Will COVID‐19 hinder or aid the transition to sustainable urban mobility? Spotlight on Portugal's largest urban agglomeration

Abstract

This study discusses the possible effects of two major COVID‐19 disruptions – that is, the widespread adoption of working from home and the ‘bio‐security’ fear affecting public transport – on travel patterns in Lisbon urban agglomeration. Drawing on evidence from the literature on the effects of teleworking on travel demand and the determinants of public transport, together with the analysis of travel data before and during the pandemic, the study concludes that the pandemic may hinder, more than aid, the transition to a sustainable mobility paradigm in Lisbon and its metropolitan area. The analysis of monthly travel demand shows that car use has recovered at a faster rate than public transport use, reaching pre‐pandemic levels in some of the main arterial roads serving Lisbon. There appears to be a substitution of public transport trips by both driving and active micromobility. The main losers so far are public transport operators: the pandemic increased operating deficits and the need for additional public subsidies, putting at risk the future financial sustainability of public transport. Without serious commitment by local and national governments, the pandemic is likely to be another missed opportunity to transition to a better, more sustainable, urban mobility system.

1. INTRODUCTION

The coronavirus disease 2019 (COVID‐19) pandemic has caused major disruptions to the way we live, work and travel. Some of these disruptions may accelerate processes that were already taking place, but their outcomes remain uncertain and difficult to predict, such as their possible effects on urban mobility patterns. During the earlier phases of the pandemic, there was a sense that cities (and societies more generally) would use this crisis as an opportunity to speed up the transition to more sustainable living and travel. The sudden improvement in the quality of the urban environment resulting from the reduction in road congestion, noise and air pollution brought to light the unresolved conflict between ecology and the economy. As the crisis progressed, and its negative impacts on the economy deepened, the urge for recovery, strongly assisted by rising vaccination levels, revealed a rather familiar reality, with mobility patterns heavily based on car use as well as road congestion and pollution increasing to pre‐pandemic levels.

The capability of cities (and countries) to shift travel mobility paradigms is influenced by many factors, notably the ways in which land use and transport systems interact with each other and how their design determines the possibility to travel using different transport modes. Portugal provides an interesting case study for over‐investment in road infrastructure, particularly motorways, and disinvestment in public transport, notably railways. According to EUROSTAT data, in 2017 Portugal had the third‐highest endowment of motorways relative to gross domestic product (GDP) and the fifth‐highest endowment of motorways relative to population in the European Union (EU). Motorway construction was particularly generous in the two metropolitan areas; for example, Lisbon metropolitan area is the NUTSII region in Europe with the second‐highest density of motorways per square kilometre (sixth highest per person) (Padeiro, 2018). Recent evidence for Portugal suggests that the expansion of the motorway network between the mid‐1990s and late 2000s contributed to suburbanization (Rocha et al., 2021), promoting further car dependency and lack of improvement of public transport systems.

This paper contributes to the debate on the challenges and opportunities created by the COVID‐19 pandemic regarding the transition to more sustainable travel mobility, with a special spotlight on Portugal's largest urban agglomeration, which accounts for approximately one‐third of the total population and employment in the country. I focus the discussion on the potential effects of two of the major disruptions caused by the pandemic, notably the large‐scale change in working practices (i.e., working from home and flexible working practices) and the health‐related penalty imposed on public transport systems owing to the fear of virus contamination. The impact of these disruptions on travel behaviour in the short and longer term is not necessarily the same because people adjust their decisions to new circumstances over time and in response to public policy. Consequently, the fate of urban mobility is likely to differ according to city‐specific measures, notably local governments’ efforts to prevent the substitution of public transport by driving.

The analysis suggests that there is a ‘mobility paradox’ in Lisbon, consisting of a strong substitution of public transport trips by both solo driving and active micromobility. I use the evidence from the literature on teleworking and travel demand, and the determinants of public transport demand, to discuss how the COVID‐19 pandemic may hinder, rather than aid, the transition from the current car‐based mobility paradigm to a more sustainable urban travel mobility paradigm. The discussion is supported by monthly travel data for different transport modes before and during the pandemic, which shows that public transport demand remains substantially below pre‐pandemic levels whereas traffic has essentially recovered to pre‐pandemic levels. In addition, the prospects for sustainable urban mobility worsened further after the local elections in September 2021, with the new, centre‐right, elected mayor promising cheaper car parking for residents and putting at risk some of the major routes of the planned cycle network, which may now face additional hindrances to its full implementation.

