Randomised Trials Show Gains Across Government

The Federal Government says randomised trials across the public sector are delivering useful results, with Assistant Minister Andrew Leigh set to outline them at the ACE and BETA showcase.

Updated on Nov. 4, 2026 in Electric Vehicles

Bold flat-color economist-style editorial illustration of two identical glass vials on a slate surface, one filled to a different level, suggesting a controlled comparison.
The Federal Government says randomised trials across the public sector are producing useful results, including a pre-call text that lifted Services Australia contact rates from 52 to 64 per cent. AI Illustration. Upload story photo >

Live Poll

Do you trust the federal government to test whether its programs actually work?

Randomised trials across the Australian public sector are producing useful evidence, according to the Federal Government. Results include a pre-call text lifting successful Services Australia contact rates from 52 to 64 per cent.

Why it matters

The trials show how small changes to government communication and policy design can be measured and tested before wider rollout. Evaluating outcomes this way helps distinguish genuine policy effects from activity that would have happened anyway.

A randomised trial of 16 developers completing 246 real tasks found AI assistance increased completion time by 19 per cent, while developers still believed it saved them 20 per cent of their time. The electric car discount's 64,000 additional sales were estimated by comparing vehicles priced below and above the eligibility threshold.

The players

Andrew Leigh

Andrew Leigh is an Australian Assistant Minister and former economics academic who has championed evidence-based policymaking.

Services Australia

Services Australia is the Australian Government agency that delivers welfare, health and child support payments to citizens.

The details

The electric car discount review combined stakeholder consultation with a difference-in-differences analysis comparing sales for vehicles just below and above the eligibility threshold. The Impact Evaluation Practitioners Network has grown to more than 700 members over two years, supporting wider use of these methods across government.

Timeline

  1. Last year: a randomised trial of AI developers was published.

  2. Over the first three years: the electric car discount generated about 64,000 additional sales.

  3. Over the past two years: the Impact Evaluation Practitioners Network grew past 700 members.

  4. 4 November 2026: the article on randomised trials and policy evaluation was published.

Roadmap

The electric car discount's measured effect of about 64,000 additional battery electric vehicle sales gives Australia a documented benchmark for EV incentive policy. The findings place the discount within a broader push to test automotive subsidies against what would have happened without them.

Buyers considering a battery electric vehicle may weigh how government incentives have affected pricing and availability. The trial findings suggest the electric car discount meaningfully changed purchasing decisions for vehicles near the eligibility threshold.

The takeaway

Testing government programs with randomised trials can show which interventions actually change behaviour and which do not. Readers can expect more policy decisions in Australia to rest on this kind of measured evidence.

Further reading

Explore more on Electric Vehicles to see how incentive policies are shaping the Australian market.

Source note: This article includes information reported by Region Canberra.

Live Poll

Do you trust the federal government to test whether its programs actually work?