## _wp04144

## Source details

**Canonical URL:** [_wp04144](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf.json)

---

### Introduction: purpose and empirical approach
- Purpose:
  - Examine why per capita income and output across Australian states did not converge during the 1990s despite comprehensive economic reforms.
- Empirical approach:
  - Panel estimation over the sample period 1990–2001.
- Hypothesized impediments to convergence:
  - Labor market rigidities (centralized wage/award system).
  - Government transfers (to households and state governments, including equalization grants).
  - Skill-biased technological change and adoption of new technologies.
- Key empirical summary statements:
  - The wage awards system may have restricted adjustment of real wages to productivity differentials and contributed to higher unemployment in some states.
  - Government transfers to households appear to have adversely affected work incentives in high-unemployment states by limiting labor force participation.
  - Growth in relatively low-income/output states partially converged toward higher-income states during the 1990s, but initial differences largely remained.
  - Since 1997, the catch-up effect on output growth has slowed, and the impact on real per capita income growth across states of government transfers appears to have increased.
  - Federal grants to the states did not have a significant impact on output growth across states in the 1990s.
  - The impact of skill-biased technological change on real per capita income and output growth is not clear.

### Stylized facts on regional disparities (1990–2001)
- Two groups of states by real per capita output and income:
  - Above-national-average group: New South Wales, Victoria, Western Australia, Northern Territory, Australian Capital Territory (ACT).
  - Below-national-average group: Queensland, South Australia, Tasmania.
- Dispersion patterns:
  - Cross-state dispersion in output and income increased from the mid-1990s; dispersion in output far higher than in disposable income.
  - Income dispersion is largely influenced by the ACT; excluding the ACT, income dispersion is lower and more or less unchanged during the 1990s.
  - Output dispersion excluding the ACT rises after 1997.
- Selected numerical indicators (Real per Capita Ratio to Australian Average; Real per Capita Household Disposable Income):
  - New South Wales: Real per Capita 1990 104.5; 1995 104.8; 2001 106.9; Disposable Income 1990 106.8; 1995 105.7; 2001 107.4
  - Victoria: Real per Capita 1990 102.2; 1995 99.9; 2001 102.9; Disposable Income 1990 101.7; 1995 99.8; 2001 103.7
  - Queensland: Real per Capita 1990 85.7; 1995 87.3; 2001 86.7; Disposable Income 1990 84.3; 1995 86.3; 2001 89.0
  - South Australia: Real per Capita 1990 89.5; 1995 84.7; 2001 83.4; Disposable Income 1990 93.3; 1995 91.8; 2001 88.7
  - Western Australia: Real per Capita 1990 106.7; 1995 115.3; 2001 107.8; Disposable Income 1990 95.5; 1995 102.1; 2001 94.0
  - Tasmania: Real per Capita 1990 79.2; 1995 77.6; 2001 72.0; Disposable Income 1990 86.1; 1995 85.4; 2001 81.2
  - Northern Territory: Real per Capita 1990 128.0; 1995 120.4; 2001 119.1; Disposable Income 1990 102.2; 1995 111.9; 2001 97.0
  - ACT: Real per Capita 1990 127.2; 1995 123.9; 2001 125.9; Disposable Income 1990 145.1; 1995 144.0; 2001 135.1
- Labor market disparities:
  - Unemployment and median duration of unemployment are higher in below-average income/output states.
  - National averages (Australia): Unemployment Rate 1990 6.7; 1995 8.2; 2001 6.7. Median Duration of Unemployment 1990 16.3 weeks; 1995 28.6 weeks; 2001 18.0 weeks.
  - Examples of state-level values for 2001:
    - South Australia: Median duration 25.0 weeks.
    - Tasmania: Median duration 35.1 weeks.
  - Example: New South Wales unemployment 1990 6.1; 1995 6.7; 2001 5.4; median duration 1990 20.1; 1995 32.7; 2001 18.3.

### Institutional factors and labor market rigidities
- Industrial relations and reforms:
  - Prior to 1996: quasi-judicial conciliation and arbitration system; federal and state awards covered about two-thirds of wage and salary earners in 1990.
  - Awards: legally enforceable, industry/occupation-wide minimum terms and conditions; generally bind union and non-union employees.
  - Main disadvantage of the award system: reduces firm-level flexibility and may lower productivity growth by not providing adequate incentives to improve skills.
  - Workplace Relations Act of 1996: facilitated transition from centralized awards to enterprise bargaining; awards restricted to a safety net; number of “allowable matters” reduced to 20.
  - Award coverage fell from nearly two-thirds of employees in 1990 to 21 percent in 2001.
  - State-level variation: Victoria transferred its industrial relations system to federal system in 1996; other states retained separate systems, creating overlapping laws and awards.
- Empirical implication:
  - With a substantial proportion of wages—particularly at the low end—still based on awards, employers have less room to adjust wages for productivity differentials, potentially causing unemployment above average in lower-productivity areas.