The paper is organized as follows. Section 2 gives a brief overview of the two studies directly relevant to this study. Section 3 describes the data collected and their respective sources, whereas Section 4 analyses travel patterns in Lisbon and the wider metropolitan region before and during the COVID‐19 pandemic. Section 5 discusses the implications of the disruptions in working practices and public transport for the future of sustainability of urban mobility in Lisbon. Section 6 provides concluding remarks.

2. WORKING FROM HOME, ‘BIO‐SECURITY’ FEAR, AND PUBLIC TRANSPORT DEMAND: WHAT CAN WE LEARN FROM EXISTING RELATED LITERATURE?

2.1. Teleworking and travel demand

Existing research on the effects of teleworking on travel demand spans back to the early 1990s and increased substantially as the adoption of information and communication technology (ICT) became more widespread. The initial view was that teleworking could be used as a travel demand management strategy to reduce the number of trips and distances travelled (e.g., Choo et al., 2005; Hamer et al., 1991; Helminen & Ristimäki, 2007; Henderson et al., 1996; Kitamura et al., 1991; Mokthtarian et al., 1995; Nilles, 1991; Pendyala et al., 1991), but more recent studies tend to conclude that teleworking can actually induce more travel or, at best, permits only minor travel reductions (e.g., de Abreu e Silva & Melo, 2018a; de Abreu e Silva & Melo, 2018b; He & Hu, 2015; Hjorthol & Nossum, 2007; Kim et al., 2015; Melo & de Abreu e Silva, 2017; Nelson et al., 2007; Zhu, 2012; Zhu & Mason, 2014).

The mechanisms underlying the relation between teleworking and travel behaviour are multiple and complex, including factors relating to individual and household preferences, demographic and socioeconomic circumstances, and the land use characteristics of the residential and workplace locations. Besides the direct effects of teleworking on commuting trips, there can also be indirect effects on non‐work travel, travel mode choice and residential location, all of which affect the overall net effect of teleworking on travel demand (e.g., de Abreu e Silva & Melo, 2018a; de Abreu e Silva & Melo, 2018b).

The effects of working from home on travel demand can also differ over time. In the shorter term, teleworking can reduce the scope of trip chaining, increasing the number of non‐work trips on teleworking days (Pendyala et al., 1991). There can also be longer‐term effects on travel demand through residential relocation away from workplace locations. The increased disposable income due to savings in commuting cost and travel time may facilitate residential relocation to peripheral urban areas, contributing to sprawl and longer commutes (Mokhtarian, 2009; Nilles, 1991). Since the quality of public transport is generally inferior in suburban areas, teleworking may increase car use levels in the longer term through residential relocation. Consequently, although the total number of commuting trips of teleworkers may be smaller compared with non‐teleworkers, the total distance travelled by car may be greater. The overall net outcome is difficult to predict and is likely to be specific to both the household and the local context.

The COVID‐19 pandemic led to the adoption of working from home by the large majority of (non‐essential) workers, as well as the elimination of most out‐of‐home activities due to lockdown restrictions. The result was a sudden fall in mobility levels and travel demand by both private and public transport. Although working from home may become a widely accepted practice in post‐COVID‐19 vaccinated economies, at least for some occupations and industries, its circumstances will be very different owing to the elimination of constraints on travel and social life. Evidence from pre‐pandemic teleworking studies may be helpful in predicting the effects of working from home on travel demand patterns, but only to some extent. People's perceptions about the value of their time, the value of their local residential environment and the safety of different modes of transport, amongst other factors, have been affected by the pandemic and are likely to influence travel behaviour in uncertain ways. The simplest scenario would be an overall reduction in the number of trips due to less commuting. This effect may be more attractive for workers living further away from their workplace, in which case there would also be a more than proportional reduction in distance travelled and air pollution levels. Reality is more complex, though, and people are likely to use the saved travel time and money to do other activities they like, which may require travel. They may also decide to use the increased disposable income to relocate for greener pastures and put up with longer, even if fewer, commutes. Changes in household location are likely to affect not only home‐to‐work commute distances, but also travel mode choice. In particular, moving out of urban areas tends to be associated with an increase in car ownership levels, so the overall effect depends on households' residential preferences and the extent to which the pandemic may alter them.

