Recent Trends in Community Economic Resilience in Australia Huong Dinh, Ben Freyens, Anne Daly, Yogi Vidyattama 22 June 2015
Background • Major shocks during 2006-2011 – The Millennium Drought – The Global Financial Crisis – The Mining Boom • A need to build or reinforce the resilience of local communities – Counteract negative impacts – Adapt to changed conditions • A rising call for measuring resilience capacity at community level – Identify communities of concern with low or declining levels of resilience – Assist policy makers to develop supportive strategies
Research Questions • What are the patterns of change in community economic resilience (CER) and its components across states, regions and over time? • How do the components of CER interact? • Is there a long-term association between CER and household income?
CER constructs, data and variables % population with a broadband connection % households with a four or more-bedroom private % people aged 15 and over with a university degree or other dwelling tertiary qualification proportion of households with at least three motor % people aged 15 and over with an advanced diploma or vehicles diploma qualification % households with a high mortgage repayment % employed people classified as professionals % households with a high rent payment % employed people who work at Skill Level 1 Physical capital % population in the national 9 th or 10 th Financial capital income deciles Human capital % of adult population in the labour force % population with internet Household affordability ratio (the use CER fraction of household income over rental % population participating Social capital payment) in voluntary work Diversity % population with English Comparing the community-level proficiency sector-specific employment rate Natural capital with the national rate Accessibility % employed people working in agricultural and mining Remoteness and accessibility in terms of access along the road industries network from populated localities to each the five categories of Service Centres Spatial unit: SA Level 1 ARIA Census
Data exclusion – Exclude SA1 with any of those characteristics: • population is less than 10; • any selected denominator variables are less than six • more than 80% of its population live in non- private dwellings.
Index computation – Principal component analysis: • turns a large number of highly correlated variables into a smaller set of independently transformed variables – Indexes range from 0 (low resilience) to 100 (high resilience) – Working data: Indexes computed for 47,490 SA1s in 2006 and 44,889 SA1s in 2011 out of a grand total of 54,805.
Fixed Effect Estimation • Adjusting for SA1-specific/state-specific ( 𝑏 𝑗 / 𝜄 𝑘 ) effects Testing if the indexes ( 𝑍 • 𝑗𝑢 ) change over time ( 2011 vs 2006): 𝑍 𝑗𝑢 = 𝑏 𝑗 +∝ 𝑢 𝑢 + 𝜁 𝑗𝑢 Testing the association between capital-related indexes ( 𝐷 𝑗𝑢 ) and diversity • and accessibility indexes ( 𝑌 𝑗𝑢 ): 𝐷 𝑗𝑢 = 𝑏 𝑗 +∝ 𝑢 𝑢 + 𝛾𝑌 𝑗𝑢 + 𝜁 𝑗𝑢 Testing the association between household income ( 𝐽 𝑗𝑢 ) and CER index • ( 𝐷 𝑗,𝑢−1 ): 𝑚𝑝𝐽 𝑗𝑢 = 𝛿 + 𝜄 𝑘 𝑇 𝑘 + 𝜀𝑚𝑝𝐽 𝑗,𝑢−1 + 𝜍𝐷𝐹𝑆 𝑗,𝑢−1 + 𝜏 𝑗𝑢
Human Capital Index 60 40 20 0 2006 2011
Social capital index 70 65 60 55 2006 2011
Natural Capital Index 15 10 5 0 2006 2011
Physical Capital Index 80 60 40 20 0 2006 2011
Financial Capital Index 45 40 35 2006 2011
Diversity Index 100 50 0 2006 2011
Accessibility Index 150 100 50 0 2006 2011
CER Index 55 50 45 40 2006 2011
CER Capital-Related Constructs by Level of Diversity Level of diversity Mean difference test Low High Mean SE Mean SE Mean diff. SE p-value Human 31.2 0.08 31.4 0.06 0.23 0.05 <0.001 Social 62.0 0.03 62.3 0.02 0.27 0.05 <0.001 Natural 5.2 0.02 4.5 0.01 -0.64 0.03 <0.001 Physical 38.4 0.06 40.0 0.03 1.58 0.08 <0.001 Financial 39.5 0.02 39.8 0.01 0.33 0.02 <0.001
CER Capital-Related Constructs by Level of Accessibility Level of accessibility Mean difference test Low High Mean Mean SE Mean SE SE p-value diff. Human 25.4 0.12 33.4 0.08 8.0 0.15 <0.001 Social 65.8 0.06 61.0 0.04 -4.7 0.07 <0.001 Natural 5.3 0.12 4.6 0.04 -0.7 0.17 <0.001 Physical 31.4 0.12 42.4 0.09 11.0 0.15 <0.001 Financial 39.5 0.03 39.8 0.01 0.4 0.03 <0.001
Association Between Household Income and CER Log of household income Coef. SE t P>|t| Lagged CER 0.04 0.0003 159.50 0.000 Lagged log of household income 0.08 0.0021 41.00 0.000 State (omitted) Cons -1.87 0.0152 -122.95 0.000 R squared 0.55 Number of observations 46012
Conclusions • There was a statistically significant improvement in the overall CER index in all states and regions. – CER improved over time in human, social and physical capital terms, declined in diversity terms. – Natural index improved in all regions and states except ACT and VIC. – Financial index declined in all regions and states except ACT. – Accessibility index improved in most states except VIC and TAS (unchanged) and the MDB (reduced)
Conclusions (cont.) • Communities with a higher level of economic diversity had higher human, social, physical and financial capitals but lower natural capital, • Communities with a lower level of accessibility reported lower levels of human, physical and financial capital but higher levels of social and natural capital. • CER has a long-term positive impact on household income
Recommend
More recommend