### Empirical evidence on wage-setting and pay methods (May 2000)
- Method of Setting Pay — Percent of Employees (selected values):
  - New South Wales: Awards Only 23.9; Collective Agreement 34.4; Individual Agreement 41.7
  - Victoria: Awards Only 21.1; Collective Agreement 34.0; Individual Agreement 45.0
  - Queensland: Awards Only 24.6; Collective Agreement 41.8; Individual Agreement 33.6
  - South Australia: Awards Only 30.5; Collective Agreement 38.7; Individual Agreement 30.8
  - Western Australia: Awards Only 18.3; Collective Agreement 35.3; Individual Agreement 46.4
  - Tasmania: Awards Only 32.2; Collective Agreement 42.2; Individual Agreement 25.6
  - Northern Territory: Awards Only 24.4; Collective Agreement 41.5; Individual Agreement 34.1
  - ACT: Awards Only 16.2; Collective Agreement 59.4; Individual Agreement 24.4
  - Australia (aggregate): Awards Only 23.2; Collective Agreement 36.8; Individual Agreement 40.0
- Sectoral variation:
  - 65 percent of the recreational industry covered by awards only; less than 6 percent in mining and finance and insurance sectors.
  - Forty-two percent of clerical, sales, and service workers covered by awards; 3 percent of managers and administrators.

### Government transfers, equalization grants, and incentives
- Transfers per household (Ratio to Australian Average, percent):
  - New South Wales: 1990 103.3 / 1995 104.1 / 2001 108.2
  - Victoria: 1990 100.3 / 1995 99.6 / 2001 98.8
  - Queensland: 1990 89.9 / 1995 91.8 / 2001 93.2
  - South Australia: 1990 106.1 / 1995 101.5 / 2001 104.3
  - Western Australia: 1990 94.6 / 1995 94.9 / 2001 88.5
  - Tasmania: 1990 103.3 / 1995 106.8 / 2001 111.5
  - Northern Territory: 1990 85.6 / 1995 91.6 / 2001 67.1
  - ACT: 1990 99.3 / 1995 89.9 / 2001 89.1
- Transfers composition and effects:
  - Transfers data include workers compensation and various social assistance payments, but exclude old age pension.
  - Income support payments can undermine work incentives for those prone to long-term unemployment.
  - Uniformity in benefits across states may provide incentives for migration from low- to high-unemployment regions with lower living costs; enforcement of anti-migration provisions is characterized as lax.
  - Migration remains an important adjustment channel; out-migration from states with higher unemployment is documented.
- Equalization grants:
  - Aim to equalize fiscal resources across subnational governments to ensure minimum public service standards.
  - Potential disincentive: equalization may reduce state governments’ incentives to introduce reforms designed to enhance growth prospects.
- Empirical finding:
  - Federal grants did not have a significant impact on output growth across states in the 1990s.
  - The impact on real per capita income growth across states of government transfers appears to have increased since 1997.

### Commonwealth–State funding structure (2001/02 overview)
- Commonwealth grants to states in 2001/02:
  - About $A30 billion of General Purpose Payments (untied grants).
  - Some $A20 billion of Special Purpose Payments (tied grants, for specific purposes such as health, education, roads, and housing).
- Vertical fiscal issue:
  - Commonwealth dominance in tax collection while delivery responsibility for services lies at the state level creates a “vertical gap”.
- Commonwealth Grants Commission (CGC) and equalization:
  - CGC's horizontal fiscal equalization principle aims to give each state the same capacity to provide services.
  - Historical shift from “rough parity” to formal requirement of “absolute equality” of capacity.
  - Allocation logic: greater share of grants to states with below average capacity to raise own revenues or higher spending needs.
- Critiques:
  - Garnaut and FitzGerald (2002) estimate a deadweight loss of about $200 million per year arising primarily from high overheads and transaction costs.
  - System described as neither simple nor transparent; sometimes judgmental and not well understood beyond those involved.
  - Contrasting views argue the system has been effective in contributing to fiscal equalization.

### Empirical framework: income and output growth equations
- Income growth equation (DIG):
  - DIG_{i,t} = β_{0i} + β_{1}(DI)_{i,(t-1)} + β_{2}(TRANSFER)_{i,t} + β_{3}(SKILL)_{i,t} + ε_{i,t}
- Output growth equation (GSPG):
  - GSPG_{i,t} = β_{0i} + β_{1}(GSP)_{i,(t-1)} + β_{2}(LG)_{i,t} + β_{3}(KG)_{i,t} + β_{4}(TRANSFER)_{i,t} + β_{5}(GPP)_{i,t} + β_{6}(SKILL)_{i,t} + ε_{i,t}
- Variable definitions (selected):
  - DIG = [(1 + real per capita disposable income growth in state i) divided by (1 + real per capita disposable income growth in Australia)]-1.
  - GSPG = [(1 + real per capita GSP growth in state i) divided by (1 + real per capita GDP growth in Australia)]-1.
  - TRANSFER = ln [(real per capita government transfers to state i) divided by (the national average for real per capita government transfer)].
  - GPP = ln [(real per capita general purpose payments to state i) divided by (average real per capita general purpose payments in Australia)].
  - SKILL = ln [(ratio of skill vacancies in state i to total vacancies in state i) divided by (ratio of skill vacancies in Australia to total vacancies in Australia)].
- Data:
  - Pooled data for all Australian states and territories for the period 1990-2001. For regressions with SKILL, ACT excluded due to absence of skill data.