2.2. Determinants of the demand for public transport

Fear of virus contamination, despite increasing levels of vaccination, makes public transport a less attractive travel mode choice. People might avoid using public transport because they perceive it as being unsafe, compared with private transport, because of having to share a confined and often crowded space, where keeping a minimum social distancing can be difficult (e.g., De Vos, 2020; Dong et al., 2021). The resulting reduction in passenger numbers, and thus in operating revenues, exacerbates the need for public subsidies, compromising the maintenance and investment in services, as well as the financial sustainability of public transport systems. The unusual circumstances created by the COVID‐19 pandemic may also limit the efficacy of the standard measures used in the past by operators and municipalities to promote the use of public transport, notably pricing policies, which may now be substantially less effective.

In this section, I draw on the conclusions from review studies of the determinants of public transport demand (e.g., Holmgren, 2007; Litman, 2004; Paulley et al., 2006) to discuss some of the possible implications of the pandemic for the future of public transport in post‐COVID‐19 vaccinated economies. The price of travelling by public transport is one of the main factors influencing the demand for public transport. Existing evidence suggests that price elasticity ranges between −0.2 and −0.5 in the short term and between −0.6 and −0.9 in the longer term (e.g., Holmgren, 2007; Litman, 2004; Paulley et al., 2006). This means that an increase of 10% in the price of travel cards is associated with a reduction in patronage of 2–5% and 6–9% in the short and longer term, respectively. The precise value of the elasticity, however, can differ with the geographical context (e.g., built environment and land uses), the type of user (e.g., income level, economic status) and journey purpose (e.g., commuting, shopping) (Litman, 2004). Importantly, the fact that demand for public transport is usually relatively inelastic means that increasing prices can also increase net revenues, ceteris paribus. This relation may be fundamentally altered by the fear of virus contagion created by the pandemic, which may make people more responsive to rises in price levels, especially for higher‐income users with more alternative travel options available, making the negative impact on operating revenues more severe.

The quality of service, notably the frequency level, is another major determinant of public transport demand. Changes in frequency levels (e.g., trains per hour) can have a relatively stronger impact on patronage, compared with changes in pricing. In other words, the demand–supply elasticity estimates can often exceed (in absolute value) the demand–price elasticity estimates (e.g., Melo & Ramli, 2014). The ‘bio‐security’ threat created by the COVID‐19 pandemic may further emphasize this pattern owing to the now even greater importance of providing more frequent, faster and safer journeys to users. Compared with pre‐COVID‐19 times, both travel times and waiting times are now perceived as being relatively more valuable than the monetary cost of a trip. To ensure minimum social distancing, public transport operators had to provide service levels that were higher than pre‐pandemic for demand levels that were lower than pre‐pandemic, increasing the gap between operating revenues and operating costs.

The price of alternative travel modes, namely driving, is another important factor affecting the demand for public transport. Cross‐price elasticities between demand for public transport and fuel prices (i.e., the main indicator for the cost of driving) reveal there is a clear substitution between public transport and driving. Furthermore, this effect becomes stronger in the medium and longer term as people adjust their behaviour to the new relative prices. Accessibility to free or under‐priced parking is also a major predictor of driving.

3. DATA AND METHODOLOGY

This section provides a description of the variables, and respective data sources, collected to analyse travel patterns in Lisbon and its metropolitan area in the recent past. Describing travel patterns in Lisbon, and more generally in Portugal, is challenging owing to the lack of regular measurement tools. Travel surveys are carried out occasionally, ruling out the possibility of developing longitudinal studies. Consequently, the majority of travel‐related analyses rely on the population censuses, which collect data on commuting distances, commuting times and main transport modes every 10 years.

The most recent travel survey was carried out in 2017 by Portugal's Statistics Office (INE) and covered the municipalities of Lisbon and Porto metropolitan areas: Inquérito à Mobilidade nas Áreas Metropolitanas do Porto e de Lisboa, IMob 2017 (INE, 2018). The IMob travel survey contains information for different types of statistical units, namely households, individuals, vehicles and trips. The survey is organized into different sections, collecting information on residential location, household size and characterization of its members (e.g., age, gender, education, occupation), household income, characterization of individual travel mobility (e.g., travel distance, travel time, mode of transport), characterization of vehicles available, daily trips (e.g., destination, mode of transport, purpose), expenditures with transport (e.g., fuel, parking, tolls, public transport) and individual opinions about/motivations for using the car or public transport in regard to comfort, safety, accessibility, travel time, cost, etc. 1 The IMob is the best, most recent and detailed data on travel behaviour and mobility patterns in Lisbon and its metropolitan area before the pandemic. To analyse travel demand during the pandemic, it was necessary to collect data from a range of sources, including public transport operators, Statistics Portugal, Lisbon municipality and the Institute for Transport and Mobility (IMT). Table 1 provides a summary of the data used in the empirical analyses.

