Research team: Luca Correani, Patrizio Morganti, and Roberta Sestini
Transport poverty (TP) broadly refers to various forms of inequalities related to transport availability, affordability and accessibility. Our contribution focuses on the affordability dimension, considering the financial burden of mobility on household budgets. Following the existing literature, we employed two indicators of TP: (i) the 10% metric, defined as the share of households spending more than 10% of their income on transport-related expenditures, and (ii) the share of transport expenditure on total household expenditure.
Although existing research shows that a high incidence of TP negatively affects well-being, empirical evidence on its determinants remains limited, especially for Italy.
Phase 1
In Phase 1 of the project, we sought to fill the existing gap in the literature by examining the socio-economic drivers of transport poverty at the regional level, with particular attention to sectoral composition and labour market heterogeneity.
The empirical analysis is based on a strongly balanced panel dataset covering the 20 Italian regions over the period 2014-2023 (annual observations). Data are drawn from ISTAT sources, including the Household Budget Survey, mobility and transportation statistics, and national accounts.
The empirical framework explores several aspects potentially related to TP, included as explanatory and control variables: (a) labor market conditions, including unemployment and sectoral employment (primary, secondary, tertiary); (b) regional economic structure, measured by real GDP per capita, GDP growth, and sectoral value-added shares; (c) socio-economic vulnerability, proxied by the share of households in absolute poverty; and (d) transport system characteristics, including reliance on public transport, rail use index, and rail satisfaction index. These variables allow us to assess both structural and cyclical determinants of transport poverty and to test for potential sectoral differences and non-linear relationships between economic growth and transport poverty.
Three econometric specifications are estimated. First, pooled OLS models examine the relationship between transport poverty and labor market variables, controlling for poverty rates and transport system indicators. Second, pooled OLS models assess the relationship between transport poverty and the level of real GDP per capita, including sectoral value-added composition. Third, we estimate a non-linear specification with regional fixed effects to examine the relationship between transport poverty and GDP growth, including a quadratic term to test for an inverted U-shaped relationship. All variables are expressed in natural logarithms, except GDP growth. A dummy variable captures the impact of the Covid-19 years (2020-2021).
Estimation results show a negative relationship between transport poverty and unemployment, suggesting that unemployed individuals spend less on transport. Employment in the primary and secondary sectors is positively associated with transport poverty, plausibly due to greater commuting needs, while employment in the tertiary sector is negatively associated, in line with its higher potential for remote work. Real GDP per capita is positively correlated with transport poverty, indicating that higher income levels stimulate transport expenditure. A higher share of value added in the primary sector increases transport poverty, while the role of the tertiary sector appears mitigating.
A key finding is the presence of an inverted-U shaped relationship between GDP growth and transport poverty. At low growth rates, transport expenditure remains limited; at intermediate growth levels, expenditure rises without being fully offset by income gains, increasing vulnerability; at high growth rates, income growth more than compensates for additional transport costs, reducing transport poverty. Transport poverty is negatively associated with absolute poverty rates, as poorer households tend to reduce mobility due to budget constraints. Higher reliance on and satisfaction with rail and public transport are associated with lower transport poverty. Finally, the Covid-19 period is associated with a significant reduction in transport poverty, due to reduced mobility during lockdowns.
The above findings were presented at the Workshop “Transport poverty in focus: insights from the HAC-VIA project” held in Viterbo on 23 May 2025.
Phase 2
During phase 2 of the project, we investigated the issue of transport poverty as intertwined with the green transition. The issue is increasingly attracting attention at the EU level. It is primarily addressed through an integrated policy approach, rather than through sector-specific legislation. Key EU strategies – such as the European Green Deal, the Sustainable and Smart Mobility Strategy, and the Fit for 55 package – acknowledge that the transition towards low-carbon mobility has to be socially inclusive and avoid disproportionate burden on vulnerable households.
Focusing once again on the affordability dimension, which is the most suitable for capturing the burden of the green transition on vulnerable groups, we enriched our panel dataset. More specifically we included within the set of socio-economic (control) variables the real GDP per capita, the share of households in absolute poverty, the rate of growth of the consumer price index for gasoline, and population density. Other covariates were selected according to our research questions, with a focus on transport-related choices and features, public transport demand and supply measures, vehicle stock characteristics – including the polluting potential of the regional vehicle fleet and the share of vehicles powered by alternative fuel types. Data were collected from several sources (e.g. ISTAT, ACI, ISPRA).