### Estimation results — labor market regressions (selected coefficients, Table 5)
- Unemployment ratio equation (dependent variable: Unemployment ratio)
  - Productivity-wage gap(-1): β = -12.441; P-value = 0.005
  - Transfers per labor force participant: β = -0.430; P-value = 0.862
  - Unemployment ratio(-1): β = 0.652; P-value = 0.000
  - Adjusted R-squared = 0.835
  - Interpretation:
    - Higher real wages relative to productivity (negative productivity-wage gap coefficient) are associated with higher unemployment.
    - Transfers variable has a counter-intuitive negative coefficient; possible explanation: transfers lower participation rates, reducing the measured unemployment rate by keeping people out of the labor force.
- Employment/Population equation (dependent variable: Employment/Population)
  - Productivity-wage gap(-1): β = 0.107; P-value = 0.000
  - Transfers per labor force participant: β = -0.087; P-value = 0.000
  - Employment/Population(-1): β = 0.646; P-value = 0.000
  - Skill bias: β = 0.005; P-value = 0.061
  - Adjusted R-squared = 0.878
  - Interpretation:
    - Both the labor market variable and government transfer variable have expected signs and are highly significant: states with above average government transfers tend to have lower employment to population ratios, implying transfers constrain work incentives and participation.
    - Higher relative skill levels are associated with higher employment-population ratios.
  - Note: Estimation excludes ACT for which skill bias data are not available.

### Detailed estimation results — Income Growth Equation (Table 6)
- Per capita disposable income (-1):
  - 1991–01: β = -0.265 (0.004)
  - 1991–96: β = -0.552 (0.002)
  - 1997–01: β = -0.704 (0.000)
- Real govt. transfers per household:
  - 1991–01: β = 0.451 (0.000)
  - 1991–96: β = 0.394 (0.027)
  - 1997–01: β = 0.766 (0.000)
- Skill bias:
  - 1991–01: β = 0.013 (0.361)
  - 1991–96: β = 0.043 (0.042)
  - 1997–01: β = -0.014 (0.279)
- Goodness of fit:
  - Adjusted R-squared: 0.374 (1991–01); 0.108 (1991–96); 0.798 (1997–01)
  - Degrees of Freedom: 67 (1991–01); 32 (1991–96); 25 (1997–01)
- Notes:
  - Pooled estimation with fixed effects; fixed effects are small and not presented here.
  - Estimation excludes ACT for which skill bias data are not available.

### Detailed estimation results — Output Growth Equation (Table 7)
- Per capita GSP ratio(-1):
  - 1991–01: β = -0.307 (0.000)
  - 1991–96: β = -0.512 (0.000)
  - 1997–01: β = -0.270 (0.074)
- Labor growth:
  - 1991–01: β = 0.366 (0.007)
  - 1991–96: β = 0.508 (0.000)
  - 1997–01: β = -0.155 (0.487)
- Investment per capita:
  - 1991–01: β = 0.063 (0.000)
  - 1991–96: β = 0.030 (0.492)
  - 1997–01: β = 0.083 (0.000)
- GPP ratio:
  - 1991–01: β = -0.001 (0.920)
  - 1991–96: β = -0.025 (0.637)
  - 1997–01: β = -0.051 (0.022)
- Real govt. transfers per capita:
  - 1991–01: β = -0.080 (0.189)
  - 1991–96: β = -0.311 (0.000)
  - 1997–01: β = 0.053 (0.655)
- Skill bias:
  - 1991–01: β = -0.002 (0.765)
  - 1991–96: β = -0.021 (0.139)
  - 1997–01: β = 0.004 (0.734)
- Goodness of fit:
  - Adjusted R-squared: 0.333 (1991–01); 0.537 (1991–96); 0.470 (1997–01)
  - Degrees of Freedom: 64 (1991–01); 29 (1991–96); 22 (1997–01)
- Notes:
  - Pooled estimation with fixed effects; fixed effects are small and not presented here.
  - Estimation excludes ACT for which skill bias data are not available.