TABLE 1.

Variables and data sources

Variables		Data source	Data type	Data provider
2017	Car ownership, modal split, mean distances and travel times	Inquérito à Mobilidade nas Áreas Metropolitanas do Porto e de Lisboa, IMob 2017.	Cross‐sectional	Statistics Portugal (INE)
2018, 2019, 2020, 2021	Average daily traffic in main urban motorways and arterial roads connecting Lisbon to the rest of the metropolitan area	Relatório de Tráfego na Rede Nacional de Autoestradas. a	Time series, monthly	Instituto da Mobilidade e dos Transportes (IMT)
	Metro passengers – Metro de Lisboa	Inquérito ao Transporte por Metropolitano (IMTM).	Time series, monthly	Statistics Portugal (INE)
	Bus passengers – CARRIS	Inquérito ao Transporte Rodoviário de Passageiros (ITRP); direct inquiry.	Time series, monthly	Statistics Portugal (INE); CARRIS
	Ferry passengers – Soflusa & Transtejo	Inquérito ao Transporte fluvial (ITF).	Time series, monthly	Statistics Portugal (INE)
	Rail passengers – Fertagus	Inquérito ao Transporte por Caminho de Ferro; Fertagus train operator. Direct inquiry.	Time series, monthly	Statistics Portugal (INE); Fertagus
	Rail passengers – CP		Time series, monthly	Statistics Portugal (INE)
	Average number of cyclists per hour	Moura et al. (2021)	Repeated cross‐sectional	Lisbon municipality

^a https://www.imt‐ip.pt/sites/IMTT/Portugues/InfraestruturasRodoviarias/RedeRodoviaria/Paginas/Relatorios.aspx.

The collection of data from different sources to obtain repeated time series data for local travel has some limitations. One limitation is that we cannot say anything about the evolution of modal split in the period studied, despite having time series data for public transport operators, traffic counts and cyclist counts. Another limitation is that the data do not permit identifying modal shift between transport modes due to specific events such as COVID‐19. This is unfortunate given its importance for policy‐making. Consequently, it is very difficult to unveil the underlying causes of changes in travel mobility patterns, notably the distinction between size effects (i.e., a genuine reduction in mobility) and composition effects (i.e., a shift in mobility between travel modes). This distinction is fundamental for public policy.

The methodology consists of two sets of empirical analyses. The first analysis describes travel patterns in Lisbon before the pandemic using data from the aforementioned IMob 2017 travel survey (i.e., Section 4.2). The second analysis describes travel demand after and during the pandemic using monthly data between 2018 and 2021 on the (i) number of passengers using the main public transport modes; (ii) traffic counts for the main urban motorways and arterial roads serving Lisbon; and (iii) the number of cyclists based on bi‐annual cycle counting campaigns (i.e., Section 4.3).

Figure 1 shows a map of public transport systems serving Lisbon and the other municipalities of the metropolitan area. Appendix B shows a map of the metropolitan area and its 18 municipalities. There is one municipal bus operator in Lisbon (i.e., CARRIS), operating essentially within the borders of the municipality, whereas bus services in the other municipalities of the metropolitan areas are provided by ten private operators. There are two railway operators, the state‐owned Comboios de Portugal (CP), which is the main rail operator in the country, and the private rail operator Fertagus, which connects the north and south banks of the metropolitan area. There are also two metro operators: the state‐owned Lisbon metro (ML), which functions mostly within its borders, and the privately owned light rail system MST in Almada connecting to the Fertagus railway corridor (MSL and Fertagus are owned by the same private group). There are also ferry connections between Lisbon and Almada operated by the state‐owned operator Transtejo & Soflusa. Figure 2 shows the main trunk roads in the metropolitan area, that is, principal itineraries (IP), secondary itineraries (IC), national roads (EN) and regional roads (ER). The main motorways connecting Lisbon to the other parts of the metropolitan area include the northbound A1/IC2 corridor, the southbound IP7‐A2 corridor, the westbound A5 corridor, the northwest‐bound IC19, IC17/A36 and A8 corridors, and the eastbound IP7 corridor.

FIGURE 1.

Public transport services in Lisbon metropolitan area. Source: www.aml.pt

FIGURE 2.

