A credible public proposal for time and opportunity

Executive summary

Australia could pursue a genuine four-day working week at unchanged weekly pay as part of a broader personal AI opportunity programme. The credible public commitment is to fund and negotiate the transition sector by sector, protect services and earnings, and give individuals a usable AI allowance independent of employment or education. This report proposes how the One World Community could put that choice before the public before the next federal election.

The campaign should seek a funded policy commitment and broad, informed support. It cannot responsibly promise that a particular number of impressions, petition signatures or supporters will secure an election commitment. Support must be measured after people see costs, alternatives and implementation limits. A3 therefore supplies a public explanation, an adoption sequence, a measurement brief and an illustrative A$1.2m campaign envelope. None is an instruction to launch or spend.

A1’s trials and historical analysis justify serious implementation work, particularly on actual hours, wellbeing and the institutions needed to secure change. The campaign is now explicitly multi-step: establish starting arrangements, win a funded transition commitment, negotiate and demonstrate sector offers, then secure successive reductions toward 2035. A2 supplies differentiated targets instead of treating 32 hours as the same gain for every worker. A2’s coverage analysis prevents a promise of a universally costless transition. B2’s central AI scenario is A$10.6bn annually, rising to A$26.55bn with full participation and redemption under the same assumptions; working-time reform and retained existing expenditure are additional. National sector costs remain unresolved. These are material facts for public discussion, not details to defer until after support is secured.

Reader orientation, premises and confidence

This is the third report in the time trilogy and the fifth report in the combined programme. A1 appraises trials; A2 designs fair time arrangements; B3 integrates the demonstration and route to 2035. The four pathways are alternative futures, not four compulsory stages. The campaign steps C1–C5 below describe actions over time; D0–D2 in B3 describe demonstration stages. Neither sequence is a relabelling of WF1–WF4.

ASSUMED FOR DESIGN — P1: Some people use shorter working hours and free AI access to experiment. P2: Individual learning by doing is the best foundation for discovering personal AI uses. P3: That capability can support employment and social-cohesion resilience. These accepted premises organise the proposal; they are not findings that the national programme has already succeeded.

Confidence is high in the conditional payroll arithmetic and the dated official rules cited here; medium in transferring selected workplace evidence; unresolved in national affordability, informed voter support, political adoption and employment/cohesion effects. Campaign choices below are design recommendations. No forecast of AI job losses is used to manufacture urgency or a funding return.

Pass 0 — State the public offer and the person it serves

Illustrative 2035 experience. An employee has genuinely fewer paid working hours, with usual ordinary weekly earnings protected. On some weeks she rests or cares for family. On another she uses her personal AI allowance to explore a new service idea. Her employer cannot require a learning course on the released day. Her neighbour, who is unemployed, has the same basic AI access. Neither must join the One World Community or submit a business plan.

The proposed public explanation is:

Australia should give people time and practical opportunity to adapt to AI. We propose a funded, staged transition toward a four-day working week with ordinary weekly pay protected, alongside personal AI access for citizens and permanent residents aged 16 and over. Starting hours and rosters differ. The proposed destination is 20% fewer ordinary hours from each verified existing full-time arrangement, with a genuine four-day pattern or an agreed equivalent for other rosters. Where needed, begin with a funded 10% reduction and a dated further step; keep essential services staffed. People would choose their own legal uses and could seek optional backing for ideas. Employers, colleges and community organisations would not control the basic allowance. Trials and earlier hours reforms support further implementation work, but national staffing costs and the effects of this combined programme still need to be established.

This is a research report for a broad public audience. It is not campaign copy. “Free” means no charge to the person within the funded allowance, not unlimited compute or no taxpayer cost. Provider terms, age suitability and lawful service operation must be resolved without adding a productivity or course test. A genuine hours reduction is distinct from compressing existing hours into four longer days. For a 38-hour baseline, a 20% reduction means 30.4 hours; a 32-hour week is a different reduction. A2’s verified examples also include 35 hours in black coal and 37.5 at PM&C: their proportional targets are 28 and 30. The named instruments do not describe every mining or public-service employee. A common 32-hour endpoint remains a comparison for informed debate, not an interchangeable description of this proposal. SH02: Black Coal Award, clause 14; SH03: PM&C agreement, clause 91

The central age case remains 16+, without an upper age. The campaign must explain the separately verified youth route. Citizens and permanent residents are programme beneficiaries; the electorate used to measure voter support is a different population. Permanent residents and 16–17-year-olds should be heard about service design without being counted automatically as federal voters.

