, which accounts for almost three-fifths of GDP.208 The CPI is expected to rise by 2 per cent per annum after its low growth in early 2019 (discussed earlier at [119]).

Employment effects of minimum wage increases

[187] In the 2017–18 Review decision we discussed significant new Australian research conducted by Bishop of the RBA, amongst other largely international research contributions. Bishop (2018) 209 used unpublished job-level data from a survey of firms undertaken for the construction of the ABS WPI survey to determine the employment effects of minimum wage increases between 1998 and 2008.210 This research found no adverse employment effects from those minimum wage increases. There has been no further Australian research since the last Review.

[188] The ACTU’s submission in this Review contained an extensive discussion of, and commentary on, recent academic research on the impacts of increases in minimum wages. Much of the research that the ACTU draws on relates to the US, but there is also new research for the United Kingdom (UK), together with a small number of studies for some European countries. 211 Additional recent studies for the US are included in the Research reference list for this Review.212

[189] The contending positions on the role of legislated minimum wages within the academic community were summarised as follows:

‘Arguments for and against minimum wage increases often invoke significantly different narratives regarding the low wage labor market. To opponents, low-wage jobs are temporary phenomena, offered to workers with little experience or skill who rapidly attain both as they spend time on the job. As their productivity increases, their wages increase, naturally removing them from the ranks of least-paid workers. Minimum wage increases threaten to disrupt this pattern by eliminating opportunities for the least-skilled and experienced. To proponents, low-wage jobs are not avenues of advancement but dead ends, where workers may linger for years if not decades. Whether held back by the lack of training and advancement opportunities, immigration status, racial or gender discrimination, or monopsony power, this narrative suggests that workers are consigned to penury in the absence of labor market intervention.’  213

[190] Undoubtedly both phenomena are at work, and a major research task is to identify the relative size of each, and the extent to which increasing minimum wages does increase the earnings of the low paid and/or ‘eliminate opportunities for the least skilled’. This is an empirical task, and one that is very challenging with no single piece of research being definitive. It is especially difficult to do this research in Australia, as fully discussed in the material prepared for the 2017–18 Review. 214 For this reason, we remain interested in new research, including from overseas, that overall will extend, support and/or challenge our understanding of the effects of increases in minimum wages. In evaluating the usefulness of international research, we remain conscious of the distinctiveness of the Australian minimum wage-setting arrangements, and that the Australian NMW and modern award minimum wages (as minimum wages in their own right) are high by international standards.

[191] Of particular interest for this Review is the broad conclusion that the extensive and increasingly sophisticated recent research continues to find, first, that increases in minimum wages which have been the subject of examination do increase the earnings of the low paid and second, that they do not, for the most part, cause job losses or increase unemployment.

[192] Possible exceptions to this second point are that inexperienced workers might find it harder to get a job and low-paid women who work part time might see some job loss; as might those who work in jobs that are more easily automated or produce internationally tradeable goods. To date, the evidence for the last point is not strong. In Cengiz et al. (2019), the estimates for the employment impact of minimum wage increases on the tradeable sector in the US are described as ‘imprecise’. 215 The authors go on to say, in respect of such increases, that ‘the effect of the minimum wage is close to zero in the non-tradable sectors (such as restaurants or retail), which employ most minimum wage workers in the U.S. today’.216

[193] The recent US research uses new analytical techniques and, perhaps more importantly, new and more suitable data. These include data that make it possible to track affected workers over time. This is a key capacity that has been available to UK researchers for some years. Both the improved techniques and the more suitable data increase the robustness of the results.

[194] There is new research also for the UK, triggered in part by the move to a higher National Living Wage (NLW) for workers aged 25 years and over. The increases in the NLW that are the subject of the investigations are 7.5 per cent in 2016 and 4.2 per cent in 2017. We have previously expressed the view that the UK experience has greater relevance for Australia than does that of the US. 217 Key propositions that may be drawn from this research are as follows:

•  In its summary of this research, the UK Low Pay Commission stated that ‘Aitken, Dolton and Riley (2018) concluded that the NLW had been a significant intervention in the labour market, raising the wages of the lowest paid, but that it had so far had little adverse impact on overall employment retention. However, consistent with previous evidence such as Dickens, Riley and Wilkinson (2015), they also found some evidence of small adverse effects on the employment opportunities of women working part-time.’ 218

•  A case study approach was used to examine in detail the impact of the increases in the NLW on the UK care homes sector. The study found that the introduction of the NLW on 1 April 2016 had no effect on employment or hours worked in that sector. There was a considerable positive spill over effect on the wages paid to workers who were under age 25 years. The study did find that the introduction of the NLW led to a reduction in the quality of care in the sector. 219

[195] The recent research, and that discussed in previous Reviews, is consistent with the position adopted by the Panel that modest and regular minimum wage increases do not result in disemployment effects or inhibit workforce participation. 220 The assessment of what constitutes a ‘modest’ increase has developed over time and the research and available evidence confirms that the increases granted by the Panel in the previous 2 Reviews, and as we have adopted in this Review, fall within that category. Further, the research is supportive of the notion that increases in minimum wages do increase the earnings of the low paid.

3. Relative living standards and the needs of the low paid

[196] Both the minimum wages objective and the modern awards objective require us to take into account ‘relative living standards and the needs of the low paid’ when setting minimum rates. These are different, but related, concepts. 221

[197] The assessment of relative living standards requires a comparison of the living standards of workers reliant on the NMW and modern award minimum wages with those of other groups, in particular other workers, especially non-managerial workers. We pay particular attention to changes in the earnings of NMW and award-reliant workers compared to changes in measures of average and median earnings more generally. The degree of dispersion (or inequality) around these measures is also relevant.

[198] The living standards of employees on the NMW and award-reliant employees are affected by their wage rate, the hours they work, tax-transfer payments and the circumstances of the households in which they live. The net effect of these factors is captured in the notion of equivalised household disposable income.

[199] A threshold of two-thirds of median (adult) full-time ordinary earnings is the benchmark we use to identify who is ‘low paid’ within the meaning of ss 134(1)(a) and 284(1)(c).

[200] The assessment of the needs of the low paid requires an examination of the extent to which low-paid workers are able to purchase the essentials for a decent standard of living and to engage in community life, assessed in the context of contemporary norms. The risk of poverty is also relevant in addressing the needs of the low paid. We accept, as we have in previous Review decisions, that if the low paid are forced to live in poverty then their needs are not being met. We also accept that those in full-time employment can reasonably expect a standard of living that exceeds poverty levels.

[201] There is no single contemporary measure of the needs of the low paid. We use a variety of measures, including budget standards, comparisons of hypothetical low-wage families with customary measures of poverty, both before and after taking account of the tax-transfer system, and survey evidence of financial stress and material deprivation among low-paid households. The Panel has generally relied on poverty lines that are based on median equivalised household disposable income, using a 60 per cent threshold on the basis that those in full-time employment can reasonably expect some margin above a harsher measure of poverty. 222

[202] The single-adult household provides the starting point for our assessment of relative living standards and needs. Such a worker receives no assistance from the transfer system, indeed their disposable income is reduced by the operation of the tax system. But we also accept that we must take into account the needs of the low paid, without limitation. Accordingly, we also give consideration to the needs of other types of families, including single-income families with dependants.

[203] The effect of taxes and transfers on disposable incomes is relevant to the needs of the low paid and relative living standards, both in terms of the impact of specific changes to the tax-transfer system and in assessing broader information in relation to measures of the relative income of the low paid. 223 The tax-transfer changes which have taken effect in the current Review period have, broadly speaking, provided a benefit to low-paid households. We later deal with these changes in more detail and note here that they are a moderating factor on our assessment of the appropriate level of increase to the NMW and modern award minimum wages arising from this Review but that it is not appropriate to apply a direct, quantifiable, discount to the increase we would have awarded in the absence of such changes.

