Question of the Day: When is using AI, and AI tools, unethical and/or verge into plagiarism? What and when does one report out regarding the use of AI tools? How does that disclaimer influence the reader?
My last post focused on my, preliminary, experimentation with agentic AI. Today I am reporting out on my next steps regarding a specific application. As you may know in this phase in my life, I am dedicating myself to poverty reduction at home (in the US) and the most promising space, in my mind, seems to be some variant of Basic Income.
Two things to note: (1) I am volunteering at the Basic Income Earth Network, and (2) trying to learn as much as I can about the issues and challenges of this space and particularly why there is so little scaling up. This is post is very much focused on the later.
Part of this challenge is trying to understand what is Basic Income and what are the arguments for – and critically – the arguments against. One of these questions is:
Does Basic Income impact labor force participation?
The proponents I have met seem firmly in the camp that this issue is long solved and the answer is there is no impact, certainly not a negative impact, and possibly a positive impact. And yet, the literature, even very recent literature suggests that the debate is live and active. Note that my position is as a proponent of some type of basic income to help address systemic and pervasive poverty in the United States.
This seems like the type of research question well suited to a LLM wiki. The process I followed is roughly:
- Using scholar.google.com identify peer reviewed articles that discuss this topic. Enter the best of those into the raw data folder.
- Query the LLM to answer the question. Ask follow up questions to improve one’s own understanding. Ask it to identify areas that need strengthening.
- If more research is needed or desired, then return to Step 1 to add more articles.
- If not, then ask the LLM to prepare a summary note discussing the strengths and weaknesses of various positions. I asked the LLM to reference the wiki and the original articles and to create a bibliography.
- Revisit periodically as needed.
My off-the-cuff estimation is that this approach allowed me to ingest 100s of pages of reasonably dense scholarly papers and understand the crux of the various arguments and provides me with additional focus on where to focus my research – total time reduced from days, or more likely weeks to hours. A 10x gain. That said, this analysis is not The Truth but a good initial step to accelerating my own education on the topic.
Question: Is this approach ethical – for personal learning, for creating a blog post, others? What needs to be disclosed?
How are you, the reader, influenced by the fact that this research used AI?
The following section is the analysis of the impacts of Basic Income on labor force participation prepared using this workflow. The summary of the AI summary is that it is complicated.
Does Basic Income Impact Labor Force Participation?
A Research Blog | May 5, 2026
The question of whether a basic income would cause people to stop working is among the most politically charged issues in contemporary policy debate. For critics, the answer feels self-evident: give people unconditional money, and they will choose leisure over labor. For advocates, the concern reflects an ungenerous view of human motivation. What does the evidence actually say? The answer is more nuanced — and more genuinely uncertain — than either camp typically acknowledges.
Defining the Terms
Definitions matter enormously here, and fuzzy language is a primary source of confusion in this debate.
Labor force participation rate (LFPR) is the share of the civilian noninstitutional population aged 16 and older who are either employed or actively seeking employment (U.S. Bureau of Labor Statistics, n.d.). A decline in LFPR means fewer people are working or looking for work — distinct from a rise in unemployment, which measures those actively searching but failing to find work.
Basic income is not a single policy. The term covers a spectrum of designs that differ in universality, conditionality, benefit level, and funding source. A universal basic income (UBI), as debated in the academic literature, is typically defined as a periodic, unconditional cash payment made to all individuals regardless of income or employment status (Widerquist, 2024). Programs actually studied in the empirical literature range from modest targeted transfers ($300–$500/month) to more substantial unconditional payments ($1,000/month). These differences in amount, conditionality, targeting, and funding mechanism substantially affect outcomes and complicate cross-study comparisons.
What Theory Predicts
Standard labor economics predicts that unconditional income transfers will reduce labor supply through the income effect: when people have more non-labor income, they can afford more leisure. This prediction is strongest for individuals with modest wages and flexible work arrangements. Microsimulation models — including Urban Institute projections for the 2021 U.S. Child Tax Credit expansion — have translated this logic into projected employment declines (Ananat et al., 2022). The theoretical case is coherent. The empirical record is less tidy.
