Where the next trial could move the field.

Which administrative-process simplifications are net-positive, and which strip out steps that were quietly providing legitimate safeguards or eligibility checks?

What is the long-run effect of guaranteed-income programs once transfers end — do gains persist, decay, or accelerate after participants exit?

Can short, high-frequency interventions (text messages, app notifications) substitute for or complement intensive case management — especially in benefits enrollment and reentry?

When do conditional cash transfers in LMICs outperform unconditional ones, and is the difference worth the administrative cost of monitoring conditions?

Do early-childhood program effects scale beyond model programs (Perry, Abecedarian) when delivered at policy-relevant scale by typical staff?

Which school-finance reforms — equalization, weighted student funding, supplemental capital aid — actually translate into measured student outcomes, given the long causal chain from dollars to learning?

Do energy-conservation feedback programs sustain their 2% effect at 5 and 10 years, or does the novelty wear off?

What is the right cost-benefit threshold for adopting financial-literacy programs given the consistently small (or null) effects in rigorous trials? Are there subgroups where effects are meaningfully larger?

Does targeted housing-first work as well for first-time homeless episodes as it does for chronic homelessness, or is rapid rehousing the better match for shorter durations?

Which behavioral interventions transfer cleanly from high-income to low- and middle-income contexts, and which depend on infrastructure (smartphones, banking, formal addresses) that limits transferability?

Do public-library experiments — programming changes, fine elimination, outreach — generalize across library systems, or do they depend heavily on staff capacity and community context?

What is the marginal effect of additional reminders past the third? Most reminder trials measure 1–3 messages; the diminishing-returns curve is not well characterized.

Do focused-deterrence violence-reduction programs (like Boston's Operation Ceasefire) sustain their effects after the initial intensive phase ends, or do shootings rebound?

When does personalization (using the recipient's name, condition, or balance) materially improve over a generic reminder?

How long does the effect persist — do recipients habituate after the third message?

Which populations respond more weakly, and is the gap explained by phone access, language, or trust?

Does the social-norm effect in tax compliance work for high-income taxpayers and corporations, or primarily for small businesses and individuals?

Can pre-populated tax returns ('return-free filing') increase compliance and reduce burden in the US? The evidence from OECD countries that have adopted it is promising but not experimental.

Which transportation-pricing interventions (congestion charges, transit subsidies, EV incentives) produce durable mode shifts versus short-term substitutions that reverse once novelty fades?

How durable are voter-mobilization effects from canvassing and SMS interventions across multiple election cycles? Most trials measure a single election.

Do the wellbeing effects sustain after transfers end, or do they fade?

How do effects differ in tight labor markets versus slack ones?

What is the right comparison — to status-quo benefits, to in-kind aid, or to nothing?