NMW, award-reliant and low-paid employees

[204] It is important to distinguish between employees on the NMW, award-reliant employees (that is, employees paid exactly an award rate) and low-paid employees generally. To some extent, these categories overlap. NMW employees and some (perhaps many, but not all) award-reliant employees are low paid. 224 According to data from the ABS EEH provided by the Australian Government, 660 700 employees were award reliant and low paid (which would comprise around half of the low paid and about 30 per cent of award-reliant employees). 225

[205] A threshold of two-thirds of median earnings provides ‘a suitable and operational benchmark for identifying who is low paid’. 226  There is, however, more than one source of data on the distribution of earnings, from which a median (and two-thirds of that median) can be derived. The two main ABS surveys are the Characteristics of Employment Survey227 (the CoE) and the ABS Survey of Employee Earnings and Hours (EEH). The most recent data for median earnings from the CoE survey is for August 2018 and is $1330.00 per week. Data on median earnings from the EEH Survey is only available for May 2018 and is $1460.00 per week. Data from the CoE survey is the usual source for calculating median earnings as it is released more frequently (annually compared with the EEH survey which is biennial) and thus provides a more complete series. The CoE survey also has a wider scope of employees, including those employed in the Agriculture, forestry and fishing industry as well as employees of private households, both of which are excluded in the EEH survey. This measure of median earnings is also the most frequently cited in the submissions.228 Based on this measure, two-thirds of median earnings was equal to $886.67 per week, or $23.33 per hour,229 in 2018.

[206] The main source of information on award-reliant employees is obtained from the EEH. According to the latest EEH, there were about 2.2 million award-reliant employees in 2018, or 21.0 per cent of all employees. 230 Using data from this survey, the Australian Government estimated that around 1.7 per cent (or around 180 200 employees) were paid the adult rate of the NMW.231 Some of these employees will not be affected by this decision as they are not in the federal industrial relations system, such as some state public sector employees and private sector employees in non-incorporated businesses in Western Australia.232

[207] In addition to these employees, the Panel’s decision is also likely to affect those paid close to the NMW rate or a modern award minimum wage rate, workers whose pay is set by a collective agreement which is linked to the outcomes of the Review, as well as workers whose pay is set by individual arrangements which are referenced to an award rate. 233 According to the RBA, the wages of a further one-fifth of employees are linked to changes in the NMW or modern award minimum wages in their collective agreements or individual arrangements.234 Research undertaken for the Commission found that 21 per cent of employees in non-public sector organisations were on ‘over-award’ arrangements—that is, where the wage is set, influenced or guided in some way by the applicable award rate.235

[208] Data from the EEH also provided by the Australian Government found that 1.25 million employees (11.8 per cent of all employees) were low paid. 236 This is based on employees earning less than $20.27 per hour as at May 2018.237 The Australian Government submission also provided data on the characteristics of low-paid workers using data from the HILDA survey for 2017. Based on this survey, low-paid adult employees were defined as having hourly earnings below $19.53 with low-pay thresholds for junior employees adjusted by the relevant minimum wage rate.238 The data provide the following profile of low-paid workers:

•  just over half were female;

•  they tended to be concentrated among younger workers (42.3 per cent aged 15–24 years and over half were aged under 30 years);

•  over half (59.5 per cent) were single;

•  most (71.3 per cent) did not have children; 239 and

•  almost two-thirds (63.6 per cent) were casual employees. 240

[209] Low-paid employees were also more likely than all employees to be working part time (51.4 per cent compared with 28.5 per cent of higher-paid employees) and were more likely to work in a small business (56.4 per cent compared to 29.8 per cent of higher-paid employees). 241

[210] Data submitted by the ACTU 242 on the award reliant suggests that the characteristics of low-paid and award-reliant employees are similar. Further, the Commission’s research found that, overall, NMW earners and low-paid adult employees share similar characteristics.243

Award-reliant and low-paid households in the income distribution

[211] The Australian Government provided an update on the distribution of low-paid workers across household incomes using data from the HILDA survey for 2017. This included data on the distribution of low-paid workers across all households (including jobless and retiree households) and across households with at least one employee. 244

[212] The chart provided by the Australian Government (Chart 3.1) shows that across the distribution of employee households, a higher proportion of low-paid employees are located in the lowest three deciles. These are likely to include low-paid workers whose wages are affected by the NMW or modern award minimum wages.

Source: Australian Government submission, 15 March 2019 at para. 45.

[213] Research undertaken for the Commission using data from the 2015 HILDA survey (Research Report 1/2017—Award-reliant workers in the household income distribution) finds that around 70 per cent of award-reliant employees are located in the bottom half of the household income distribution of employee households. 245

[214] On the basis of this evidence, we remain of the view that low-paid workers, whose wages are likely to be affected by the NMW or modern award minimum wages, are disproportionately located in the lower deciles of equivalised household disposable income of employee households.

Real wages

[215] The real value of the NMW and modern award minimum wages rose by 6.2 per cent over the 5 years to the December quarter 2018. 246 Over this period, uniform percentage increases from Review decisions have resulted in the real value of the NMW and all modern award minimum wages increasing at the same rate. Wages as measured by the WPI grew by 2.3 per cent in real terms over the same period.247

Household disposable income and the tax-transfer system

[216] Changes in the tax-transfer system, whether providing a benefit or a detriment to low-paid households, are relevant to our consideration of relative living standards and the needs of the low paid because of their effect on disposable incomes. In many cases, the net effect of the tax-transfer system is to increase the disposable income of households that are reliant on low wages. But the impact of taxation and the means testing of many transfer payments means that increases in the NMW and modern award minimum wages are not fully reflected in household disposable income.

[217] Changes to the tax-transfer system were discussed in previous Reviews and have impacted disposable incomes in 2018–19. There have been additional changes to the tax-transfer system effective from July 2018 that have impacted disposable incomes and these have been taken into account for this Review.

[218] In July 2018, the Child Care Subsidy replaced the previous Child Care Benefit and the Child Care Rebate with a single means-tested subsidy. 248 In 2018–19, families earning up to $66 958 per year are eligible to receive the subsidy of 85 per cent. The amount of subsidised hours will be determined by the activity level of both parents in recognised activities. An annual subsidy cap of $10 190 per child per financial year applies to families whose annual income is between $186 958 and $351 248 (higher than the current cap of $7 613 per child per year under the Child Care Rebate). Families earning less than $186 958 per year have no annual cap.249

[219] In last year’s Review, we invited parties to provide further material on the financial impact of the changes to child care assistance on low-wage families to this Review. 250

[220] The Australian Government’s submission for this Review did not address the child care changes or their impact. The Federal opposition repeated the claim that the changes would leave 279,000 families worse off. 251 ACCI submitted that the changes ‘had a positive impact’ on NMW-reliant employees with children.252

[221] The ACTU commented that the Child Care Subsidy is not ‘specifically targeted at the low paid’ and would ‘amount to little change for those paid at minimum wages.’ 253 ACBC submitted that the financial impact on low-paid working parents would depend to ‘a large extent on the actual costs of child care compared to the amounts set out in these caps.’254

[222] As we noted in last year’s decision, it is challenging to evaluate the impact of changes to child care assistance on the low paid and on the material before us, as we are unable to come to a conclusive view on the impact of these changes on low-paid employees. 255

[223] The 2018–19 Budget proposed changes to income tax rates that have since been legislated. These were not taken into account in last year’s Review as the earliest of the proposed reforms were payable at the end of the 2018–19 financial year. 256 We propose to take them into account in the present Review. We note that other changes to the tax-transfer system have been proposed for 2019–20 and beyond. Consistent with how the Panel has previously dealt with proposed changes to legislation,257 we have not taken these measures into account in this Review.

[224] An information note on Changes to the tax-transfer system was published as additional material by Commission staff. 258 As well as discussing the Child Care Subsidy, the note highlighted the following legislated changes that also took effect following the 2017–18 Review:

•  the introduction of the LMITO;

•  raising the top income threshold for the 32.5 per cent marginal tax bracket from $87 000 to $90 000; and

•  increasing the low-income threshold to the Medicare levy in line with inflation.