What the Evidence Shows
A systematic review applying PRISMA methodology to over 1,200 documents (de Paz-Báñez et al., 2020) found no significant overall reduction in labor supply from UBI-type programs across 38 contrasting empirical studies. Where reductions were observed — children exiting child labor, elderly workers retiring earlier, mothers shifting time to childcare — they were characterized as “functional” reductions that serve social goods, and they were generally offset by labor supply increases elsewhere in the community.
More recent and rigorous evidence complicates this reassuring picture. Vivalt et al. (2024), in the largest U.S. guaranteed income randomized controlled trial (RCT) to date — $1,000/month for three years, n=3,000 — found labor force participation fell by 4.1 percentage points in the treatment group relative to controls, with earned income down approximately $1,800 per year. The time freed from paid work was primarily absorbed by leisure rather than caregiving, education, or other productive activities.
A companion study from the Compton Pledge (Balakrishnan et al., 2024) found part-time workers reduced labor force participation by 13 percentage points, though single mothers — a group often poorly served by conditional programs — increased work hours by 6.4 hours per week. Subgroup heterogeneity, in other words, is large and cuts in both directions.
By contrast, the expanded U.S. Child Tax Credit of 2021 ($250–$300/month per child) produced no detectable employment effect across a sample of over 500,000 households (Ananat et al., 2022), and the largest UBI RCT conducted in Kenya (Banerjee et al., 2023), spanning approximately 23,000 adults over two years, found no reduction in total hours worked — with a notable shift from wage employment to self-employment and enterprise formation.
Why the Studies Disagree
Several variables appear to drive the divergence in findings.
Transfer size matters. The Vivalt et al. study’s $1,000/month represents a larger effective income floor than most programs reviewed in the 2020 systematic review. A labor supply effect that is negligible at modest transfer levels may become detectable as the transfer approaches or exceeds a meaningful fraction of a recipient’s prior earnings.
Duration matters. Short-term pilots may not capture permanent income effects. Participants who know a program ends in 12 months may not substantially revise their career plans. The Kenya study is a partial exception, but its 12-year results are not yet available (Banerjee et al., 2023).
Context matters. Developing-country experiments, where recipients face credit constraints and limited self-employment opportunities, may generate productivity gains through business formation that do not generalize to high-income labor markets with well-developed credit systems.
The funding mechanism is often ignored entirely. Almost no pilot study imposes the tax increases that a full-scale UBI would require. Research on 28 OECD countries finds that higher tax burdens significantly reduce LFPR, independently of any transfer (Yıldırım & Kuştepeli, 2023). A full-scale UBI would require effective tax rates of 50–70% or more on income to be fiscally sustainable (Sawyer, 2024) — a distortion that no pilot has ever faced, and that macro models suggest could reduce GDP by 10–18% under most funding designs (Jaimovich et al., 2022).
Where the Genuine Uncertainty Lies
The most defensible summary of current evidence is this: moderate labor supply reductions are plausible at high transfer levels in high-income countries, and may not be offset by productive time use — but effects are highly heterogeneous by subgroup, and the tax-financing channel may matter as much as the transfer itself.
It is worth noting that public intuition systematically overstates labor withdrawal. A 2018 survey by Duflo and Banerjee found 49% of Americans believed “many people” would stop working under a $13,000/year basic income — but only 13% said they themselves would stop. Whether this asymmetry reflects accurate knowledge about others or motivated self-presentation is itself an open empirical question.
Areas for Additional Research
The following gaps are particularly important to close before strong policy conclusions can be drawn:
- Long-run experiments. Most pilots last 2–3 years at most. Career trajectories, human capital investment, and business formation unfold over decades. The Kenya study’s 12-year endpoint will be uniquely valuable, but analogous long-run designs in high-income countries remain rare.