Pass 1 — Use trial and historical evidence without overstating it

The evidence brief should retain three distinctions from the preceding reports:

The study and original evaluation are available directly; A1/A2 contain the fuller appraisal and denominators. A3-01: Fan et al., A3-02: council evaluation

History adds a political and institutional lesson. Australia's arbitration route, German bargaining and Sweden's staged legislation show different ways to turn demand into operative hours. They do not show that a public campaign alone caused reform or that an Australian election commitment guarantees delivery. France and Portugal further show why pay, subsidies and flexibility belong in the public explanation. The fuller evidence, including adverse findings and the French correction, is in A1 and the historical background. HH02: Australian order; HH03: German agreement chronology; HH08: Swedish parliamentary record

The strongest implementation objection deserves a numerical answer. If required output is unchanged, hours fall 20%, and effective output per hour does not improve, the affected payroll rises 25% to cover the work. With a 10% effective gain it rises about 13.64%. These are modelled consequences, not observed national costs. Direct-care requirements cannot be replaced by assumed efficiency. The public proposal should publish the sector costing work as it becomes available, including who pays and where workforce supply constrains delivery.

The initial 10% step still requires about 11.11% additional covered payroll in the same no-gain constant-output model. Staging spreads the transition and can create learning opportunities; it does not make the final recurring cost disappear. Cost the actual target and start date for each covered workforce. A2 compares the smaller benefits and costs of a fixed 32-hour destination.

The campaign should also explain why this is an additional route beyond traditional education. A person can explore a problem before knowing what qualification, occupation or business might fit. That freedom is the accepted design premise. It does not justify saying that courses never work, that every person wants to experiment, or that AI assistance automatically produces retained skill.

Pass 2 — Build broad adoption and measure informed support

The recommended sequence is to make the offer understandable, expose the trade-offs, record public responses and then seek a documented policy commitment. All public communications should use the same substantive explanation. Accessible formats and translations improve access; they should not introduce different promises to different audiences.

This is a proposal for future work. No organisations have been contacted or appointed, and no supporters have been recruited. Petition participation can show voluntary engagement but is not a population estimate. Candidate responses can show a position but cannot establish votes in Parliament or implementation capacity.

C1–C5: a campaign that secures successive, reviewable commitments

These are proposed campaign and delivery steps toward the WF4 endpoint. Dates are planning windows; no parties, employers or funders have agreed to them. They allow overlapping sector progress while keeping delayed groups visible.

The first political ask is therefore a credible funded transition with dates and responsibilities, rather than a claim that everyone will work four days immediately after the election. The ultimate objective remains a genuine reduction and four-day opportunity by 2035. A nine-day fortnight can be an intermediate arrangement for a conventional five-day roster; existing RDOs or compressed shifts cannot be counted again. Workers may choose equivalent shorter-time patterns, which must be labelled separately from four-day attendance.

This sequence addresses two opposing risks. A simultaneous national switch can outrun staffing capacity; endless voluntary pilots can exclude difficult sectors permanently. The proposed response is phased enforceable arrangements, transparent funding gaps and corrective deadlines. Progress must not be bought by extending the hours of workers already on shorter standards or by transferring work to colleagues without compensation.

Negotiated arrangements can establish specific offers within the applicable system. A broader statutory or award-standard route may be needed to extend reach; its federal/state coverage, bargaining interaction, funding and enforcement require detailed policy drafting. A change to the NES alone must not be advertised as automatically amending every roster, pay entitlement and service budget. This is an implementation recommendation, not a legal determination about a particular workplace. SH01: ordinary hours and maximum-hours framework

Survey commissioning brief

The primary target population is eligible federal voters resident in Australia; the sampling contractor must specify the frame and its coverage. Commission a probability-based design if feasible, with telephone/offline accommodation so the survey does not select only enthusiastic AI users. A separate service-design sample can include eligible non-voters. Do not mix their estimates.

Each of three proposed waves would seek 2,000 completed voter interviews. Record unprompted awareness first; then present the dated, balanced proposal including BAU, sector staffing constraints and both the central and full-redemption AI costs. Ask overall support/opposition/uncertainty, whether respondents understood the offer, perceived priority relative to other public spending, and reasons. Measure support for the funded initial step and the final destination separately, using the same disclosed substantive offer for everyone. Test comprehension of the 35/37.5/38-hour examples and the fixed-32 alternative; do not count support for a smaller first step as endorsement of the final package. Retain “don’t know”. Record actual field dates, wording, recruitment, weighting and nonresponse. Randomise option order where appropriate and retain a stable core for comparisons. This design specification is the report’s recommendation; the questionnaire still requires cognitive testing.

At 2,000 independent probability-sample responses, a proportion near 50% has an approximate 95% sampling half-width of 2.19 percentage points; a design effect of 1.5 makes it 2.68. These exclude coverage, response and measurement bias. They are not valid margins for an open website poll. The disclosure approach follows A3-03: AAPOR standards. The accompanying script reproduces the arithmetic; no survey has been conducted.