[225] Ai Group submitted that these changes reduce the income tax payable for low and middle-income earners by up to $530 per annum and that this equates to an increase in disposable income for a person earning the current NMW of 0.63 per cent. 259

[226] The relevant budget measures took effect on 1 July 2018, namely:

•  the introduction of the LMITO (which provides an offset of $200 for those whose personal income is less than $37 000, increasing to $530 for incomes from $37 000 to less than $48 000 at a rate of 3 cents per dollar phasing out at a rate of 1.5 cents per dollar for incomes above $90 000); 260 and

•  increasing the Medicare levy thresholds for the 2018–19 income year (the Medicare levy’s low-income threshold for singles, families and seniors and pensioners was increased to take into account movements in the CPI; the threshold for singles increased from $21 980 to $22 938, while the threshold for families increased from $37 089 to $37 794). 261

[227] All parties were invited to comment on Ai Group’s submission regarding the impacts of the LMITO announced in the 2018–19 Budget. While ACCI agreed with the analysis, 262 the ACTU responded that this increase amounted to $5.97 before tax and would be less for households eligible for highly means-tested deductions and payments and may even result in a fall in income after taxes and transfers.263

[228] The Australian Government submitted that ‘[t]he introduction of the Low and Middle Income Tax Offset from 1 July 2018 will provide some tax relief for minimum wage earners’. 264 In response to a question on notice from the Panel, the Department of Jobs and Small Business and the Treasury provided an analysis of the impact of the 2018–19 Budget measures on the disposable income of a range of NMW household types.

[229] The Australian Government’s modelling shows that all but 2 household types receive an increase in their disposable income of $4 to $5 per week due to the 2018–19 Budget measures. The two household types which did not receive an increase were a single parent working part time with 1 child and a single parent working part time with 2 children. 265

[230] Table 3.2 presents the changes to nominal disposable income for selected hypothetical NMW-reliant households as at July 2018. It also shows the changes to each household’s disposable income between July 2017 and July 2018 (in dollar and percentage terms) which incorporates the increase to the NMW and modern award minimum wages from the 2017–18 Review decision and the tax-transfer settings effective from 1 July 2018. The final column presents the percentage change in household disposable income over the five years to July 2018. The percentage changes in household disposable income over the year and over the five years are compared with increases in the CPI over the year and five years to the June quarter 2018.

Source: Statistical report, Table 8.4; ABS, Consumer Price Index, Australia, Mar 2019, Catalogue No. 6401.0; Fair Work Commission modelling.

Note: The percentage change in the CPI over July 2017 to July 2018 and July 2013 to July 2018 were calculated with reference to the June quarter. CPI data are expressed in original terms. Tax/transfer parameters as at July each year. For assumptions to modelling, see Statistical report, Table 8.4.

[231] Table 3.2 shows that the disposable income of a single adult on the NMW was $646.15 per week as at July 2018, an increase of $22.12 over the year (1.4 per cent in real terms). The increase in nominal disposable income for this household was 13.4 per cent over the five years.

[232] The table also shows that over the year to July 2018, nominal disposable incomes across all of the 14 selected household types increased. This compares with the previous year when the nominal disposable incomes increased in only 8 of the 14 households. 267 However, 2 household types that included income from the NSA had incomes that rose by less than inflation.

[233] Over the year to July 2018, households without children and not receiving NSA received a 3.5 per cent increase to their nominal disposable income (a 1.4 per cent increase in real terms), while the increase for all other households was less due to the operation of the tax-transfer system.

[234] Over the year to July 2018, 11 household types experienced an increase in real household disposable income, compared with only 3 household types over the year to July 2017. 268

[235] Over the five years to July 2018, 7 of the household types had a real increase in income. Again, this is an improvement on the four household types whose real incomes rose over the 5 years to July 2017. 269

[236] Table 3.3 shows the increase in nominal household disposable income from the 2017–18 Review decision considering the tax-transfer settings as at July 2018. It also displays the amount and proportion of the increase actually retained by each household. The net increase in each household’s disposable income is separated into the change from the wage increase as a result of the 2017–18 Review decision as well as the changes to tax payable and transfer income received effective from 1 July 2018.

Source: Fair Work Commission modelling.

Note: Statistical report, Table 8.5; Tax/transfer parameters as at July 2018. The NMW increase retained is calculated as the change in disposable income as a proportion of the wage increase. Other assumptions as per Statistical report, Table 8.4.

[237] Table 3.3 shows that the proportion of the NMW increase retained by different households varied significantly due to the tax-transfer system. Single parents working part time retained 100 per cent of the increase while a single-earner couple receiving NSA retained less than one-eighth of the increase. For the 5 household types that retained the lowest proportion of the NMW increase, the withdrawal of transfer payments played a larger role than the impact of higher income tax. Households that received NSA retained the smallest share of the increase to the NMW.

[238] Table 3.3 also shows that the gross increase in household income for a single adult following the 2017–18 Review was $24.30 per week. The single-adult household retained $18.28 of this increase (or 75.2 per cent) due to the deduction of income tax, of $6.02. This is lower than the proportion retained by single adults following the increase awarded in the 2017–18 Review (79.0 per cent). 270 The reduction is due to some of the increase to the NMW being taxed at the 32.5 per cent marginal tax rate; a result of the gross yearly income of a full-time NMW worker increasing above the tax threshold of $37 001.

[239] However, the effect of ‘bracket-creep’ is offset by the changes to the tax-transfer system over 2018–19, including the LMITO. As shown in Table 3.2, disposable income for the single-adult household increased by 3.5 per cent to $646 per week, the same increase as that awarded in the 2017–18 Review decision. The response prepared by Treasury and Department of Jobs and Small Business to questions put by the Panel showed that the disposable income of this household would have been $642 per week before the changes to the tax-transfer system in 2018–19 and was $646 when the changes are included. 271

[240] Minimum wages continue to play an important role in maintaining the disposable incomes of households not receiving income support payments. 272 A large majority of adult low-wage workers have no dependent children and the many who work full time are not assisted by the tax-transfer system: indeed, they have their disposable incomes reduced by income tax.273

[241] As noted in last year’s Review, we accept that changes to the tax-transfer system are important to our consideration of the relative living standards and needs of the low paid. 274 The identified changes to the tax-transfer system and their effects on low-paid employees have been taken into account.

[242] Tax-transfer changes which have taken effect in the current Review period have, broadly speaking, provided a benefit to low-paid households. These changes are a moderating factor on our assessment of the appropriate level of increase to the NMW and modern award minimum wages arising from this Review. But it is not appropriate to apply a direct, quantifiable, discount to the increase in the NMW and modern award minimum wages we would have awarded in the absence of such changes in the tax-transfer system. As the Panel said in the 2012–13 Review decision:

‘The range of considerations we are required to take into account calls for the exercise of broad judgment, rather than a mechanistic approach to minimum wage fixation.’ 275

[243] We would add that reducing the level of the increase that we would otherwise determine by an amount directly referable to the increase in disposable income resulting from the 2018–19 Budget measures would be inappropriate for two reasons.

[244] First, such a direct discount would unfairly impact on some low-paid employees who did not benefit from the Budget measures.

[245] Second, the Budget measures are plainly intended to benefit low-paid households. A direct discount which entirely negates that legislative intention is not appropriate.

Relative earnings and earnings inequality

[246] The relative living standards of workers are affected by the degree of dispersion in earnings. If the earnings of workers in the lowest deciles are growing more slowly than those in the higher deciles, then the relative earnings of the low paid will fall. 276 We take into account a broad range of measures when assessing relative living standards and inequality.

[247] In this section we consider earnings inequality with reference to changes in the minimum wage relative to median earnings of full-time employees (the minimum wage bite) and, more broadly, in the distribution of real weekly earnings for full-time non-managerial adult employees.

Minimum wage bite

[248] The minimum wage bite is a commonly used indicator of relative earnings that compares the NMW with median earnings. In last year’s Review, we said that although the minimum wage bite is usually based on median earnings for full-time employees, it is also useful to consider the minimum wage bite for hourly earnings because it includes the many part-time employees who are paid at or near to the NMW. 277

[249] Chart 3.2 presents the minimum wage bite for several measures of median earnings, both weekly and hourly. Each measure shows a decline in the minimum wage bite in 2018.