Main road networks in Lisbon metropolitan area. Source: www.aml.pt

4. MOBILITY PATTERNS BEFORE AND DURING THE COVID‐19 PANDEMIC

4.1. Overview of COVID‐19 phases and the impact on public transport operators

So far, there have been two major lockdown periods in Portugal. The first lockdown state of emergency was declared on 19 March 2020 and lasted until 2 May 2020, after which the country moved into a calamity status until 30 June. Between 1 July and 14 September 2020, the country lived in a general alert status, although some areas of Lisbon metropolitan area remained in a contingency status with more stringent lockdown restrictions on travel mobility. From 15 September until 14 October 2020, there was a new general contingency status with specific teleworking arrangements in the metropolitan areas. From 15 October to 8 November 2020, the country moved into a calamity status with limited cross‐municipality mobility. From 9 November 2020 until 7 January 2021, the country re‐entered a state of emergency, but there was an exceptional easing of restrictions on mobility during the Christmas period. The state of emergency was continually renewed until 15 March 2021, after which the government rolled out a plan for the re‐opening of the economy, in tandem with a cohort‐based vaccination plan for the population. As of 15 November 2021, approximately 87% of the population had been vaccinated and domestic travel mobility conditions were essentially the same as in pre‐pandemic times. 2 Appendix A provides a diagrammatic overview of the pandemic phases in terms of the severity of constraints on travel mobility and social meetings.

Public transport operators responded to the decline in travel demand by reducing the level of services provided while maintaining designated minimum service levels (i.e., 50% of usual pre‐pandemic winter frequency level) for essential workers and travel. Passenger levels were limited to either one‐third or two‐thirds of the maximum capacity, depending on the severity of the pandemic and lockdown status, a limitation which was abolished only in August 2021. According to a report produced by the transport authority for Lisbon metropolitan area (TML, 2021), whilst the level of services provided by public transport operators had resumed to pre‐pandemic levels in September 2020, demand remained well below pre‐pandemic levels. The number of monthly travel cards sold by all public transport operators in the metropolitan area between April and December 2020, compared with the same period in 2019, varied between 12% (April) and 65% (October). In addition, passenger numbers in 2020 were only about 57% of the passengers transported in 2019. Although comprehensive comparable data for 2021 are not yet available, the statistics reported by the different public transport operators suggest that passenger numbers are still considerably below pre‐pandemic levels, whilst service levels are at least as high as before the pandemic. The widening gap between operating revenues and costs has worsened the deficits faced by public transport operators, putting at risk their financial sustainability. To compensate public transport operators for the increasing operating deficits, the national government created bespoke legislation allowing for the allocation of pre‐existing funds to COVID‐19 financial support (i.e., PART, 3 PROTransP 4 and funding for discounted travel cards for children, students and the elderly).

4.2. Mobility patterns in Lisbon and the metropolitan before the COVID‐19 pandemic

Mobility patterns in Lisbon and its metropolitan area are strongly car dependent. According to data from the decennial censuses and the recent IMob travel survey, car use increased drastically over the past 40 years: from 13% of total commuting trips in 1981 to 22% in 1991, 43% in 2001, 54% in 2011 and 59% in 2017. In contrast, commuting by public transport decreased from 62% in 1981 to 47% in 1991, 34% in 2001, 28% in 2011 and 16% in 2017 (INE, 1992; INE, 2002; INE, 2003; INE, 2012; INE, 2018). If we consider trips for all purposes, walking and cycling accounted for 23% and 0.5% of trips, respectively, according to the IMob travel survey carried out in 2017. There are no data referring specifically to cycling in the decennial censuses.

In this section, I provide a brief characterization of travel patterns in Lisbon and the metropolitan area, based on the recent IMob travel survey, in terms of car ownership, travel distance and travel time by transport mode for intra‐ and cross‐municipality travel. Where possible, I focus specifically on commuting data and their characteristics because of the direct effects of working from home on this type of trip. I also distinguish between municipalities according to data availability (see Appendix B for a map of the 18 municipalities in the metropolitan area).