We estimated several econometric model specifications using panel fixed effect models, as this allows to control for time invariant unobserved region-specific characteristics (that may affect both income and the TP metric).
First, our results indicate, across all model specifications, a negative and statistically significant relationship between the share of households being in absolute poverty and the incidence of transport poverty. This finding may be driven by the phenomenon of “hidden transport poverty”, which refers to situations where individuals limit their mobility below their actual needs due to cost pressure or affordability constraints. Moreover, an increase in the population density variable worsens TP, meaning probably that in our country TP is not merely a rural phenomenon. Not surprisingly, and in line with a stream of literature showing that increasing fuel prices hamper well-being, an increase in gasoline prices is associated with a higher incidence of transport poverty. This is most likely to occur in contexts characterised by limited public transport provision, where low-income groups may have no option but to absorb higher petrol costs, being in a condition of forced car ownership. As expected, all estimates point to a lower incidence of transport poverty during the COVID-19 pandemic years.
Regarding the role of transport choices and attitudes, the results indicate (in all specifications) that rising passenger car motorization rates are associated to increased transport poverty. In fact, data show that in Italy transport demand is increasingly met by individual road transport, which for its growth and largely stable modal share remained dominant with respect to other transport modes. Forced car ownership, combined with a strong car ownership culture, contributes to explaining these findings.
Besides, an increase in the polluting potential index of passenger cars exacerbates transport poverty. This may be driven by more vulnerable population groups, which disproportionately rely on older, less efficient and more polluting vehicles. In this context, targeted policy incentives aimed at facilitating a transition towards hybrid vehicles (battery electric – HEV- and plug-in hybrid electric – PHEV- vehicles) may help alleviate the effect. Rail use satisfaction index – a measure of public transport quality– in our analysis contributes positively to reduce TP, whilst other variables related to public transport demand and supply are not statistically significant.
Hence, we argue that the quality, frequency, time competitiveness relative to private cars, and accessibility of public transport are key factors influencing individuals’ choice to commute by public transport rather than by car. Intervening on these factors may contribute to a modal shift in favour of public transport, with positive effects also on transport poverty.
Finally, a non-linear relationship emerges between real per capita income and transport poverty. We developed an interpretative framework for this finding and explored its policy implications. There is a growing body of research establishing that, over the past several decades, and due to rising housing costs, many cities have experienced a spatial redistribution of low-income population from central to suburban neighbourhoods. A potential negative impact of these trends is the increasing concentration of low-income households in more car-oriented areas, which results in higher barriers to daily travel and greater costs associated with mobility and activity participation. We examined whether this pattern also applies to Italian regions. Marked spatial differences in house price index dynamics are observed, with stronger growth concentrated in the more affluent macro-regions—the North-West and especially the North-East. Notably, these regions are also characterised by comparatively higher motorization rates. At the same time, we find no strong evidence supporting the hypothesis of a modal shift towards public transport, at least for the North-East.
These results suggest that progress towards key sustainable mobility objectives, such as decoupling private transport growth from economic growth and achieving modal rebalancing remains uneven for Italian regions. This, in turn, negatively affects transport poverty.
We are allowed to conclude that, on the one hand, in regions with low per capita income, expanding public transport supply, upgrading infrastructure, and subsidising its use may help break the transport poverty trap and foster sustainable mobility. On the other hand, in regions with higher per capita income one of the critical drivers of increasing transport poverty might be represented by inner-city gentrification coupled with the so-called suburbanization of poverty. Reduced employment accessibility is likely to result in longer commuting distances, which in turn can exacerbate transport poverty and increase transport-related emissions. Thus, in this scenario we call for targeted measures to increase accessibility and time competitiveness of public transport as compared with private car use, which can align social and environmental goals.
The above findings were presented at the IAERE Annual Conference (Trento, 12-13 February 2026).
The research team is currently finalizing a research paper based on these results.