Source: Statistical report, Chart 8.3; ABS, Characteristics of Employment, Australia, various, Catalogue No. 6333.0; ABS, Employee Earnings, Benefits and Trade Union Membership, Australia, various, Catalogue No. 6310.0; ABS, Weekly Earnings of Employees (Distribution), Australia, various, Catalogue No. 6310.0; Metal, Engineering and Associated Industries Award 1998; Manufacturing and Associated Industries and Occupations Award 2010.

Note: Median earnings are measured in August of each year. The NMW represents the minimum wage of each year, which is equal to the C14 rate. Following the amendments to the Workplace Relations Act 1996 (Cth) taking effect in 2006, the NMW (then called the Federal Minimum Wage) was set at $12.75 per hour, equivalent to $484.50 per week. Earnings are for employees including owner managers of incorporated enterprises. Median earnings from 2004 onwards are taken from the COE survey. The median earnings data reflect revised estimates as a result of rebenchmarking estimates.

[250] It is evident from Chart 3.2 that the minimum wage bite has been trending upwards in recent years, but fell in 2018. When compared with weekly earnings of full-time employees, the minimum wage bite was estimated at 54.1 per cent in 2018, 0.8 percentage points lower than in 2017, though it remains higher than its low point in 2008.

[251] The reported decrease in the minimum wage bite in 2018 appears counterintuitive as the increase in the NMW (3.5 per cent) was greater than the increase in the WPI (2.3 per cent). The reduction in the minimum wage bite implies an increase in median earnings of 5.1 per cent. A possible explanation may be the data sources used and that annual changes in median earnings have tended to be more volatile than other measures of average wages. (In 2017, median weekly earnings of full-time employees increased by only 1.5 per cent, lower than the WPI (2.0 per cent) and AWOTE (2.4 per cent), 278 but rose to 5.1 per cent in 2018.) This volatility is reduced over the longer term. Over the decade to 2018, median weekly earnings of full-time employees increased by 33 per cent, lower than AWOTE (38.6 per cent).279 Accordingly, it may be more useful to analyse the minimum wage bite over a longer period rather than focussing on movements from year-to-year. Like a number of measures of inequality, the minimum wage bite fell substantially between the mid 1990s and the mid-2000s. It has remained fairly steady since.

[252] Chart 3.3 compares the changes in the nominal value of the C14 and C10 rates with changes in measures of average earnings between the December quarter 2009 and the March quarter 2019. Over this period, growth in the C14 and C10 rates was stronger than growth in the average measures, particularly the WPI and average weekly earnings (AWE), due to slower growth in these average measures over the last few years. The faster growth in the NMW and modern award minimum wages has resulted in some improvement in the relative position of the low paid.

Source: Statistical report, Chart 8.1; ABS, Average Weekly Earnings, Australia, Nov 2018, Catalogue No. 6302.0; ABS, Wage Price Index, Australia, Mar 2019, Catalogue No. 6345.0; Metal, Engineering and Associated Industries Award 1998; Manufacturing and Associated Industries and Occupations Award 2010.

[253] Table 3.4 compares the growth in full-time real weekly earnings for 1998–2008 with 2008–2018 280 and shows that in the latter decade, earnings dispersion continued to rise, but at a significantly reduced rate. As with the minimum wage bite, there was a substantial rise in inequality in the earlier period which has since been attenuated, but not reversed. It should be noted that every percentile has had increases, in both periods, in the real value of their earnings.

Source: Australian Government submission, 15 March 2019 at para. 275, Table 8.2.

Note: Growth from 1998 to 2008 is for total cash earnings (excluding salary sacrifice) and from 2008 to 2018 is for total cash earnings (including salary sacrifice).

[254] Table 3.5 provides further information on real earnings dispersion for full-time non-managerial employees over the past decade. It shows that from 2008 to 2018 there was a degree of stability in the relative earnings distribution when compared with either mean or median real earnings.

Note: Earnings data for 2014 and 2016 are based on full-time non-managerial employees paid at the adult rate.

Source: Statistical report, Table 8.3; ABS, Consumer Price Index, Australia, Sep 2018, Catalogue No. 6401.0; ABS, Employee Earnings and Hours, Australia, various, Catalogue No. 6306.0.

Income and wealth inequality

[255] We take into account a range of measures when assessing inequality, relevant to our consideration of relative living standards. In this section we discuss a recently published research paper by the PC and look at 2 measures: the Gini coefficient and the distribution of equivalised household disposable income.

Productivity Commission Research Paper ‘Rising inequality? A stocktake of the evidence’, 2018

[256] The stated purpose of the PC Paper was to contribute to an informed discussion on inequality and its social impacts using the latest and most complete evidence measuring the level of, and trends in, inequality, poverty and disadvantage across Australian society. 281 The PC Paper uses ‘an array of indicators’ to examine distributions of household incomes, consumption and wealth, their composition and movements within the distributions over time as this cannot be done by one single metric.282

[257] The Australian Government submitted that the PC Paper showed that ‘[i]ncome inequality in Australia has been broadly stable for more than a decade. The Gini coefficient for equivalised household disposable income stood at 0.323 in 2015–16 (latest data), slightly below the 0.336 recorded in 2007–08’. 283

[258] In their reply submission, the ACTU contended that the PC Paper downplays the significant increases in the incomes of the top 1 per cent of households. 284 The ACTU also submitted that it agreed with the PC Paper that ‘a more complete picture of economic well-being is obtained by jointly considering income and wealth inequality’.285 We return to the issue of wealth inequality shortly.

[259] The Australian Government also submitted that the PC Paper found that ‘Australia’s tax and transfer system has consistently acted to substantially reduce income inequality’ with these payments reducing the Gini coefficient.²⁸⁶ They argued that ‘the impact of minimum wage increases on income inequality is more ambiguous, as minimum wage and award-reliant workers can be found across the household income distribution.’ 287 The PC Paper highlighted the important role that Australia’s tax-transfer system plays in reducing income inequality:

‘… Australia’s progressive tax system and highly targeted transfer system substantially reduce income inequality. Income tax and government transfers have typically lowered the measure of overall income inequality (the Gini coefficient) by 30 per cent, an equalising effect that far outweighs the overall increase in the measure since the late 1980s. This equalising effect has fluctuated over time, but overall there has been no material change in the past thirty years.’ 288

[260] The PC Paper is an important contribution to the debate on inequality and we agree with the PC that it is important to take account of multiple measures and data sources when trying to understand complex and highly contested issues such as inequality, poverty and disadvantage. Of particular relevance to the relative living standards of the low paid are the findings which emphasise the importance of measures of inequality which focus on equivalised disposable income.

[261] The PC Paper provides a useful way of conceptualising disadvantage (Chart 3.4). 289 The three main elements that contribute to disadvantage—poverty, material deprivation and social exclusion—are related, and often overlap, but it is also possible for someone to experience only 1 or 2 elements at a time.

Source: PC Paper, p. 109.

[262] We also note the observation in the PC Paper that the distribution of wealth is important to consider alongside that of income as ‘the sources of widening wealth inequality … can be dynamic factors that influence future income inequality’. 290 Further, as the PC Paper points out, wealth is much less evenly distributed than income:

‘In 2015–16, the person at the 90th percentile of the wealth distribution had about forty times as much wealth as the person at the 10th percentile. In contrast, the person at the 90th percentile of the income distribution had four times as much income as the person at the 10th percentile.’ 291

[263] However, as we noted in last year’s Review, 292 despite the evidence of substantial and increasing wealth inequality, it is unlikely that any moderate adjustment to the NMW or modern award minimum wages arising from the Review would have any discernible effect upon wealth inequality.

Gini coefficient

[264] The Gini coefficient is the most commonly used indicator depicting inequality over the whole income distribution and is expressed as a value between 0 and 1. A zero value indicates perfect equality (all people have the same income) and a value of 1 indicates perfect inequality (one person has all the income). In last year’s Review, we noted that based on equivalised household disposable income, the Gini coefficient was higher in 2015–16 than it was in the mid-1990s, though it had stabilised from at least 2007–08. 293 This was based on data from the ABS Household Income and Wealth. The ABS has not released any new data on the Gini coefficient since the last Review.