Define “critical mass” operationally as a decision assessment using sustained informed support, a credible implementation coalition and concrete political commitments. No numeric threshold is known to make adoption likely. A majority estimate, even if its interval excludes 50%, is evidence about expressed support under that question, not a probability of legislation. If support collapses when costs are disclosed, revise the proposal or pace; do not omit the cost question. Publish adverse findings alongside favourable ones when publication is separately authorised.

Pass 3 — Resource the campaign and respect the election timetable

The Parliamentary Library’s ordinary simultaneous-election window is 7 August 2027 to 20 May 2028. A House election can occur earlier, so these dates are planning bounds rather than a fixed election date. Direct page retrieval returned 403; the official indexed text was checked on 9 September 2026. A3-04: Parliamentary Library timetable

The design should have a reviewable public package by early 2027 and readiness for an earlier election. It must not promise that a twelve-month Australian demonstration will finish before an election. If called early, present existing evidence, the proposed funded transition and the demonstration protocol, clearly labelled.

The AEC’s current reform page says major funding/disclosure changes commence 1 January 2027; older material referring to July 2026 is superseded for this plan. The planned campaign crosses that boundary. Before expenditure or fundraising, the responsible legal entity needs a dated assessment of applicable classification, accounts, donation, expenditure and disclosure obligations. A3-05: AEC reform timetable

Separately, authorisation requirements can apply outside election periods and depend on the communication and responsible entity. Calling a campaign “nonpartisan” does not settle its legal treatment. The proposed aim of influencing election-related support makes a review of the actual material necessary before release. A3-06: AEC authorisation guidance

Illustrative campaign envelope

These are gross resource assumptions in September 2026 dollars, not supplier quotes, lawful expenditure caps or a reach forecast. This twelve-month envelope covers an initial campaign cycle only. It is not the cost of C1–C5 through 2035; later campaign rounds, sector negotiation, policy development and implementation need separately priced resources and authority. Funding sources and the OWC legal entity remain unresolved. Cost per informed supporter cannot be estimated before valid reach and response data exist. Separate advocacy funding and data from personal allowances and independent evaluation. Receiving public AI support must never depend on supporting this policy.

Pass 4 — Compare four pathways on the same terms

All rows concern the same citizens/permanent-resident population to 2035, with voter support separately measured. B0 denotes continuing-policy expenditure; T2/T3/T4 denote incremental national time-reform costs still to be priced. Amounts below reuse B2’s illustrative annual AI assumptions, not new forecasts.

The campaign envelope is additional and distinct from national service budgets. WF3/WF4 AI expenditure can rise to A$26.55bn at full participation/redemption within the B2 assumptions. A larger allowance alone does not define WF4. Apply the same faster AI disruption, fiscal restraint, provider exit and workforce shortage to every pathway. Under fiscal pressure, expose a slower funded rollout or an explicitly revised allowance; do not conceal unpaid work or quietly ration an established entitlement. Under workforce shortages, protect services and use narrower time arrangements while retaining individual AI opportunity.

Pass 5 — Backcast from 2035 and hand off to action

Objections and current judgement

A campaign may win attention while the policy remains unaffordable; national time costs therefore stay on the critical path. A four-day week may benefit established employees first; A2’s part-time, casual and service-coverage arrangements must be part of the public offer. Released time may be used for rest; this is compatible with voluntary learning by doing. Greater AI use may intensify harms; B3 measures them rather than treating confidence as protection.

Current judgement: DESIGN PLAUSIBLE; DELIVERY UNRESOLVED. A credible broad public proposal can be prepared now. No representative informed-support data, campaign funding, appointed operator or legislative commitment has been established. The integrated action design remains subject to human acceptance or amendment. B3 supplies the demonstration sequence, evidence limits and conditions for any future launch.

References and source locations

A3-01–A3-06 are linked beside the supported claims. Original trial details and appraisal are retained in A1/A2; fiscal, infrastructure and eligibility assumptions are retained in B1/B2. The public source register records the primary sources and design limits; the programme data directory contains the reproducible budget and precision inputs and outputs. This is a bounded implementation synthesis, not a new systematic review of every trial or an electoral legal opinion.

Glossary

BAU: continuing approaches, including their evolution. Informed support: a response after exposure to a disclosed substantive proposal. Critical mass: a policy-adoption judgement requiring multiple forms of evidence, not a validated percentage. Entitlement: the proposed base personal offer. T: unpriced incremental working-time costs. OWC: One World Community, a proposed community whose legal and operating capacity is not assumed.

Authorship and method

Rick Molony owns the proposition and human decisions. The research, synthesis and reproducible arithmetic were developed with AI assistance. No interviews, polling, campaign or field experiment were conducted. AI generation is disclosed; it is not independent validation. Sources and policy dates were checked on 9 September 2026, with retrieval limits recorded.

Suggested citation

How this page was derived: the full substantive source body was rendered after the declared public-context transformations, removal of private delivery/workflow material and omission of the private change log.