Unraveling the Dynamics of Affordability Transport Poverty

Abstract

Transport poverty is a growing concern in the Netherlands, with affordability emerging as one of its most pressing and persistent dimensions. Rising transport costs and evolving urban infrastructure increasingly limit certain population groups' access to essential services such as employment, education, and healthcare. While shared mobility offers a potentially flexible and cost-efficient solution, policymakers still lack insight into how affordability transport poverty evolves over time and how such policies influence its dynamics.

This thesis quantitatively investigates if shared mobility can alleviate affordability transport poverty through a System Dynamics (SD) modeling approach, using selected demographic groups of the municipality of Almere as a case study. The model captures key feedback mechanisms linking employment income, transport costs, and travel behavior across distinct population segments, referred to as transport groups. It simulates how these groups adjust travel patterns based on available budgets, transport needs, and employment conditions. Core model components include travel budget thresholds, transport expenditures, and employment income. The main research question of this study is:

“What are the key dynamics of affordability transport poverty, and to what extent do shared mobility policies influence these dynamics across different demographic groups?”

To address this, two shared mobility policies were simulated:

- (Policy 1) Last-mile regional hubs: This policy provides regional shared mobility to extend the range of accessible travel.

- (Policy 2) Subsidized micro-hubs: This policy locally provides two days of free shared mobility options per week.

The analysis starts with a detailed base case simulation, focused on individuals in Almere with two defining characteristics: (1) they have either an income below social minimum or a sub-modal income, and (2) they either own a car or do not. These factors define four distinct transport groups used in the simulations.

Base case results show variation in vulnerability across these groups. Individuals with car access and with incomes below social minimum levels consistently exhibit the highest risk of falling into affordability transport poverty. In these groups, high transport costs eventually led to limited access to jobs, constrained travel budgets, and reduced salary, forming self-reinforcing feedback loops. This creates a poverty trap that further restricts social and economic mobility. By contrast, those who did not own a car and with a sub-modal income experienced less severe or no affordability issues, even under worsening economic conditions.

Both policies tested introduce improvements, with policy 1 being the most effective in increasing both income and reducing transport poverty levels. However, while somewhat effective, their impact is relatively modest in terms of increasing income or reducing severe transport poverty, and neither policy is sufficient to reverse the broader trends that contribute to affordability transport poverty. Yet still, the importance of moderate improvements in income must not be understated, especially for lower-income groups, who can have significant quality of life improvements with only modest amounts of increased income.

These results suggest that while shared mobility policies offer a degree of support, they are unlikely to resolve transport poverty when implemented in isolation. Affordability constraints are closely tied to broader socio-economic conditions, including housing, employment, and urban form. This highlights the importance of integrated approaches. Effectively addressing affordability transport poverty requires these measures to be complemented by strategies like targeting wages, job distribution, housing policy, and long-term transport costs.

Moreover, the results also underscore the importance of timing: early-stage interventions yield greater impact, especially before feedback loops fully entrench poverty. Delayed implementation reduces effectiveness, even for the same policy design.

The presented results do have some limitations. One of them is that this research adopted a general approach to shared mobility, without considering the associated implementation costs. It is important to acknowledge that alternative policies may be more financially viable. In that case, shared mobility could instead be applied in more targeted and specific contexts where other measures prove ineffective—one example being individuals who are highly car dependent but are not able to afford one.

Moreover, although the model simplifies real-world complexities, it offers a robust framework for understanding the core dynamics of affordability transport poverty. It contributes a novel application of SD modeling to this domain and provides insights for local policy. By uncovering the underlying feedback mechanisms and evaluating targeted interventions, this study lays the groundwork for future SD-based research and supports more equitable, data-driven policy design in cities like Almere and beyond.