[265] The Australian Government submitted that assessing the Gini coefficient over time is difficult as the ABS has changed its method for measuring income. 294

[266] A longer-term analysis of income inequality using the Gini coefficient was undertaken by the PC in the PC Paper. The ABS data showed that income inequality had risen between 1988–89 and 2015–16, particularly during the mining boom, before stabilising; while data from the HILDA survey found that inequality has been fairly stable since the early 2000s. 295 The PC explained that the discrepancy between the ABS and HILDA data is partly caused by the changes to the ABS definitions of income and their survey methodology and that while the changes improved the accuracy of measuring income, it creates difficulties with comparisons with previous years.296 As the PC observed, the trend in income inequality is ‘contested territory’.297

[267] The Gini coefficient is ‘a single statistic that summarises the extent of inequality for the whole distribution’. 298 In their reply submission, the ACTU raised concerns and limitations around the use of the Gini coefficient as a ‘comprehensive measure of inequality’299 as the middle of the income distribution ‘gets more weight’ than either the higher or lower ends (which are ‘not well captured’) such that greater disparity between the top and bottom of the distribution can result in the same Gini coefficient.300

[268] The User Guide to the ABS Survey of Income and Housing explains that the Gini coefficient can be sensitive to changes around the middle of the income distribution because the derivation ‘reflects the ranking of the population, and ranking is most likely to change at the densest part of the income distribution, which is likely to be around the middle of the distribution.’ 301 However, the ABS concludes that ‘[t]he Gini is preferred over other summary measures because it is not overly sensitive to the very low income or wealth values that can be reported, and it is relatively simple to interpret’.302 The PC Paper notes a similar limitation, stating that the Gini coefficient ‘does not indicate where in the distribution any changes in inequality have occurred.’303 To complement this measure, the PC also examined alternative measures of inequality, such as quantile ratios (comparison of incomes at different points of the distribution), finding that these ratios ‘have varied in broadly similar ways to Gini coefficients’.304 Similar additional measures are also presented in the Statistical report.305 For the same reason we also consider changes in the 10^th and 25^th percentiles of the earnings distribution as ratios of the mean and median, as presented in Table 3.5 above.

[269] Despite its limitations, the Gini coefficient provides useful information on overall income inequality, and our assessment of relative living standards.

[270] The Australian Government also provided data from the OECD showing that Australia has the 14^th (out of 36) highest level of inequality as measured by the Gini coefficient using data for 2016, down from 11^th based on data for 2015. 306 Australia remains more unequal than a number of major European countries (including France and Germany) and Canada. It remains more equal than the US, UK and New Zealand (Chart 3.5).

[271] Australia’s position in the distribution suggests that a number of similar countries, including Canada, the Netherlands, France and Germany, have achieved higher degrees of equality than in Australia. But, as noted in last year’s Review, it is not possible to make such international comparisons with precision. 307

Chart 3.5: Gini coefficient measure of inequality among OECD countries: household disposable income, 2016 or latest year

Source: Australian Government submission, 15 March 2019 at para. 269, Chart 8.1.

Note: Latest available data refer to 2016, except for Chile, Denmark, Iceland, Ireland, Japan, Korea, where the latest data available were for 2015; and Hungary, Mexico, New Zealand and Switzerland where latest data available were 2014.

Needs of the low paid

[272] There are broadly 3 ways in which to assess the extent to which the needs of the low paid are being met.

[273] The first approach uses experts, informed by community consultations, to decide what types of goods and services an individual or family needs in order to be able to maintain a defined standard of living. Having constructed this basket of necessary goods and services, they must be priced and all the costs added to reach a total that reflects the income needed to reach the minimum budget standard.

[274] The second approach is to accept that the needs of the low paid are established by comparison with others in their community. On this approach, it will be presumed that the needs of the low paid will be met if the incomes of the low paid are not ‘too far’ below typical incomes. This is a strictly relative measure, and it is a matter of informed judgement to decide what constitutes ‘too far’. Its practical application is to identify a threshold income that is some proportion of median income and 60 per cent is a commonly used threshold or ‘poverty line’.

[275] The third method is to focus on indicators of deprivation, such as whether the low paid have to seek help from charities or are unable to buy certain essentials (such as heating their houses).

[276] Both the first and the second approaches have to be adjusted for differences in the size and composition of the household. All 3 approaches have something to offer.

Budget Standards

[277] A budget standard estimates how much money a particular family type needs to achieve a particular standard of living in a particular place at a particular time. It is derived by specifying the standard and then identifying all of the items required to reach it, pricing each item and calculating the cost of the total ‘basket’ of goods and services required. 308 Budget standards are generally used to estimate the income levels required to achieve a minimally adequate standard of living. Contemporary budget standards can provide an effective means of measuring the ‘needs of the low paid’.

[278] Budget standards require revision from time to time to maintain relevance. The components of a budget standard need to be revised annually to reflect movements in prices and, over a longer timeframe, to reflect changes in, among other things, the ‘basket’ of goods and services required to meet the relevant standard. 309 In past Reviews, parties have referred to research on budget standards conducted in the mid-1990s and the Panel has concluded that such research ‘lacks contemporary relevance’.310 The failure to adjust the 1997 SPRC standards for anything other than the overall level of consumer prices led the Panel to make the following observation in the Annual Wage Review 2013–14 (2013–14 Review) decision:

‘The budget standards material submitted by ACOSS and ACCER is based on a 1997 study by the SPRC, commissioned by the former Department of Social Security, to assist in the assessment of the adequacy of social security payments. We accept that contemporary budget standards measures can provide an effective means of measuring the needs of the low paid, which can be considered, together with other relevant data. However, the budget standards measures derived from the 1997 SPRC study do not provide useful contemporary information about the needs of the low paid.’ 311

[279] In 2017 the SPRC at UNSW published new budget estimates 312 (the 2017 Budget Standards Report) based on the Minimum Income for Healthy Living (MIHL) standard developed in the UK public health literature.313 The MIHL standard is designed to ensure that all individuals are able to lead healthy lives and participate in society. The MIHL standard is thus designed to ensure that each individual is able to achieve levels of consumption (of food, clothing, medications, transportation, personal care, and so on) and participation (in lifestyle, exercise and social activities) that are consistent with healthy living.

[280] The 2017 Budget Standards Report is intended to provide ‘a set of contemporary budget standards that reflect the needs of low-income working and unemployed individuals and families that can be used to assess income adequacy and guide decision-making.’ 314

[281] The 2017 Budget Standards Report provides budget estimates for 5 different family types and these are compared to their disposable income. The 5 basic family types are: single people; couples without children; couples with 1 and 2 children; and a sole parent with 1 child. Adults were assumed to be between 35 and 40 years of age, while the children are both of school age: a 6 year-old girl and a 10 year-old boy. The 2017 Budget Standards Report notes that the decision to include only school-age children reduces the need for child care that would have added to the budgets in some instances if the children had been younger, while the inclusion of older children would have also added to the costs of things like food, clothing and social participation. Each family is assumed to contain one adult who is assumed to be either in work and receiving the minimum wage, or unemployed and receiving social security assistance in the form of NSA or Parenting Payment for the sole parent.  315

[282] The budgets in the 2017 Budget Standards Report draw on 3 kinds of data: expert (normative) data that reflects prevailing judgments on how much is needed to achieve specific standards; survey data that describes the spending patterns of actual families; and focus group data that captures how real families budget and make ends meet. 316 The ‘grossed-up’ weekly budgets for low-paid families in June 2016 are: $597.31 for a single person; $833.24 for a couple; $969.90 for a couple with one child; $1173.38 for a couple with two children; and $827.70 for a sole parent with one child.317

[283] The descriptor ‘Grossed-up’ is apt to mislead and in this context simply means that the cost of housing (estimated using the rental threshold approach) is added to the price of the budget basket of goods and services. 318

[284] The budgets in the 2017 Budget Standards Report are contemporary, in the sense that the data upon which they draw is comparatively recent; but they do require revision to take account of movements in prices. We return to that issue shortly.