- General equilibrium effects. Pilots measure individual-level responses in partial equilibrium — they cannot capture economy-wide wage, price, and firm behavior. Macro models that attempt to account for these (Jaimovich et al., 2022; Sawyer, 2024) require empirical calibration that only full-scale implementation can provide.
- Funding mechanism effects. Almost no empirical study isolates the tax-side effect of UBI from the transfer-side. Research designs that vary revenue instruments — wealth taxes, VAT, income surcharges — while holding transfer amounts constant would substantially advance the field.
- Subgroup heterogeneity. The data consistently show that aggregate effects mask large variation: single mothers, part-time workers, the elderly, and young people respond quite differently. Policy design should be grounded in granular understanding of who changes their labor supply and why.
- Interaction with existing welfare programs. In high-income countries, a UBI would partially replace or interact with existing conditional programs. Whether that interaction removes welfare traps or creates new disincentives is underresearched.
Conclusion
The empirical record does not support the claim that basic income will cause widespread labor market withdrawal. It also does not warrant confidence that effects will be negligible. The most responsible position is to treat transfer amount, funding mechanism, population, and program duration as critical design variables — and to be skeptical of both alarmist and dismissive readings of the current evidence base.
References
Ananat, E., Glasner, B., Hamilton, C., & Parolin, Z. (2022). Effects of the expanded Child Tax Credit on employment outcomes: Evidence from real-world data from April to December 2021 (NBER Working Paper No. 29823). National Bureau of Economic Research.
Balakrishnan, R., Bhattacharya, J., Hamilton, C., Hendra, R., Howard, G., Hyra, D., Karlan, D., Lucchesi, A., Mendes, M., Rozo, S., & Slocum, M. (2024). The Compton Pledge: 18-month endline findings (NBER Working Paper No. 33209). National Bureau of Economic Research.
Banerjee, A., Faye, M., Krueger, A., Niehaus, P., & Suri, T. (2023). Universal basic income: Short-term results from a long-term experiment in Kenya. Working paper, September 15, 2023.
de Paz-Báñez, M. A., Asensio-Coto, M. J., Sánchez-López, C., & Aceytuno, M.-T. (2020). Is there hard evidence of the impact of universal basic income on labour supply? A systematic review. Sustainability, 12(22), 9459. https://doi.org/10.3390/su12229459
Jaimovich, N., Saporta-Eksten, I., Setty, O., & Yedid-Levi, Y. (2022). Universal basic income: Inspecting the mechanisms (CEPR Discussion Paper No. 16996). Centre for Economic Policy Research.
Sawyer, M. (2024). Macroeconomic observations on paying for and funding universal basic income. Basic Income Studies, 19(2), 227–252. https://doi.org/10.1515/bis-2023-0032
U.S. Bureau of Labor Statistics. (n.d.). Labor force participation rate [Data series CIVPART / LNS11300000]. Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/series/CIVPART/
Vivalt, E., Rhodes, E., Haggag, K., Bridgman, A., Brodeur, A., Feinberg, F., Gentile, A., Ghai, S., Knight, B., McDonald, R., Oreopoulos, P., Price, J., Sampson, C., Shaefer, H. L., Shannon, M., Woda, J., & Evans, D. (2024). The employment effects of a guaranteed income: Experimental evidence from two U.S. states (NBER Working Paper No. 32719). National Bureau of Economic Research.
Widerquist, K. (2024). The debate over the definition of basic income. Basic Income Studies, 19(2), 155–181. https://doi.org/10.1515/bis-2024-0021
Yıldırım, D. Ç., & Kuştepeli, Y. R. (2023). The effect of tax burden on labor force participation rate: A panel data analysis for OECD countries. Trends in Business and Economics, 37(2), 91–97.
This blog was prepared by the author with the assistance of an LLM wiki. The underlying data is human curated and the LLM creates a knowledge wiki which is queried by the author. An initial blog draft responding the author’s question is prepared by the LLM and refined under the guidance/hand of the author.