Where and how much? - A modelling framework to estimate land value uplifts from transport interventions
Due to the COVID-19 crisis measures the PhD defence of Andres Morales will take place (partly) online in the presence of an invited audience.
The PhD defence can be followed by a live stream.
Andres Morales is a PhD student in the research group Urban and Regional Planning and Geo-Information Management (PGM). His supervisor is prof.dr.ir. J.A. Zevenbergen from the Faculty of Geo-Information Science and Earth Observation (ITC).
Sustainable mobility, that provided by Light Rail Transit (LRT) and Bus Rapid Transit (BRT) systems, is of growing interest in many Global South countries as a sustainable strategy to face the challenges of rapid urbanization. Sustainable mobility is fundamental to boosting economies and promoting equality. It is well known that accessibility improvements also have wider effects such as land values uplifts. Thus, practitioners are calling for options to include these effects in Cost Benefit Analyses (CBAs). Direct effects (i.e. travel-time savings) cannot be easily translated into wider often economic benefits. This makes it difficult to justify the high investment required. In contrast, a direct translation of the monetized accessibility improvements could provide useful information for design processes and for plan evaluation/selection processes. Moreover, land value uplift analyses could provide the grounds for the formulation of mechanisms (e.g. land value capture) to secure funding for sustainable transport investments. However, practitioners continue to face methodological difficulties in estimating land value uplifts reliably.
Studies that aim to predict value uplifts before the interventions are implemented are scarce. Some limitations are identified in such literature as there is. First, estimating the elasticities of land values as a function of accessibility would be the preferred approach rather than an urban economic formulation in areas where market imperfections are accentuated or even unknown. Second, a comprehensive concept and operationalization of urban accessibility as it accrues to users and the valorisation of land is missing. Accounting for public mobility benefits by Euclidian distance to transit stops makes it impossible to associate the value uplift with an interpretable metric of improved geographic access (i.e. travel times reduction). Introduction of new transport technologies commonly include modifications to existing road networks. Hence, impacts on the geometric accessibility (i.e. network centrality) can also be expected. This is important since recent studies have suggested that geometric accessibility, as analysed in Space Syntax (SSx), adds spatial information that improves the understanding of land and property values. Third, the spatial scope of analysis must not be restricted to pre-assumed catchment areas. Instead, a city scale approach would allow the calibration of elasticities based on richer datasets while gaining broader spatial insights into the potential land value effects of a transport investment.
The objective of this research was to propose and implement a modelling framework to estimate the spatially distributed land value uplifts of future transport infrastructure by means of operationalizing a comprehensive accessibility definition into a predictive model. To do so, we implemented a quantitative correlational research using Guatemala City as the case study area. We first compared location-based and Space Syntax methods to map urban accessibility in data scarce contexts. Multiple data layers were obtained from official and Volunteered Geographic Information (VGI) sources and then consolidated. Second, we incorporated a comprehensive definition of accessibility and uncovered its relations with residential land values by means of a multivariate regression model. Third, we used a geostatistical method to develop a parsimonious land value model. The predictive model was then used to construct a land value map for Guatemala City. Finally, we structured data and methods from previous steps into a modelling framework. The framework’s applicability is demonstrated by estimating the future impacts of an LRT system on urban access and residential land values.
The findings revealed that SSx metrics consistently reflect urban accessibility conditions that have had previously only been addressed using location-based methods. Furthermore, an increased ability to explain residential land-values and prediction accuracy was achieved when including geometric-access metrics and addressing the disparity of geographic-access opportunities due to available transport modes. From the results, it appears that geometric- accessibility capitalizes land not only at its location, but also as a resource that is reachable by means of geographic-access.
Implementing a geostatistical method was demonstrated to be important in developing a parsimonious model. Some SSx metrics were confirmed to contribute significant statistical information, even under spatialized modelling conditions. Moreover, it was observed that geographic access to urban centrality contributed more statistical information compared to access to the central business district. A land value map, in combination with inferential modelling and accessibility analyses, allowed us to conclude that Guatemala City is predominantly monocentric. The road network hierarchy identified via SSx was observed to play an important role in the distribution of urban access and land values.
Operationalizing the modelling framework revealed where and how much the accessibility improvements and land value uplifts can be expected after introducing an LRT system. The effects were observed to be higher in areas near to the corridor but gradually lower at locations farther away. However, the city-wide modelling scope revealed that such effects spill beyond a distance of 3.4 km and even beyond 4 km from the access stations. The spatial distribution of such effects is rather heterogeneous along the corridor and across the city. The way in which such effects are distributed is associated with the availability, or its lack, of other public transport infrastructures. Higher access benefits and value uplifts were identified in areas where either MetroRiel’s use leads to a real reduction in travel time, or when its interaction with other transport infrastructure (in a perpendicular layout) represents increase in the coverage of the overall transport network. Furthermore, new road connections in the form of bridges or minor transversal connections along the new transport corridor have a positive influence on the city-wide access. This improvement, as measured by means of geometric via geographic access, has been shown to greatly benefit certain peripheral areas of the city. Consequently, this also leads to significant land value uplifts. Strikingly, less than a quarter of the net land stock value uplifts would meet the required investment to build the LRT system.
Our findings and discussions expand the literature available about Space Syntax from two perspectives: first, about the applicability of such a method in the Global South context and derivable knowledge to map urban access; second, its contribution to improving land value modelling while spatial dependence is not ignored. The framework demonstrates how a comprehensive operationalization of urban accessibility allows the estimation of impacts of future transport infrastructure, even when such infrastructure is new to the city being considered. The modelling workflow offers a reference for practitioners in overcoming the difficulties of estimating land value uplifts. Overcoming such difficulty contributes to the ongoing regional discussions about the feasibility of adopting financial mechanisms to secure funding for sustainable transport investments. It is this aspect that means the framework could provide a bridge for collaborations between planners and land administration practitioners.
The limitations reported in this dissertation suggest directions for future research. It is vital to advance on the validation and reproducibility of the framework. The modelling approach could be further automated and developed focussed, and be more centred, on an end-users approach. Capitalization of land as a function of accessibility is a result of a rather complex, and not fully measurable, process that is also interlinked with other dynamics over time and with feedback loops. The modelling framework could be further embedded in more complex modelling architectures, such as those dedicated to modelling the Land Use and Transport Interactions (LUTI). Naturally, the understanding of how new transport infrastructure influences cities should not remain narrowly focused on economic benefits. It is therefore paramount to link the findings and the opportunities that the contributions of this research offer towards achieving an integral and balanced perspective within sustainable transport planning.