[285] In last year’s Review proceedings the weight to be given to the 2017 Budget Standards Report was in dispute. The ACTU argued that ‘the budget standards research should be given no special status in the array of material for evaluating relative living standards and the needs of the low paid’ 319 because, it submitted, the budget standards were ‘far too modest to in fact reflect the needs of a person or a household’320 and had ‘failed to enumerate or address the many costs of working that are not immediately apparent’.321 ACOSS pointed out that the budgets proposed were ‘extremely tight’ and that ‘the new budgets would support only a very frugal living standard, arguably below that which the National Minimum Wage should support’.322 ACCI submitted that the research did not ‘elevate relative living standards and the needs of the low paid above other considerations’323 and urged the Panel to ‘use caution’ in forming a decision based on the new budget standards research.

[286] The Australian Catholic Council for Employment Relations (ACCER) (now ACBC) broadly supported the use of the budget standards research on the basis that it identifies and measures a contemporary decent standard of living consistent with the Panel’s statements that an assessment of the needs of the low paid ‘requires an examination of the extent to which low-paid workers are able to purchase the essentials for a ‘decent standard of living’ and to engage in community life, assessed in the context of contemporary norms.’ 324

[287] In the 2017–18 Review decision, the Panel noted that, overall, there was low support from parties regarding the consideration to be given to the budget standards in the Review. 325 However, we concluded:

‘The new budget standards research is the first time that a serious effort has been made, using contemporary scholarship in this field, to estimate the needs of low-paid employee households. We judge it to be useful and relevant, while recognising its limitations and the Panel has taken it into account along with all of the relevant material we have before us. We note the comment from the authors that ‘[b]udget standards are not a panacea but they provide important information that can inform and assist decisions taken about adequacy…’

We agree with ACCER’s submission that the research is the ‘best evidence available in regard to the needs of the low paid Australian workers and their families,’ but also with ACCI’s submission that the “budget standards cannot of themselves be determinative of the NMW or any uprating of minimum award rates”.’ 326

[288] In the current Review proceedings, the ACBC reiterated its commitment to ‘an evidence-based minimum wage system that pays due regard to the needs and living costs of low income families’ and submitted that ‘[a]though imperfect, the budget standards research is the best evidence that we have about those needs’. 327 The ACBC went on to point out, in Chapter 4:D(5) of its submission, that Table 5.17 in the 2017 Budget Standards Report ‘contains significant errors in relation to the calculation of safety net income’ and ‘gave a very misleading summary of the living standards of low income families with children’.328

[289] Table 5.17 from the 2017 Budget Standards Report is reproduced below as Table 3.6.

Family type	Grossed-up budget Standard (1)	Low-Paid Safety-Net Income (2)	(2) minus (1)	Grossed-up Budget Standard (1)	Unemployed Safety Net Income (2)	(2) minus (1)
Single adult	597.31	659.22	61.91	433.68	337.68	–96.00
Couple, no children	833.24	794.21	–39.03	660.25	552.84	–107.41
Couple, 1 child (G, 6)	969.90	978.74	–8.84	766.74	708.28	–58.46
Couple, 2 children (G, 6 & B, 10)	1173.38	1084.64	88.74	940.37	814.13	–126.24
Sole parent, 1 child (G, 6)	827.70	872.56	44.86	675.18	627.79	–47.39

Source: 2017 Budget Standards Report, p. 103.

Note: G, 6 = girl aged 6 years; and B, 10 = boy aged 10 years.

[290] The reference to the ‘Grossed-up’ budget standard in Table 3.6 is explained earlier, at [283].

[291] The basis for the ‘Safety Net’ calculations in the above table is explained in the 2017 Budget Standards Report:

‘The ‘safety net’ calculations are based on the minimum wage levels and social security payment rates and conditions that exist in April 2016 and assume that each family receives every entitlement for which it is eligible. This means, for example, that they do not include the increase in the minimum wage (of 2.4 per cent) announced on 31 May 2016 and paid from 1 July 2016. Nor do they include the automatic indexation of social security payment rates that took effect at the beginning of July 2016, only those that came into effect on 1 March 2016. Note that the timing difference between April 2016 to which the safety net incomes apply and the June Quarter 2016 to which the budget standards apply will not impact on the comparisons presented below because the safety net (and taxation) provisions that existed in April remained in force throughout the June Quarter.’ 330

[292] The short point put by ACBC is that the ‘Safety Net’ calculations are erroneous, in particular it is submitted that:

•  the single adult is said to have a safety net income of $659.22 per week when in fact the net wage of a NMW dependent in April 2016 was $594.01 per week; 331

•  the calculations of safety net income of the other 4 household types are inconsistent with the data in Table 8.2 of the FWC Statistical Report of April 2016, regarding disposable incomes from wages and transfers at December 2015; 332 and

•  the disposable incomes of the two couple households (one child and two children) erroneously include NSA for the person who is not employed. 333

[293] We posed a question on notice to all parties inviting comment on the points raised by the ACBC and also invited a response from Professor Saunders, one of the authors of the 2017 Budget Standards Report.

[294] In his response, Professor Saunders acknowledged that there are some errors in safety net income calculations contained in Table 5.17 of the 2017 Budget Standards Report which resulted in an over-estimation of the ‘Safety Net income’ for some families, but observed that:

‘The errors in Table 5.17 are important, although they do not effect the overall content and context of the report, which is most valuable for the new budget standard estimates that are not affected by the error.’ 334

[295] In response to the question on notice the ACTU submitted that ‘what appear to be errors’ identified by ACBC in what the 2017 Budget Standards Report describes as the safety net income of the low paid group raises a ‘real issue’ about whether the 2017 Budget Standards Report is accurate in its assessment of which particular household types exceeded the relevant budget standard. 335 In these circumstances the ACTU submitted that:

‘The appropriate course to definitively determine this would be to seek information from the Department of Social Services about how the estimates of Safety Net Income provided to the authors were calculated.’ 336

[296] Ai Group responded that the general nature of the claims raised by ACBC ‘reinforce the cautionary views presented in Ai Group’s submission to the 2017–18 Review’. 337

[297] The ‘cautionary views’ advanced by Ai Group in last year’s Review concerned the use of the budget standards to support an increase in minimum wages, rather than a critique of the budget standards themselves. As Ai Group put it:

‘While we welcome the Budget Standards research, we are very wary of it being used to support an argument to raise minimum wage rates to address discrepancies between calculated Grossed-up Budget Standards and the safety net level of income for particular household types. We are particularly concerned that adjusting wage rates to address this discrepancy would raise the risks of disemployment both for this household type and for all other low-paid households including low-paid households types for which no discrepancies between the calculated Grossed-up Budget Standards and the Safety-net level of income, is suggested.’ 338

[298] ACCI submits that ‘were the ACBC correct, the conclusion that must be drawn is that the budget standards research would need to come off the table’. 339 Later, ACCI submitted that:

‘There is nothing new in 2019. The budget standards research has not been revised, it is known rather than new information, and there is no basis for it [to] have any more or less determinative value in 2019 than it had in 2018.’ 340

[299] In the alternative, ACCI contends that the Budget Standards research has become dated and this ‘favours not revisiting it or treating it as current or determinative in 2019’, or, if this is not correct, the ‘only available approach’ would be to note ACBC’s criticisms and ‘to no longer conclude that the budget standards research can be deemed useful and relevant.’ 341

[300] We also note that in the present proceedings ACOSS reiterated its view that ‘the new budgets would support only a very frugal living standard, arguably below that which the NMW should support and significantly below that of the previous ‘modest but adequate’ budget standard’. 342 There is considerable force in the point advanced by ACOSS and indeed the authors of the 2017 Budget Standards Report concur, as is evident in the following statements from the Report:

‘the new MIHL standard for low-paid workers is intended to be lower than the previous MBA standard, which was designed to apply to a standard of living that is around the median …

The focus was … very much on producing minimum budgets, as required by the MIHL standard. Whatever the merits of the view that the previous budgets erred on the generous side it is very difficult to sustain the claim that this is also the case in relation to the new budgets. They are extremely tight and leave no room for even the most modest of “special treats”.’ 343

[301] So where does this leave us? As Professor Saunders points out in his reply, the issues raised by ACBC do not detract from the budget standards presented in the Report; rather they are directed at the estimate of the safety net disposable income of various household types and the comparison between their disposable income and the relevant budget standard. The errors alleged by ACBC do not affect the validity of the budget standards themselves; that said, the budget standards themselves are contested and, as acknowledged by the authors of the Report, they are ‘extremely tight’.