### Key empirical findings (summary)
- Convergence:
  - Results show convergence in growth rates although most of the initial level differences between states are preserved.
  - States with lower income and output tend to grow faster (negative estimates on lagged income and output variables).
  - For the output equation, the catch-up coefficient for 1997–2001 is about half the size of the coefficient for 1991–1996, consistent with increased output dispersion since 1997.
- Transfers and convergence:
  - Government transfers have helped to reduce income disparities across states.
  - Transfers may have constrained or delayed convergence in output growth across states, possibly reflecting adverse effects on incentives to work.
- Other factors:
  - Relative labor force growth and capital accumulation explain most of the states’ output growth.
  - Labor is insignificant and has the wrong sign in the sub-sample 1997–2001; technological advances in the second half of the 1990s may not be fully accounted for.
  - Relative skill bias coefficient in 1997–2001 is positive but insignificant.
- Equalization principle effect:
  - General purpose payments coefficient is negative suggesting an adverse effect on output growth, although coefficients are statistically insignificant in two out of three regressions.

### Conclusions and policy-relevant implications
- Empirical conclusions:
  - Economic reforms of the last two decades have not brought about significant convergence in real per capita income and output across Australian states; differences in regional unemployment have also persisted.
  - Wage awards system may have impeded adjustment of real wages to productivity differentials and contributed to higher unemployment rates in some states.
  - Government transfers to households may have adversely affected work incentives in high unemployment states by limiting participation in the labor force.
  - Federal grants to the states do not appear to have had a significant impact on output growth across states.
  - Impact of skill-biased technological change on real per capita income and output growth is not clear.
- Policy considerations implied by the analysis:
  - Further decentralization of wage-setting and enhancement of firm-level/enterprise bargaining could improve wage flexibility and labor market adjustment.
  - Review of transfer and equalization arrangements to mitigate adverse work incentives while preserving minimum service standards.
  - Complementary policies to support skills upgrading and labor market participation in lower-income states to harness gains from technological change.
- Additional note:
  - Estimation of the output equation with National Competition Policy Payments (NCPP) for the subsample 1997–01 does not alter the above story—the NCPP remains an insignificant determinant of regional growth differentials.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf*

### References..............................................................................................................

### _wp04144 - References..............................................................................................................

### Introduction: purpose and empirical approach
- The paper examines why per capita income and output across Australian states did not converge during the 1990s despite comprehensive economic reforms.
- Empirical approach: panel estimation over the sample period 1990–2001.
- Hypothesized impediments to convergence:
  - Labor market rigidities (centralized wage/award system).
  - Government transfers (to households and state governments, including equalization grants).
  - Skill-biased technological change and adoption of new technologies.
- Key empirical summary statements:
  - The wage awards system may have restricted adjustment of real wages to productivity differentials and contributed to higher unemployment in some states.
  - Government transfers to households appear to have adversely affected work incentives in high-unemployment states by limiting labor force participation.
  - Growth in relatively low-income/output states partially converged toward higher-income states during the 1990s, but initial differences largely remained.
  - Since 1997, the catch-up effect on output growth has slowed, and the impact on real per capita income growth across states of government transfers appears to have increased.
  - Federal grants to the states did not have a significant impact on output growth across states in the 1990s.
  - The impact of skill-biased technological change on real per capita income and output growth is not clear.

### Stylized facts on regional disparities (1990–2001)
- Two groups of states emerge by real per capita output and income:
  - Above-national-average group: New South Wales, Victoria, Western Australia, Northern Territory, Australian Capital Territory (ACT).
  - Below-national-average group: Queensland, South Australia, Tasmania.
- Dispersion patterns:
  - Cross-state dispersion in output and income increased from the mid-1990s; dispersion in output far higher than in disposable income.
  - Income dispersion is largely influenced by the ACT; excluding the ACT, income dispersion is lower and more or less unchanged during the 1990s.
  - Output dispersion excluding the ACT rises after 1997.
- Selected numerical indicators from Table 1 (Real per Capita Ratio to Australian Average; Real per Capita Household Disposable Income):
  - New South Wales: 1990 104.5; 1995 104.8; 2001 106.9; Disposable Income: 1990 106.8; 1995 105.7; 2001 107.4
  - Victoria: Real per Capita 1990 102.2; 1995 99.9; 2001 102.9; Disposable Income 1990 101.7; 1995 99.8; 2001 103.7
  - Queensland: Real per Capita 1990 85.7; 1995 87.3; 2001 86.7; Disposable Income 1990 84.3; 1995 86.3; 2001 89.0
  - South Australia: Real per Capita 1990 89.5; 1995 84.7; 2001 83.4; Disposable Income 1990 93.3; 1995 91.8; 2001 88.7
  - Western Australia: Real per Capita 1990 106.7; 1995 115.3; 2001 107.8; Disposable Income 1990 95.5; 1995 102.1; 2001 94.0
  - Tasmania: Real per Capita 1990 79.2; 1995 77.6; 2001 72.0; Disposable Income 1990 86.1; 1995 85.4; 2001 81.2
  - Northern Territory: Real per Capita 1990 128.0; 1995 120.4; 2001 119.1; Disposable Income 1990 102.2; 1995 111.9; 2001 97.0
  - ACT: Real per Capita 1990 127.2; 1995 123.9; 2001 125.9; Disposable Income 1990 145.1; 1995 144.0; 2001 135.1
- Labor market disparities:
  - Unemployment and median duration of unemployment are higher in below-average income/output states.
  - National averages (Australia): Unemployment Rate 1990 6.7; 1995 8.2; 2001 6.7. Median Duration of Unemployment 1990 16.3 weeks; 1995 28.6 weeks; 2001 18.0 weeks.
  - Examples of state-level values for 2001:
    - South Australia: Median duration 25.0 weeks.
    - Tasmania: Median duration 35.1 weeks.
  - Table 2 highlights variability: New South Wales unemployment 1990 6.1; 1995 6.7; 2001 5.4; median duration 1990 20.1; 1995 32.7; 2001 18.3.

