Public support for cultural consumption raises a familiar question in public economics. Participation in cultural activities varies markedly with income and family background, yet the benefits of cultural consumption may extend beyond the individual consumer. If exposure to books, cinema, live performances, and museums contributes to the accumulation of cultural capital, broadens opportunities for participation, and fosters persistent consumption habits, reducing barriers to access can provide a rationale for public intervention. Whether vouchers achieve this objective, however, is ultimately an empirical question. For individuals who would have purchased the same goods anyway, public spending may simply crowd out private expenditure. The relevant policy question is therefore not only whether cultural demand should be subsidized, but also whether, for whom, and to what extent such subsidies generate additional cultural consumption.

The Italian 18app programme provides a particularly informative setting in which to address these questions. In The impact of a public voucher for the arts, published in the Journal of Economic Behavior and Organization, Andrea Baldin, Anna Marenzi, and Francesca Zantomio provide the first causal evaluation of the effects of a cultural voucher on young people’s consumption. The contribution extends beyond the Italian case: 18app inspired similar initiatives in France, Spain, and Germany, while the design of cultural voucher schemes continues to evolve.

The analysis draws on the Italian National Institute of Statistics (Istat) survey Aspects of Daily Life for 2013–2019 and exploits age-based variation in eligibility before and after the introduction of the voucher. Eligible individuals aged 18–20 are compared with 21–23-year-olds, who were never eligible. Nineteen- and twenty-year-olds are included because the benefit, awarded at age eighteen, could be used for more than one year, whereas the survey records age at the time of interview and cultural consumption over the previous twelve months.

The identification strategy combines a difference-in-differences design with entropy-balancing reweighting. Before estimation, the treatment and comparison groups are made comparable along a set of observable characteristics, including education and employment status, co-residence with parents, parental education and household economic conditions, and municipality size. Because the survey does not record actual voucher use, the estimates should be interpreted as intention-to-treat effects.

The results point to a clear increase in cultural participation along the extensive margin. Eligibility raises the probability of going to the cinema at least once a year by 5.6%, attending a non-classical music concert by 4.1%, reading at least one book by 4.2%, and purchasing books or e-books online by 7.1%. Along the intensive margin, the average effect is statistically significant only for cinema attendance, with an increase of 0.56 visits per year.

The most pronounced heterogeneity emerges by family background. Socio-economic status is proxied by the highest level of education attained within the household, an indicator that the literature identifies as a strong predictor of cultural participation. The effects are concentrated among young people living in households with no university graduate. For this group, the voucher increases both entry into the consumption of cinema, non-classical concerts, and books, and the frequency of participation: cinema attendance rises by 20%, non-classical concert attendance by 23%, classical concert attendance by 45%, and museum visits by 16%. By contrast, effects are close to zero for children of university graduates. This pattern is consistent with economic and access constraints playing a larger role among young people with lower levels of family cultural capital.

The paper also quantifies the extent to which voucher-funded spending is genuinely additional rather than replacing purchases that would have occurred in any case. Combining administrative data on voucher use with average prices and the estimated treatment effects, the authors find that each euro spent through the voucher generates €0.13 of additional cinema expenditure and €0.19 of additional spending on non-classical concerts. Among young people from households with no university graduate, additional expenditure reaches €0.28 per euro for classical concerts. Crowding out of private spending is therefore substantial, but incomplete, and is less pronounced where baseline cultural consumption is lower.

Two additional findings suggest that the effects of the programme extend beyond the recipient’s immediate purchase. Among parents of eligible eighteen-year-olds, online purchases of books and e-books increase by 21%, consistent with within-household spillovers. Moreover, young people who have exhausted the voucher continue to attend the cinema more frequently than never-eligible peers, with a 28% differential. This persistence is consistent with habit formation and the development of cultural tastes.

These findings are particularly relevant in light of subsequent reforms to the programme. Introduced in 2016, 18app granted €500 to all residents turning eighteen and was therefore universal with respect to income and educational performance. The 2023 Budget Law replaced it with two selective €500 cards that can be combined. The Youth Culture Card is available to young people in households with an ISEE indicator not exceeding €35,000, while the Merit Card is awarded to students who obtain an upper-secondary-school diploma with the maximum grade of 100 by the year in which they turn nineteen. This reform marked a clear shift from age-based universalism to eligibility criteria based on household economic conditions and educational performance. From 2027, both cards will be replaced by the Carta Valore, available to those who obtain their diploma by the year in which they turn nineteen. The new scheme removes both the ISEE threshold and the grade requirement, but it does not restore the universalism of 18app, since eligibility remains conditional on timely completion of upper-secondary education.

These results have direct implications for the design of future cultural vouchers. A universal voucher generated the largest responses among young people furthest from cultural consumption. Eligibility rules based on merit or regular educational progression may therefore exclude precisely those groups, since lower grades and less linear educational trajectories are more common in less educated families. An ISEE threshold can reduce transfers to individuals who would have consumed cultural goods anyway, but it may also generate non-take-up and exclude lower-middle-income households just above the cutoff. Finally, cultural consumption among young people is often shared: narrowing eligibility may weaken peer and family spillovers that, as the evidence suggests, contribute to the effectiveness of the policy.