[302] In the circumstances it is not necessary for us to reach any concluded view as to the validity of the errors alleged by ACBC. We later compare the budget standards to measures of nominal disposable income from the Statistical report and the Australian Government submission (see Table 3.8).

[303] In last year’s decision we expressed the view that the 2017 Budget Standards Report was ‘useful and relevant’, 344 while acknowledging that budget standards cannot of themselves determine the outcome of the Review.345 We remain of that view, though with a qualification which acknowledges the force of the submission put on behalf of ACOSS.

[304] As mentioned earlier, the budgets in the 2017 Budget Standards Report are contemporary, in terms of the data upon which they draw, but require revision to take account of movements in prices.

[305] The submissions before us do not address the appropriate method for adjusting the budget standards. It may be inferred from ACBC’s submission that they are to be adjusted by movements in the CPI. The ACBC submitted:

‘Since June 2016 wage increases have been greater than CPI increases and the growth in disposable incomes have been modest. Assessing the impact of these changes is not a precise arithmetical exercise.

On the evidence, it is reasonable to conclude, consistent with the FWC’s assessment in the June 2018 decision, that, in early 2019, a full time job at the NMW rate is sufficient to provide a single adult with a reasonable standard of living. It may be that further evidence (perhaps including a refinement of the estimation of relevant changes since June 2016) will show that it is not sufficient, but at the present time, the evidence supports that conclusion’. 346

[306] The authors of the 2017 Budget Standards Report suggest that the use of the CPI to adjust the budgets over the short term is ‘standard practice in the budget standards literature’ and ‘will not induce major errors into the estimates’. 347 The short term is a period ‘covering up to about 7 years’ and, ‘beyond the seven-year time horizon, it is preferable to review and revise the entire budgets to ensure that items, quantities and lifetimes as well as prices are reviewed and adjusted to reflect changes in community norms and average living standards’.348

[307] The budgets were priced in the second half of 2013 (1 July to 31 December). We note that the finalised budgets in the 2017 Budget Standards Report were updated by the authors using CPI data to the June quarter 2016 ‘to maintain their relevance’. 349 The updating appears to have been undertaken in line with movements in the CPI at the main budget area level.350 The table below adopts the same approach and adjusts the various components in the budget of a single employed adult by movements in the corresponding component of the CPI since the June quarter of 2016 to the March quarter 2019.

Budget category	Budget priced at latter half of 2013		Budget updated using CPI June 2016			Budget updated using CPI March 2019
	$	Index^	$	Index^	% change from 2013	$	Index	% change from 2016
Food	60.19	101.1	61.80	103.8	2.7	64.90	109.0	5.0
Clothing and footwear	11.05	100.2	10.81	98.0	–2.2	10.25	92.9	–5.2
Household goods and services	77.11	101.9	79.23	104.7	2.7	78.02	103.1	–1.5
Transport	81.59	103.1	77.71	98.2	–4.8	82.14	103.8	5.7
Health	6.40	109.5	7.33	125.5	14.6	7.96	136.3	8.6
Personal care	14.99	104.4	15.59	108.6	4.0	16.38	114.1	5.1
Recreation	28.49	101.1	29.04	103.1	2.0	29.60	105.1	1.9
Education	0.00	108.8	0.00	124.6	14.5	0.00	135.9	0.0
Budget standards	279.82	–	281.51	–	0.6	289.25	–	2.7
Weekly rental costs*	nd	–	315.80	110.2	–	321.82	112.3	1.9
Grossed-up budget standards	nd	–	597.31	–	–	611.07	–	2.3

Source: Fair Work Commission (2019), Information note—Budget standards research, additional material to the 2018–19 Review, 8 May; Saunders P & Bedford M (2017), New Minimum Income for Healthy Living Budget Standards for Low-Paid and Unemployed Australians, SPRC Report 11/17, Social Policy Research Centre, UNSW Sydney, Table 5.7; ABS, Consumer Price Index, Australia, Mar 2019, Catalogue No. 6401.0.

Note: Highlighted columns are updated data to March 2019 based on indexes from June 2016. ^Based on CPI indexes at Table B.2 of report. *March quarter 2016 rent figures were used in the report (see p. 87 of the 2017 Budget Standards Report); the indexed March 2019 figure reflects this. nd = not derived.

[308] The revised budget standard for a single adult employee is $611.07, the disposable income of a single adult on the NMW is $646.15, a difference of $35.08 per week. Hence, the disposable income of a single adult earning the NMW is above the corresponding MIHL budget standard.

[309] As mentioned earlier, there are some errors in the calculation of the ‘safety net’ income of the other family types in Table 5.17 of the 2017 Budget Standards Report which have resulted in an over-estimation of the disposable income of some families. In these circumstances, it is difficult to draw any firm conclusion about whether the disposable income of the selected family types is above or below the estimated budget standard. However, based on other data about the nominal disposable income of selected NMW-reliant households, some reasonable inferences can be drawn. Table 3.8 below compares the MIHL budget standard for certain family types (as at June 2016) with the nominal disposable income of those family types, as at July 2018 (from Table 8.4 of the Statistical Report) and with the Australian Government submission.

	MIHL Budget(1)	Nominal disposable income, Statistical report(2)	Nominal disposable income, Australian Government submission(3)
	June 2016	July 2018	January 2019
Single-earner couple, no children	833.24	660.54 824.50 (with NSA)	763 (with NSA)
Single-earner couple, 1 child	969.90	906.87 1014.01 (with NSA)	937/960* (with NSA)
Single-earner couple, 2 children	1173.38	1012.71 1120.74 (with NSA)	1069 (with NSA)
Sole parent working PT, 1 child	827.70	605.93	np

Source: (1) See Table 5.14 of the 2017 Budget Standards Report. (2) See Table 8.4 of the Statistical Report. (3) Australian Government submission, 15 March 2019 at para. 286, Table 8.4.

Note: *Child can be aged 3 years or 9 years. np = not provided.

[310] On these figures the nominal disposable income of all but one family group (the single-earner couple with one child in receipt of NSA) is below the relevant budget standard. This is particularly telling given that the above MIHL budget estimates are for June 2016 and it is likely that the budget estimates will be revised upwards when price movements between June 2016 and March 2019 are taken into account.

[311] The measure of disposable income in the Statistical report and that provided by the Australian Government do not perfectly match the family types and assumptions discussed in the 2017 Budgets Standards Report. The main differences are due to the inclusion of NSA (although the Statistical report includes household types with and without NSA) and assumptions on the age of the children:

•  the budget standards assumes children are aged 6 years and 10 years. Assuming the child is aged 6 years means that the sole parent can be eligible for Parenting Payment Single.

•  the Statistical report assumes the child is aged between 8 and 12 years.

•  the Australian Government assumes the child is aged either 3 years or 9 years for couples with one child, and aged 3 and 9 years for couples with two children.

[312] A further difference is in the number of hours worked by the sole parent—the budget standards assume the sole parent works 20 hours a week, while the Statistical report assumes 19 hours a week (i.e., earning 50 per cent of the weekly NMW rate). The Australian Government modelling does not consider this family type.

[313] In addition, the Australian Government modelling also assumes that single-earner couples incur no child care costs as the non-working partner will look after the children. This means that there would be an even greater difference between disposable incomes and the budget standards for families that require child care. It appears also to assume that the single earner couple with no children does not receive rental assistance.

[314] It appears that the assumptions in the Statistical report more closely align with those in the Budget Standards report than those of the Australian Government.

[315] We see merit in future research addressing how the budget standards can be appropriately updated to take account of price changes over time to accurately match the various budget standards with the relevant household disposable income. In relation to the latter issue we note that the differences between the family disposable incomes reported in the second and third columns in Table 3.8 are not trivial, even though they purport to measure the same thing. This most likely reflects the tailored, and complex, nature of the tax-transfer system.