Level of car ownership varies across municipalities, with lower values observed for municipalities with higher population density (Table 2). This is a well‐known regularity in travel demand: denser areas tend to have shorter journeys and can sustain better public transport systems, ceteris paribus, contributing to lower levels of car use (e.g., Cervero & Kockelman, 1997). Car ownership also increases with income level, which tends to be higher in denser urban areas owing to agglomeration economies (e.g., D'Costa & Overman, 2014; Melo & Graham, 2009). On average, across the metropolitan area, 61% of households have one car, 29% of households have two cars and 6% of households have three or more cars. The mean number of cars per household, considering only households with at least one car, is 1.44 for the whole metropolitan area and ranges between 1.30 for Amadora and 1.55 cars for Alcochete and Cascais. The share of households owning at least one car is very high, exceeding 90% in all municipalities; the share of households with three or more cars is also considerable, ranging between 4% for Almada and Amadora and 8% for Alcochete, Barreiro and Palmela.

TABLE 2.

Car ownership by municipality

Municipality	Population	Share population (%)	Population density (people per km2)	Percentage of households with 0, 1, 2 or 3+ cars
				0	1	2	3+	Mean number of cars1
Alcochete	19,020	0.67	148	3.2	52.5	36.0	8.4	1.55
Almada	169,330	6.00	2,419	5.3	62.4	28.2	4.2	1.39
Amadora	178,169	6.32	7,492	5.4	70.2	20.2	4.2	1.30
Barreiro	75,978	2.69	2,088	2.1	61.3	28.4	8.1	1.48
Cascais	210,889	7.47	2,165	3.5	52.9	36.3	7.3	1.55
Lisboa	504,964	17.90	5,047	4.4	66.1	24.9	4.5	1.36
Loures	207,567	7.36	1,241	4.9	57.7	32.2	5.1	1.45
Mafra	82,581	2.93	283	4.8	54.0	34.7	6.4	1.51
Moita	64,767	2.30	1,172	4.7	61.9	26.0	7.4	1.44
Montijo	55,742	1.98	160	3.8	57.4	31.3	7.5	1.49
Odivelas	156,083	5.53	5,881	1.6	66.5	25.5	6.3	1.40
Oeiras	174,249	6.18	3,798	2.9	57.3	32.8	6.9	1.50
Palmela	64,146	2.27	138	5.8	54.0	32.1	8.1	1.53
Seixal	165,123	5.85	1,730	2.2	65.0	28.2	4.6	1.39
Sesimbra	50,972	1.81	260	5.6	56.5	30.8	7.1	1.49
Setúbal	116,979	4.15	508	3.8	57.5	32.1	6.6	1.48
Sintra	383,946	13.61	1,203	2.0	62.4	29.4	6.2	1.44
Vila Franca de Xira	140,844	4.99	443	3.3	57.5	34.1	5.2	1.47
AML	282,1349	100	936	3.7	61.2	29.3	5.8	1.44

Notes: ¹Considering households with at least one car. The figures reported were estimated using sampling weights for households.

Source: Author, based on IMob 2017 microdata.

Given the high levels of car ownership, it is not surprising to observe, in Table 3, that the car is the main travel mode, accounting for 59% of all trips in the metropolitan area and ranging between 45% of all trips in Lisbon and 78% of all trips in Mafra. In 7 out of the 18 municipalities, at least two‐thirds of all trips are made using the car. Public transport (i.e., bus, metro, train, ferry) accounts for a maximum of 21% of total trips in Lisbon, followed by Almada and Odivelas with 19% each, Amadora, Barreiro and Loures with 18% each, and Seixal and Sintra with 15% each. The remaining municipalities have substantially lower levels of public transport use. Compared with other European capital cities, Lisbon has relatively worse performance in terms of public transport use, even when compared with its main Iberian neighbours Madrid and Barcelona, which have a share of public transport use of circa 40% and 30%, respectively (Urban Audit, 2011).

TABLE 3.

