During Victoria’s outbreak of influenza A(H1N1)pdm09 in 2009, parents appeared to be somewhat more likely to take time off work to care for their children when a parent in the household had access to paid sick or carer’s leave, compared to households without access to leave, but there is insufficient statistical evidence to reject the null hypothesis of no difference. Taking time off work was associated with two indicators of compliance with quarantine recommendations: quarantined children staying home for most of the time on all days and not making trips to populated places. However, this study found no evidence that access to leave, per se, was associated with overall compliance with quarantine recommendations. On the other hand, lack of access to leave had measurable negative impacts on families. In households without this benefit available, nearly three-quarters had a parent who lost pay, compared to one in five households with leave, and nearly 40% of households that lost pay experienced further financial difficulties as a consequence.
The chief explanation for the lack of association between access to leave and compliance with quarantine appears to be that families frequently chose to follow public health recommendations even when that meant absorbing the collateral employment-related effects due to inadequate leave entitlements: in 42% of households that did not have access to leave, a parent still took time off work to care for the quarantined child. This behavioural response is particularly selfless in light of the fact that financial consequences are borne privately whereas the benefits of home quarantine and social distancing measures accrue to the community in the form of reduced risks of transmission. While some of this behaviour may have been driven by the need to care for sick children, there were no confirmed influenza A(H1N1)pdm09 diagnoses in the vast majority (94%) of households in our sample. This suggests that, absent the strict quarantine restrictions, other childcare options may well have been attractive to parents to enable them to attend work during the period of school closure. Twenty-two per cent of households where a parent did have access to leave still lost pay as a result of taking time off work. The likely explanation is that, because leave was defined at a household level, a parent without access to leave was the one who took time off work.
Our study is the first we know of to have considered the effect of parental leave entitlements on quarantine compliance during the 2009 outbreak of influenza A(H1N1)pdm09. In Western Australian school closures during this outbreak, a parent took time off work in 45% of households
 — a similar finding to our study. However the Western Australian study did not examine whether time taken off work influenced compliance or whether taking leave had a financial impact. Our finding contrasts with findings from studies in the US, both hypothetical and real, which have suggested a lack of access to paid sick leave is a barrier to social distancing
The study had several limitations. First, despite beginning with a sample frame consisting of all households in Victoria affected by school closures, our relatively small analytic sample meant the study was underpowered to detect differences unless they were large. A good example of this is the relationship between parents’ access to leave and their decision to take time off work to care for their children; the difference in proportions was substantial (16 percentage points) but did not attain statistical significance, likely due to the small sample size. Second, our response rate was not high, despite the use of incentives to boost participation rates. This has implications both for power and the risk of Type II errors. Nonetheless, the response rate is comparable to that achieved in other similar school-based studies of pandemic influenza in the US and England
[13–15] and our study had the advantage of covering a larger number of affected schools than most other studies. As we showed in an earlier publication from this study, we received a disproportionately low level of response from less advantaged schools, reducing the generalisability of our findings and potentially biasing our results
. It could be expected that non-responding households were less likely to have access to paid leave and may have experienced greater financial consequences, resulting in the estimates presented in this paper being conservative. Unfortunately, the survey had to be administered through schools due to privacy constraints, and we are therefore not able to characterize non-respondents in more detail.
The study was also limited by the fact that the survey was administered several months after the school closures occurred, and all information was obtained via self-report, introducing the possibility of recall bias. In some cases, parents were reporting on behaviours of their children at times when parents may not have been present.
All pandemic plans must balance the likely benefits and social and economic costs of implementing social distancing measures. Characterising the costs incurred by families during quarantine and social distancing of school children during Victoria’s 2009 outbreak of pandemic influenza contributes to the evidence base for future assessment of the costs and benefits of these containment strategies. Models of pandemic influenza have shown that the greatest impact of school closure on transmission is observed when closures are widespread, initiated early, and sustained beyond the epidemic peak
[16–18]. In Victoria, school closure was localised, short-lived (often less than 7 days) and reactively initiated following case identification.
In households where parents are forced to take leave from work due to public health emergencies, foregoing wages is a high price to pay for honouring a public duty. Employers should be encouraged to provide flexible working arrangements, such as allowing employees to work from home or to make up hours at a later date. Setting aside the question of whether access to paid sick leave should be available to all workers, there are strong ethical arguments
 and community support
 for the provision of compensation to individuals who experience loss of income as a result of public health measures such as quarantine. Policy initiatives along these lines are not unprecedented: several countries affected by the SARS outbreak introduced some form of compensation for affected households
. In Australia, this might involve government and employers sharing the costs of compensating quarantined employees. This could operate similarly to the current legislated arrangements for jury service, whereby employers are required to release employees for jury service and pay them the difference between the set jury payment provided by the courts and what they would have received as earnings for that period had they not been on jury service