Poverty and poverty lines

[316] As mentioned earlier, considering the needs of the low paid involves an assessment of an employee’s capacity to purchase the essentials for a ‘decent standard of living’ and to engage in community life, assessed in the context of contemporary norms. If the low paid are forced to live in poverty, then their needs are not being met. Further, those in full-time employment can reasonably expect a standard of living that exceeds poverty levels.

[317] There is no consensus on how to measure poverty. It can be measured in absolute or relative terms, although an absolute poverty line is not calculated in Australia and is more common in developing countries. 351 The PC Paper described absolute poverty indicators as those that reference a minimum level of income necessary to support a minimum acceptable standard of living.352

[318] A relative income poverty line approach regards people as living in poverty if their income is below a certain percentage of a measure of median income. By design a relative income poverty threshold reflects movement in the median standard of living. As the PC Paper observes, relative poverty measures can be sensitive to changes in median incomes. Hence an increase in relative poverty could indicate that real incomes at the bottom of the distribution are falling, but could also reflect strong growth in median incomes. 353

[319] Relative poverty lines are used to measure incomes in comparison to the broader community. Typically this shows the proportion of people earning less than a certain proportion (e.g., 50 or 60 per cent) of median income. These lines are customarily set based on the distribution of equivalised household disposable income as this takes into account not only private income but also tax and transfers and the impact of resource sharing by taking into account household size. Relative poverty ‘captures the notion that preferences and norms in society on what is an acceptable standard of living—and the costs associated with it—change over time as incomes rise’. 354 However, relative poverty lines are more a measure of inequality than an assessment of whether low-income households have enough income to meet their basic needs.355

[320] As noted in the PC Paper, a relative poverty threshold is simple and self-adjusting, but the selection of a threshold is arbitrary and can produce counterintuitive results:

‘… a fall in the United Kingdom’s median income resulted in a lowering of the relative poverty line and a consequent decline in the proportion of people living in relative poverty – even though deprivation increased as living standards declined (Fahmy 2014). Conversely, an increase in all incomes could produce a rise in the relative poverty rates, as was observed in Australia during the mining boom.’ 356

[321] In its analysis of poverty, the PC Paper considered 50 per cent of median income as a threshold. Based on this measure, the report showed that there was an increase in poverty from 1993–94, peaking at 12 per cent in 2009–10 due to strong growth in median income during the mining boom. 357 An estimated 9 per cent of Australians (or 2.2 million people) were below this measure of relative poverty in 2015–16,358 which is around its long-term average.359

[322] The Panel has generally relied on poverty lines that are measured relative to median equivalised household disposable income and has considered that a threshold of 60 per cent of median equivalised household disposable income is more appropriate when using relative poverty lines to set minimum wages, as those in full-time employment can reasonably expect a standard of living that exceeds harsher poverty levels. 360 We propose to continue to adopt that approach.

[323] Table 3.9 compares the equivalised household disposable income for a range of hypothetical households reliant on the NMW and selected modern award minimum wages with a relative poverty threshold calculated at 60 per cent of median equivalised household disposable income. These poverty lines are based on estimates of median equivalised household disposable income provided by the ABS. As these data are provided infrequently, they are adjusted (on a quarterly basis) for movements in household disposable income per head as calculated by the Melbourne Institute of Applied Economic and Social Research (Melbourne Institute), and further adjusted for household composition using the modified OECD equivalence scale. 361

[324] These relative poverty lines differ between the household types because of the different size and composition of the households. As we have stated previously, because of these differences, it is not feasible for minimum wages on their own to ensure that all families with a full-time minimum wage worker have incomes that exceed poverty levels. Larger families may need help from the welfare system. 362 In comparison, a single adult receives no assistance from transfer payments and is entirely reliant on their earnings.363

[325] Further, the margin between the selected relative poverty line and the equivalised disposable income of households reliant on the NMW and selected modern award minimum wages represents, at best, a broad indicator of the extent to which the needs of the low paid are met. 364

[326] Table 3.9 shows that, for the December quarter 2018, all households receiving the C4 rate or AWOTE had disposable incomes above the relative poverty line, apart from single parents working part time with 2 children (at the C4 rate). Households with equivalised disposable income below the relative poverty line were single parents with children working part time, single-earner couples and single-earner couples with two children with the non-earning partner not receiving NSA, where the employed member of the household was receiving the C10 rate.

[327] However, there were 5 households receiving the C14 rate that were below the 60 per cent median income poverty line in December 2018, all single-earner households—single parents working part time with 1 or 2 children; single-earner couples without children and single-earner couples with 1 or 2 children. This is fewer than the 7 households below the 60 per cent relative poverty line in December 2013.

[328] The table also shows that, between the December quarter 2017 and the December quarter 2018, the equivalised household disposable income increased or stayed the same for each of these hypothetical households at the C14, C10 and C4 rates as well as AWOTE relative to 60 per cent of median equivalised disposable income. However, for those households on these wage rates that had disposable income below the relative poverty line, the increase was not enough to move any household above the poverty line.

	December 2013					December 2017					December 2018
	60% median income PL	Disposable income as a ratio of 60% median income PL				60% median income PL	Disposable income as a ratio of 60% median income PL				60% median income PL	Disposable income as a ratio of 60% median income PL
	($ pw)	C14	C10	C4	AWOTE	($ pw)	C14	C10	C4	AWOTE	($ pw)	C14	C10	C4	AWOTE
Single adult	506.40	1.13	1.28	1.47	2.19	520.32	1.20	1.34	1.54	2.29	528.70	1.22	1.37	1.58	2.32
Single parent working FT, 1 child	658.32	1.25	1.37	1.50	1.91	676.41	1.31	1.42	1.55	1.94	687.32	1.32	1.43	1.58	1.95
Single parent working PT, 1 child	658.32	0.85	0.92	1.02	1.36	676.41	0.88	0.96	1.05	1.40	687.32	0.88	0.97	1.07	1.40
Single parent working FT, 2 children	810.24	1.15	1.25	1.36	1.69	832.51	1.19	1.28	1.39	1.68	845.93	1.20	1.29	1.41	1.70
Single parent working PT, 2 children	810.24	0.82	0.88	0.96	1.24	832.51	0.84	0.90	0.98	1.26	845.93	0.84	0.91	0.99	1.27
Single-earner couple (with NSA)	759.60	0.98	1.00	1.01	1.46	780.48	1.04	1.04	1.05	1.53	793.06	1.04	1.05	1.06	1.54
Single-earner couple	759.60	0.77	0.87	0.98	1.46	780.48	0.82	0.90	1.03	1.53	793.06	0.83	0.92	1.05	1.54
Single-earner couple, 1 child (with NSA)	911.52	1.02	1.04	1.08	1.38	936.57	1.06	1.07	1.12	1.40	951.67	1.07	1.07	1.14	1.41
Single-earner couple, 1 child	911.52	0.90	0.99	1.08	1.38	936.57	0.94	1.02	1.12	1.40	951.67	0.95	1.03	1.14	1.41
Single-earner couple, 2 children (with NSA)	1063.44	0.98	1.00	1.03	1.28	1092.67	1.01	1.02	1.06	1.28	1110.28	1.01	1.02	1.08	1.29
Single-earner couple, 2 children	1063.44	0.87	0.95	1.03	1.28	1092.67	0.90	0.98	1.06	1.28	1110.28	0.91	0.99	1.08	1.29
Dual-earner couple	759.60	1.16	1.33	1.54	2.31	780.48	1.24	1.41	1.62	2.40	793.06	1.27	1.44	1.66	2.43
Dual-earner couple, 1 child	911.52	1.18	1.28	1.40	1.93	936.57	1.24	1.33	1.45	2.00	951.67	1.25	1.35	1.48	2.02
Dual-earner couple, 2 children	1063.44	1.12	1.20	1.30	1.70	1092.67	1.16	1.24	1.34	1.72	1110.28	1.17	1.25	1.36	1.74