Modal split (%) of total trips by municipality of residence

	% trips by mode1								Work‐related trips/all trips2
	Car	Bus	Train	Metro	Ferry	Bicycle	Walk	Other
Alcochete	67.8	8.7	0.4	1.6	0.8	1.3	18.2	1.1	32%
Almada	57.2	10.7	2.4	4.8	0.7	0.3	20.9	3.0	31%
Amadora	50.7	8.9	3.9	5.1	0.0	0.4	29.9	1.0	30%
Barreiro	56.6	8.9	1.2	3.0	4.9	0.5	24.3	0.3	32%
Cascais	70.4	4.6	4.6	0.8	0.0	1.4	15.6	2.5	32%
Lisboa	45.1	12.2	1.1	7.8	0.0	0.6	29.8	3.3	29%
Loures	60.9	14.0	1.0	3.0	0.0	0.3	18.4	2.5	33%
Mafra	77.7	6.4	0.0	0.3	0.0	0.2	11.5	3.9	31%
Moita	51.1	7.4	3.8	0.8	1.5	0.3	33.3	1.8	27%
Montijo	66.5	4.9	0.4	0.8	2.2	1.0	23.2	0.5	32%
Odivelas	56.1	12.2	0.3	6.8	0.0	0.1	23.2	1.3	29%
Oeiras	64.7	8.4	3.8	1.1	0.0	0.2	19.7	2.0	30%
Palmela	69.7	6.9	2.9	0.4	0.6	1.1	17.2	1.2	29%
Seixal	59.6	6.9	6.2	1.6	0.3	0.8	22.8	1.8	31%
Sesimbra	70.4	3.3	2.4	0.4	0.1	0.4	19.7	3.3	34%
Setúbal	66.0	5.0	1.7	0.1	0.2	0.2	24.9	1.9	25%
Sintra	63.4	6.6	7.8	0.5	0.0	0.2	19.5	1.9	33%
Vila Franca de Xira	58.2	7.0	5.2	0.8	0.0	0.3	27.0	1.4	35%
AML	58.9	8.8	3.2	3.1	0.3	0.5	23.0	2.2	31%

Notes: ¹Based on data from Table IV.1 of the dataset containing the results for Lisbon metropolitan area. ² Based on Figure 121 of INE (2018) and Table IV.1 of the dataset containing the results for Lisbon metropolitan area.

Source: Author, based on municipal‐level tables from IMob 2017.

With regard to trip distance and travel time by travel mode and purpose, Table 4 shows that, for work‐related journeys, people using the car spend, on average, 25 min (15 km) travelling, in contrast to 41 min for bus (18 km) and metro (9 km), and 51 min (19 km) for train. The respondents using the bicycle for work‐related trips spend, on average, 32 min (8 km) travelling. These figures provide further evidence of low public transport performance in relation to other modes of travel.

TABLE 4.

Mean distance and travel time of all trips and work‐related trips by transport mode in the metropolitan area

Transport mode	Total trips		Work‐related trips
	Distance (km)	Travel time (min)	Distance (km)	Travel time (min)
Car	13.1	21.3	15.2	25.2
Bus	14.7	39.2	18.0	40.8
Train	19.1	53.4	18.5	51.0
Metro	8.6	39.7	8.9	41.4
Ferry	19.5	58.1	19.0	54.2
Bicycle	8.8	36.2	8.0	32.2
Walk	1.5	17.0	1.4	15.1
Other	14.0	25.0	20.5	32.7
Overall mean	11.0	24.5	14.8	29.5

Notes: Based on Tables IV.16 (travel time) and IV.17 (distance) of the dataset containing the results for Lisbon metropolitan area.

Source: Author, based on municipal‐level tables from IMob 2017.

Table 5 and Figure 3 reveal interesting differences in travel patterns and modes across municipalities. Despite the mean distances and travel times of intra‐municipal trips being substantially shorter compared with inter‐municipal trips, they are nonetheless heavily based on car use. Amadora and Lisboa have the lowest levels of car use, each with 39% of all intra‐municipal trips compared with the metropolitan‐wide value of 53%. In contrast, Mafra and Cascais have the highest levels of car use, with nearly 75% and 68% of all intra‐municipal trips done by car, respectively. Lisbon has the highest public transport use for intra‐municipal trips, with 25% of all trips. Despite the short distances travelled – the mean travel distance of intra‐municipal trips in the metropolitan area is 3.7 km – mobility is still heavily dependent on private motorized vehicles, which is further evidence of an inefficient and unsustainable mobility system in the metropolitan area. Considering inter‐municipal trips, the lowest shares of car use are recorded again for Lisbon and Amadora with 62% and 68%, respectively, whereas the highest shares are reported for Mafra (95%), Montijo (85%), Setúbal (85%) and Cascais (84%).

TABLE 5.