### Institutional factors and labor market rigidities
- Industrial relations context (Box 1) — pre-1996 and reforms:
  - Prior to 1996: quasi-judicial conciliation and arbitration system; federal and state awards covered about two-thirds of wage and salary earners in 1990.
  - Awards: legally enforceable, industry/occupation-wide minimum terms and conditions; generally bind union and non-union employees.
  - Main disadvantage of the award system: reduces firm-level flexibility and may lower productivity growth by not providing adequate incentives to improve skills.
  - Workplace Relations Act of 1996: facilitated transition from centralized awards to enterprise bargaining; awards restricted to a safety net; number of “allowable matters” reduced to 20.
  - Award coverage fell from nearly two-thirds of employees in 1990 to 21 percent in 2001.
  - State-level variation: Victoria transferred its industrial relations system to federal system in 1996; other states retained separate systems, creating overlapping laws and awards.
- Empirical implication: with a substantial proportion of wages—particularly at the low end—still based on awards, employers have less room to adjust wages for productivity differentials, potentially causing unemployment above average in lower-productivity areas.

### Empirical evidence on wage-setting and pay methods
- Table 3 (Method of Setting Pay, May 2000; Percent of Employees):
  - New South Wales: Awards Only 23.9; Collective Agreement 34.4; Individual Agreement 41.7
  - Victoria: Awards Only 21.1; Collective Agreement 34.0; Individual Agreement 45.0
  - Queensland: Awards Only 24.6; Collective Agreement 41.8; Individual Agreement 33.6
  - South Australia: Awards Only 30.5; Collective Agreement 38.7; Individual Agreement 30.8
  - Western Australia: Awards Only 18.3; Collective Agreement 35.3; Individual Agreement 46.4
  - Tasmania: Awards Only 32.2; Collective Agreement 42.2; Individual Agreement 25.6
  - Northern Territory: Awards Only 24.4; Collective Agreement 41.5; Individual Agreement 34.1
  - ACT: Awards Only 16.2; Collective Agreement 59.4; Individual Agreement 24.4
  - Australia (aggregate): Awards Only 23.2; Collective Agreement 36.8; Individual Agreement 40.0
- Sectoral variation (textual example): 65 percent of the recreational industry covered by awards only; less than 6 percent in mining and finance and insurance sectors. Forty-two percent of clerical, sales, and service workers covered by awards; 3 percent of managers and administrators.

### Government transfers, equalization grants, and incentives
- Government transfers per household are higher in states with the largest output and unemployment gaps (Table 4 referenced in text).
- Transfers composition and effects:
  - Transfers data include workers compensation and various social assistance payments, but exclude old age pension.
  - Income support payments can undermine work incentives for those prone to long-term unemployment (Industry Commission, 1993).
  - Uniformity in benefits across states may provide incentives for migration from low- to high-unemployment regions with lower living costs; enforcement of anti-migration provisions in Social Security Act characterized as lax and mainly applied to those already receiving benefits.
  - Migration remains an important adjustment channel; Debelle and Vickery (1999) and Lawson and Dwyer (2002) find out-migration from states with higher unemployment.
- Equalization grants:
  - The equalization principle in federal grants aims to equalize fiscal resources across subnational governments to ensure minimum public service standards.
  - Potential disincentive: equalization may reduce state governments’ incentives to introduce reforms designed to enhance growth prospects (Craig, 1997).
- Empirical findings:
  - Federal grants did not have a significant impact on output growth across states in the 1990s.
  - The impact on real per capita income growth across states of government transfers appears to have increased since 1997.

