skip to primary navigationskip to content

John Douglas Hunt: Abstract and background reading

Title image HuntJD.AUM2020.Assessing Spatial Economic Impacts COVID Edmonton PECAS.TitleImage

SHORT ABSTACT: The Province of Alberta in Canada has sponsored the development and application of ASET, the Alberta Spatial Economic Transport model based on the PECAS theoretical framework and software.  The City of Edmonton in Alberta is using this model to consider the potential impacts of the changes in behavioural attitudes arising with COVID-19 on the spatial economic and transport system in the Edmonton Region. This presentation describes the assumptions made about the changes in behavioural attitudes arising with COVID-19 starting in 2020 and the indications provided by ASET regarding the impacts of these changes to the year 2039.  This demonstrates the inner workings and results of the mechanisms and interactions represented in ASET, and how these as an integrated set of explicit model components can inform economic, land use and transportation planning and adjustments in this planning in response to COVID-19.

READ MORE: The Province of Alberta in Canada has sponsored the development and application of ASET, the Alberta Spatial Economic Transport model based on the PECAS theoretical framework and software.  The City of Edmonton in Alberta is using this model to consider the potential impacts of the changes in behavioural attitudes arising with COVID-19 on the spatial economic and transport system in the Edmonton Region. This work is running the full ASET model that provides coverage for the entire Province of Alberta, but is considering the changes in expected trends and potential impacts of policy options for just the City of Edmonton.

This presentation describes the assumptions made about the changes in behavioural attitudes arising with COVID-19 starting in 2020 and the indications provided by ASET regarding the impacts of these changes to the year 2039.  This demonstrates the inner workings and results of the mechanisms and interactions represented in ASET, and how these as an integrated set of explicit model components can inform economic, land use and transportation planning and adjustments in this planning in response to COVID-19.  These inner workings are described in relation to their operation with the scenarios considered here.  Full descriptions of ASET and of the PECAS Framework and its use in the development of ASET are not included in this presentation.  See the references listed below for such descriptions.

This presentation describes one in a set of “what if” style analyses underway.  A specific set of assumptions about the changes in behavioural attitudes from 2020 forward because of COVID-19 are input to the model system, and the model simulation for the period from 2019 to 2039 is compared with a reference case without these assumptions.  This provides indications of the impacts of COVID-19 based on these assumptions.  In this case, the set of assumptions is viewed as a “reasonable extreme” that can be used to explore what would help establish a reasonable outer-bound of the impacts for planning considerations. A range of other less extreme sets of assumptions together with different elements of policy responses are also being considered with ASET.

The set of assumptions about shifts in behavioural attitudes arising with COVID-19 and the corresponding changes to ASET considered here are set out below as 4 groups of changes, with single-letter labels established for each group of changes as follows:

“t” changes: Challenges with social distancing and sharing air with strangers makes the use of bus and rail forms of public transport less attractive into the future.  The initial very extreme reactions giving rise to drops in ridership of 90% are tempered, but there is a remaining negative impact on transit use utilities. In ASET, the utility function for transit use is adjusted for both bus and rail in the mode choice models for all purposes. The sensitivities to transit ride times are doubled, making the coefficients twice as negative, and the alternative specific constants for transit are decreased by the equivalent of an additional 10 minutes of ride time.  The resulting reductions in utility values feed through the mode choice and destination choice models, impacting accessibilities and thereby impacting decisions about activity patterns, exchange locations, technologies and activity locations, and therefore rents and the spatial patterns in the economy and the built form.

“w” changes: With the widespread support of employees and employers, Work-at-Home is adopted extensively for white collar occupations that use office space, with a “hot desk” system adopted such that only 10% of the previous office space requirements are maintained by employers for these functions and the rest of these space requirements are off-loaded to households. COVID-19 is a catalyst for encouraging Production activities using other than white collar labour in office space are not impacted in this way specifically. In ASET, the vectors of technical coefficients for the use of office space by management activities across industries are factored by 0.1, impacting the magnitudes and elasticities of demand for office space and thereby impacting rents and the further development of supply for office space.  Work-at-Home decisions are expanded accordingly for white-collar occupations.  These changes impact the accessibilities for both labour make and use and the spatial distribution of demands for office support services, impacting the spatial patterns in the economy and the built form. 

“e” changes:  Visiting shops and gathering for entertainment and recreation in person are discouraged initially more by government request or edict and eventually more by ongoing personal concerns about COVID-19 transmission and the discovery of stay-at-home alternatives. One quarter of in-person retail activity shifts to stay-at-home with on-line/telephone/text selection and home delivery and one half of recreation and entertainment activity shifts to stay-at-home. In ASET, the generation rates of visits to production locations per unit of the retail margin put is reduced by 25% across all household types and the generation of commercial vehicle visits from retail employment to residential locations is increased correspondingly. These changes impact the travel by households for shopping and the transport of goods to residential locations, which in turn impacts the use of gasoline and diesel fuels and the associated CO2 and PM10 emissions. The technical coefficients for the consumption of recreation and entertainment services is factored by 0.5 for all household types.  These changes impact the demand for recreation and entertainment services, impacting accessibilities to labour and other services and thus the spatial patterns in the economy.