### Overall conclusions and policy-relevant implications
- Labor market institutions:
  - Centralized award-based wage setting limited regional wage adjustment to productivity differentials, contributing to higher unemployment in lower-productivity states.
  - Workplace Relations Act of 1996 reduced award coverage substantially (from nearly two-thirds in 1990 to 21 percent in 2001), but award-based wage setting remained substantial—especially in lower-income states.
- Transfers and incentives:
  - Household transfers can weaken work incentives and labor force participation in high-unemployment states.
  - Equalization grants may blunt state-level incentives to pursue growth-enhancing reforms.
- Dynamics of convergence:
  - Some convergence in growth of low-income/output states occurred during the 1990s, but initial gaps largely persisted; catch-up on output slowed after 1997.
  - The role of skill-biased technological change in driving divergence is inconclusive in this analysis.
- Policy considerations implied by the analysis:
  - Further decentralization of wage-setting and enhancement of firm-level/enterprise bargaining could improve wage flexibility and labor market adjustment.
  - Review of transfer and equalization arrangements to mitigate adverse work incentives while preserving minimum service standards.
  - Complementary policies to support skills upgrading and labor market participation in lower-income states to harness gains from technological change.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf*

### Box 2 provides an overview of the current system for Commonwealth-State funding and its

### _wp04144 - Box 2 provides an overview of the current system for Commonwealth-State funding and its

### Overview of Commonwealth-State funding system
- Grants are allocated to the states on the basis of each state’s ability to generate its own-source revenue in an attempt to at least partially equalize revenue across states, so that each state has sufficient resources to provide a minimum standard of public services.
- If a policy change improves a state’s revenue, and the state is assessed as having above average capacity, then federal grants would decline, partially offsetting the revenue gain.
- Consequently, this grants arrangement may impede actions which would boost the real per capita income and output levels of the lower income states.
- A study was commissioned by the governments of New South Wales, Victoria, and Western Australia to review the methods of allocating Commonwealth grants to the states and territories, and the appropriateness of the outcomes (see Garnaut and FitzGerald (2001 and 2002)).
- The Ministerial Council on Commonwealth State Financial Relations is also considering a review of the methodology (see Australian Government, 2004).

### Table 4 — Ratio of Real Government Transfers Per Household to the Australian Average (In percent)
- 1990 / 1995 / 2001
  - New South Wales: 103.3 / 104.1 / 108.2
  - Victoria: 100.3 / 99.6 / 98.8
  - Queensland: 89.9 / 91.8 / 93.2
  - South Australia: 106.1 / 101.5 / 104.3
  - Western Australia: 94.6 / 94.9 / 88.5
  - Tasmania: 103.3 / 106.8 / 111.5
  - Northern Territory: 85.6 / 91.6 / 67.1
  - ACT: 99.3 / 89.9 / 89.1
- Sources: Australian Bureau of Statistics; and IMF staff calculations.

### Limitations and incentive effects of the grants arrangement
- The decline in federal grants following an assessed increase in a state's revenue capacity creates high effective marginal tax rates on policy-induced revenue gains.
- This situation is roughly comparable to that faced by welfare recipients who confront high effective marginal income tax rates owing to the withdrawal of benefits when they return to work.

### Empirical evidence — Labor market regressions and variables
- Estimated pooled regression (with fixed effects) used to explain dispersion of unemployment across states:
  - UR_i,t = β_0i + β_1 (PWGAP)_{i,(t-1)} + β_2 (TRANSFER)_{i,t} + β_3 UR_{i,(t-1)} + ε_{i,t}
  - Definitions:
    - UR is the unemployment rate in state i relative to the national average.
    - PWGAP is the differential between productivity and real wages in state i relative to the national level (indicative of the influence of wage-setting arrangements).
    - TRANSFER is real government transfers per labor force participant in state i relative to the national average.
  - Rationale:
    - Contemporaneous information on the government transfer variable is usually available to labor force participants in making their work-leisure decisions.
    - The labor productivity-wage gap variable is likely to impact the unemployment rate with a lag.
    - The lagged dependent variable controls for persistence or hysteresis in unemployment.
- Data limitation note:
  - Due to data limitations on the components of social assistance at the state level, old age pensions are subtracted from the transfer data by applying the ratio of old age pensions to total social assistance at the Commonwealth government level to the state level data. This implicitly assumes that the ratio of old age transfers across all states are the same.
  - The Annex provides further detailed data definitions.

### Estimation results and interpretation (Table 5)
- Unemployment ratio equation (dependent variable: Unemployment ratio)
  - Productivity-wage gap(-1): β-coefficient = -12.441; P-value = 0.005
  - Transfers per labor force participant: β-coefficient = -0.430; P-value = 0.862
  - Unemployment ratio(-1): β-coefficient = 0.652; P-value = 0.000
  - Adjusted R-squared = 0.835
  - Interpretation:
    - Higher real wages relative to productivity (negative productivity-wage gap coefficient) are associated with higher unemployment.
    - The transfers variable has a counter-intuitive negative coefficient (i.e., higher transfers associated with lower relative unemployment); possible explanation: transfers lower participation rates, reducing the measured unemployment rate by keeping people out of the labor force.
- Employment/Population equation (dependent variable: Employment/Population)
  - Productivity-wage gap(-1): β-coefficient = 0.107; P-value = 0.000
  - Transfers per labor force participant: β-coefficient = -0.087; P-value = 0.000
  - Employment/Population(-1): β-coefficient = 0.646; P-value = 0.000
  - Skill bias: β-coefficient = 0.005; P-value = 0.061
  - Adjusted R-squared = 0.878
  - Interpretation:
    - Both the labor market variable and government transfer variable have the expected signs and are highly significant: states with above average government transfers tend to have lower employment to population ratios, implying transfers constrain work incentives and participation.
    - Higher relative skill levels are associated with higher employment-population ratios.
  - Note: Estimation excludes Australian Capital Territory for which skill bias data are not available.
- Note on statistical significance: P-values are the probability values for the null hypothesis of a coefficient equal to zero. Thus, a value of .005 indicates coefficient significance at less than one percent level.
- Methodology note: Results are for pooled estimation with fixed effects; fixed effects are small and not presented here.

*Sources: Australian Bureau of Statistics; and IMF staff calculations.*

### Box 2. Overview of the Current System of Intergovernmental Transfers

### Box 2. Overview of the Current System of Intergovernmental Transfers

### Overview of Commonwealth–State Funding Structure
- In 2001/02, Commonwealth grants to states included about $A30 billion of General Purpose Payments (untied grants) and some $A20 billion of Special Purpose Payments (tied grants, for specific purposes such as health, education, roads, and housing).
- Heavy reliance of states on the Commonwealth reflects the Commonwealth’s dominance in tax collection while delivery responsibility for important services lies at the state level, creating a “vertical gap” for state governments in terms of revenue collection and expenditure responsibilities.
- There is an underlying inter-governmental agreement between the Commonwealth and the states.

### Commonwealth Grants Commission (CGC) and the Equalization Principle
- The CGC’s horizontal fiscal equalization principle aims to give each state the same capacity to provide services.
- Historical shift: from “rough parity” to a formal requirement of achieving “absolute equality” of capacity among states.
- CGC funding logic: if each state made the same effort to raise revenue from its own resources and operated at the same level of efficiency, each would have the capacity to provide the same level and standard of service.
- Allocation implication: a greater share of grants goes to states with below average capacity to raise their own revenues and/or need to spend more to provide the same standard of services as other states.
- Objective of “policy neutrality” seeks to avoid states influencing grant amounts through policy choices; critics argue the CGC methodology may not be policy neutral and that states can influence long-term need for services and revenue raising capacity.

### Critiques of the Current System: Equity, Efficiency, Transparency
- Efficiency critique: Garnaut and FitzGerald (2002) estimate a deadweight loss of about $200 million per year arising primarily from high overheads and transaction costs of administering the system.
- Other critiques:
  - System described as neither simple nor transparent, often judgmental, and not well understood beyond those involved.
- Contrasting views: Buchanan and Wagner (1971) and Mathews (1983) argue that the Australian intergovernmental transfer system has been effective and achieved considerable success in contributing to fiscal equalization.

### Empirical Framework: Income and Output Growth Equations
- Income growth equation (DIG):
  - DIG_{i,t} = β_{0i} + β_{1}(DI)_{i,(t-1)} + β_{2}(TRANSFER)_{i,t} + β_{3}(SKILL)_{i,t} + ε_{i,t}
- Output growth equation (GSPG):
  - GSPG_{i,t} = β_{0i} + β_{1}(GSP)_{i,(t-1)} + β_{2}(LG)_{i,t} + β_{3}(KG)_{i,t} + β_{4}(TRANSFER)_{i,t} + β_{5}(GPP)_{i,t} + β_{6}(SKILL)_{i,t} + ε_{i,t}
- Variable definitions (selected):
  - DIG = [(1 + real per capita disposable income growth in state i) divided by (1 + real per capita disposable income growth in Australia)]-1.
  - GSPG = [(1 + real per capita GSP growth in state i) divided by (1 + real per capita GDP growth in Australia)]-1.
  - TRANSFER = ln [(real per capita government transfers to state i) divided by (the national average for real per capita government transfer)].
  - GPP = ln [(real per capita general purpose payments to state i) divided by (average real per capita general purpose payments in Australia)].
  - SKILL = ln [(ratio of skill vacancies in state i to total vacancies in state i) divided by (ratio of skill vacancies in Australia to total vacancies in Australia)].
- Data set: Pooled data for all Australian states and territories for the period 1990-2001. For regressions with the skill bias variable, the Australian Capital Territory has been excluded due to absence of skill data.
- Data Sources: Australian Bureau of Statistics, Department of Employment and Workplace Relations, and the Australian Treasury.