“m” changes: Concerns about social distancing and sharing common areas (in particular their air and surfaces) and the potential mixing of air by HVAC systems makes living in multifamily residential structures less attractive in the future. Media reports of outbreaks in multifamily act to reinforce this continuing bias.  In ASET, the utility functions for multifamily economy and multifamily luxury space types are adjusted by decreasing the alternative specific constants by 40% of the difference with the alternative specific constant for the corresponding single family detached type, which is equivalent to increases in rent of about 250 $ per month. These changes reduce the demand for multifamily space, which lowers rents and slows development of this space and increases the use of space thereby reducing densities; and the concentration of this space in central areas results in differential impacts on demands for services and accessibilities and thus impacts the spatial patterns in the economy and the built form.

Comparing the model simulations for 2020 to 2039 with COVID-19 vs reference case (without COVID-19) it is seen that:

  • Overall growth slowed slightly, still forced in ASET by Province-wide forecasts; slows more in Edmonton City than Edmonton Region and more for employment than population
  • Office Space market collapses Province-wide, including Edmonton City and Region – based on assumption re-purposing not permitted – so consider alternative uses for office space, including residential with low-cost emphasis
  • Employment decentralized, shifts from Centre and Inner City to Outer City and Exurbs; driven by work-at-home, reduced transit appeal and shifting accessibilities
  • Population and residential units growth shifts from Centre; driven by reduced multifamily appeal and shifting accessibilities
  • Slower growth in travel; auto use, both SOV and HOV, less growth in aggregate and larger proportion; transit use less in aggregate and in proportion; driven by lower transit utilities and reduced centralization
  • Slower growth in vehicle flows, range of 60-80 % of growth in previous forecasts for 2040
  • Similar growth in vehicle fuel use overall; less growth in gasoline use and slower increase in CO2 from vehicles; more growth in diesel use and slightly greater increase in PM10
  • Lower rents for both residential and non-residential space, particularly in Centre and Inner City; lower densities in occupation and built form

These results indicate the larger planning goals of increasing the density of built form, reducing sprawl, shifting to an increased reliance on transit and cycling and walking modes together with a reduced reliance on the private vehicle will all be even more difficult to pursue with the changes in behavioural attitudes arising from COVID-19 as assumed in this analysis.  It is an extreme case considered here, so these indications are to be viewed as indicative of a reasonable outer bound for expectations.  Further analysis with ASET using some more tempered assumptions and with policy changes seeking to mitigate and/or re-direct some tendencies  – including in particular allowing and even encouraging the transformation of office space into residential – can help further guide policy actions.

These results also indicate some slowing of roadway infrastructure development and some actions to address the use of diesel in delivery vehicles should also be considered. 

Acknowledgements

The work described here and the development of the ASET model overall has required a large effort.  It has benefitted from substantial contributions by John Abraham, Geraldine Fuenmayor, Amila Silva, Graham Hill, Cecilia Barbosa, Shena Kaul, Harley Gorrell, Kevin Stefan, Alan Brownlee, Paul McMillan, Sandeep Datla, Arun Bhowmick, Joydip Majumder, Cherry Liu, Robert Duckworth, Jun Yang, Lorraine Jonasi and Tariq Shahriar. 

Any errors contained herein are solely the responsibility of the author.  The material presented and the opinions expressed are entirely those of the author, and cannot in any way be taken to represent the policy or considerations of The City of Edmonton, The Province of Alberta or any other agency.

BACKGROUND READING

Abraham JE and Hunt JD, 2007, Random utility location/production/exchange choice, the additive logit model, and spatial choice microsimulations.  Transportation Research Record 2003:1-6.

Fuenmayor G, Abraham JE and Hunt JD, 2019, Building a PECAS activity allocation module: The experience from Caracas. Journal of Transport and Land Use 12(1):443-474. DOI: https://doi.org/10.5198/jtlu.2019.1188

Hunt JD and Abraham JE, 2005, Design and implementation of PECAS: A generalized system for the allocation of economic production, exchange and consumption quantities; Chapter 11. In: Foundations of Integrated Land-Use and Transportation Models: Assumptions and New Conceptual Frameworks. Elsevier, Oxford UK, pp:217-238.

Hunt JD, Abraham JE, Silva A, Fuenmayor G, Hill GT and Kaul S, 2019, Residential demand and supply in the Alberta PECAS Model; Paper F1_OS3_03. Online Compendium of Papers of the 15th World Conference on Transport Research, Mumbai, India, May 2019

Hunt JD, Abraham JE, Fuenmayor G, Silva A and Hill GT, 2018, Update on Alberta Spatial Economic and Transport Model Project. Presented at the Symposium on Innovation in Applied Urban Modelling (AUM2018), Cambridge UK, June 2018.

Hunt JD, Zhong M, Hill GT, Kaul S and Abraham JE, 2017, PECAS Update: Some Current Projects and Methods. Presented at the Symposium on Innovation in Applied Urban Modelling (AUM2017), Cambridge UK, June 2017.