### Key Empirical Findings (Summary)
- Convergence:
  - Results show convergence in growth rates although most of the initial level differences between states are preserved.
  - States with lower income and output tend to grow faster (negative estimates on lagged income and output variables).
  - For the output equation, the catch-up coefficient for 1997–2001 is about half the size of the coefficient for 1991–1996, consistent with increased output dispersion since 1997.
- Transfers and convergence:
  - Government transfers have helped to reduce income disparities across states.
  - Transfers may have constrained or delayed convergence in output growth across states, possibly reflecting adverse effects on incentives to work.
- Other factors:
  - Relative labor force growth and capital accumulation explain most of the states’ output growth.
  - Labor is insignificant and has the wrong sign in the sub-sample 1997–2001; technological advances in the second half of the 1990s may not be fully accounted for.
  - Relative skill bias coefficient in 1997–2001 is positive but insignificant, suggesting higher skill bias implies higher output growth though not robustly significant.
- Equalization principle effect:
  - General purpose payments coefficient is negative suggesting an adverse effect on output growth, although coefficients are statistically insignificant in two out of three regressions.

### Detailed Estimation Results — Table 6. Income Growth Equation (Dependent Variable: Growth in Per Household Disposable Income)
- Independent variables — β-coefficient (with p-value in parentheses)
  - Per capita disposable income (-1):
    - 1991–01: -0.265 (0.004)
    - 1991–96: -0.552 (0.002)
    - 1997–01: -0.704 (0.000)
  - Real govt. transfers per household:
    - 1991–01: 0.451 (0.000)
    - 1991–96: 0.394 (0.027)
    - 1997–01: 0.766 (0.000)
  - Skill bias:
    - 1991–01: 0.013 (0.361)
    - 1991–96: 0.043 (0.042)
    - 1997–01: -0.014 (0.279)
- Goodness of fit and sample:
  - Adjusted R-squared: 0.374 (1991–01); 0.108 (1991–96); 0.798 (1997–01)
  - Degrees of Freedom: 67 (1991–01); 32 (1991–96); 25 (1997–01)
- Notes:
  - Results for pooled estimation with fixed effects; fixed effects are small and not presented here.
  - Estimation excludes Australian Capital Territory for which skill bias data are not available.
  - P-values for the null hypothesis of a coefficient equal to zero are within parentheses.

### Detailed Estimation Results — Table 7. Output Growth Equation (Dependent Variable: Growth in Per Capita Gross State Product)
- Independent variables — β-coefficient (with p-value in parentheses)
  - Per capita GSP ratio(-1):
    - 1991–01: -0.307 (0.000)
    - 1991–96: -0.512 (0.000)
    - 1997–01: -0.270 (0.074)
  - Labor growth:
    - 1991–01: 0.366 (0.007)
    - 1991–96: 0.508 (0.000)
    - 1997–01: -0.155 (0.487)
  - Investment per capita:
    - 1991–01: 0.063 (0.000)
    - 1991–96: 0.030 (0.492)
    - 1997–01: 0.083 (0.000)
  - GPP ratio:
    - 1991–01: -0.001 (0.920)
    - 1991–96: -0.025 (0.637)
    - 1997–01: -0.051 (0.022)
  - Real govt. transfers per capita:
    - 1991–01: -0.080 (0.189)
    - 1991–96: -0.311 (0.000)
    - 1997–01: 0.053 (0.655)
  - Skill bias:
    - 1991–01: -0.002 (0.765)
    - 1991–96: -0.021 (0.139)
    - 1997–01: 0.004 (0.734)
- Goodness of fit and sample:
  - Adjusted R-squared: 0.333 (1991–01); 0.537 (1991–96); 0.470 (1997–01)
  - Degrees of Freedom: 64 (1991–01); 29 (1991–96); 22 (1997–01)
- Notes:
  - Results for pooled estimation with fixed effects; fixed effects are small and not presented here.
  - Estimation excludes Australian Capital Territory for which skill bias data are not available.
  - P-values for the null hypothesis of a coefficient equal to zero are within parentheses.

### Conclusions (from the source)
- Economic reforms of the last two decades have not brought about significant convergence in real per capita income and output across Australian states; differences in regional unemployment have also persisted.
- Empirical analysis considered labor market rigidities, government transfers to households and state governments, and increased adoption of new technologies in some states relative to others as potential contributors to regional disparities.
- Results suggest:
  - Wage awards system may have impeded adjustment of real wages to productivity differentials and contributed to higher unemployment rates in some states.
  - Government transfers to households may have adversely affected work incentives in high unemployment states by limiting participation in the labor force.
  - Federal grants to the states do not appear to have had a significant impact on output growth across states.
  - Impact of skill-biased technological change on real per capita income and output growth is not clear.
- Additional note: Estimation of the output equation with National Competition Policy Payments (NCPP) for the subsample 1997–01 does not alter the above story—the NCPP remains an insignificant determinant of regional growth differentials.

*Source: Box 2. Overview of the Current System of Intergovernmental Transfers (content unit), IMF Working Paper excerpt.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04144.pdf_