Inter‐ and intra‐municipal trips by transport mode, mean distance and travel time, by municipality

	Intra‐municipal trips					Inter‐municipal trips
	% intra‐municipal trips1,2	Mean distance3	Mean travel time4	% trips by car5	% trips by public transport5	Mean distance6	Mean travel time7	% trips by car8	% trips by public transport8
Metropolitan area	65.4	3.7	16.6	53.2	11.2	16.4	35.0	70.8	25.0
Alcochete	56.3	2.3	10.3	61.1	4.1	19.3	30.8	79.3	18.0
Almada	68.5	4.4	16.7	55.4	14.1	15.0	33.3	72.5	20.9
Amadora	57.1	2.6	12.2	38.7	7.3	11.1	30.8	67.8	25.5
Barreiro	65.7	3.1	12.9	56.0	7.3	18.8	31.6	69.9	23.7
Cascais	71.3	4.2	14.7	67.7	6.6	20.1	35.8	83.5	14.7
Lisboa	63.4	4.2	23.1	39.3	24.6	16.8	40.0	61.9	34.9
Loures	58.1	3.5	14.8	54.6	9.5	13.9	31.3	73.2	22.1
Mafra	80.4	5.0	13.0	73.4	6.3	29.2	36.3	94.9	4.6
Moita	66.2	2.0	10.6	41.2	5.5	17.0	29.3	68.2	21.5
Montijo	68.5	3.3	12.3	59.3	2.5	20.3	30.3	85.4	13.1
Odivelas	58.6	2.4	13.8	50.0	7.1	11.4	32.3	69.1	26.8
Oeiras	54.4	2.9	13.9	59.6	6.4	13.3	32.0	77.5	18.8
Palmela	62.3	3.7	13.0	63.5	7.0	19.3	30.0	79.5	18.3
Seixal	68.3	3.7	18.2	57.0	5.0	16.5	33.4	69.9	24.2
Sesimbra	68.9	4.0	15.0	64.8	1.3	19.6	33.9	85.4	10.0
Setúbal	79.9	3.4	13.9	64.4	3.7	23.8	37.4	73.6	18.6
Sintra	69.3	4.1	15.0	61.2	6.3	17.0	35.8	72.4	24.2
Vila Franca de Xira	74.9	3.5	14.3	51.3	4.9	22.0	38.8	69.2	26.2

Notes: ¹The share of incoming trips only considers trips within the metropolitan area. ²Based on Table V.1 of the dataset containing the results for Lisbon metropolitan area. ³Based on Table V.6 of the dataset containing the results for Lisbon metropolitan area. ⁴Based on Table V.7 of the dataset containing the results for Lisbon metropolitan area. ⁵Based on Table V.8 of the dataset containing the results for Lisbon metropolitan area. ⁶Based on Table VI.10 of the dataset containing the results for Lisbon metropolitan area. ⁷Based on Table VI.11 of the dataset containing the results for Lisbon metropolitan area. ⁸Based on Table V.12 of the dataset containing the results for Lisbon metropolitan area.

Source: Author, based on municipal‐level tables from IMob 2017.

FIGURE 3.

Percentage of trips by car: all trips (top‐left panel), intra‐municipal trips (top‐right panel), inter‐municipal trips (bottom panel). Source: Author, based on municipal‐level data from IMob 2017

4.3. The effect of the COVID‐19 pandemic on travel mobility patterns in Lisbon

To gain a better understanding of the effects of the pandemic on travel mobility in Lisbon, I collected monthly data between 2018 and 2021 for the number of passengers transported by the main public transport operators and mean daily traffic on the main arterial roads serving the city. It is difficult to separate the effect of COVID‐19 from the ‘natural trend’ in travel patterns before the pandemic, especially because the implementation of the PART programme in April 2019 changed the structure and pricing of public transport in Lisbon metropolitan area in a drastic way. Programa de Apoio à Redução Tarifária (PART) is a national programme introduced in 2019 by the Portuguese government with the objectives of increasing public transport use, mitigating the effect of the negative externalities of car use and promoting greater territorial cohesion. It provides financial support to transport authorities to reduce the price of travel cards and to coordinate public transport services with the aim of improving service levels; most of the financial support was used to reduce the price of travel cards. In the case of Lisbon metropolitan area, the myriad of existing travel cards was replaced with a single travel card for the municipality and the metropolitan area, costing 30€ and 40€, respectively. Before PART, the price of travel cards varied with distance to Lisbon, with values over 100€ for the more peripheral municipalities. PART represented savings of more than 60% for five municipalities and between 50% and 60% for four municipalities (Figure 4. According to the evaluation report published by the Institute for Transport and Mobility (IMT, 2020), there was an immediate increase in public transport demand: between May and December 2019, passenger numbers and travel card sales rose by 18% and 25%, respectively, together with a reduction in the sales of occasional tickets. Consequently, the analysis carried out in the paper cannot assess the effect of the pandemic, and home working, on the change in travel demand levels in 2020 and 2021, compared with 2019 (after PART) and 2018 (before PART). Nevertheless, it can provide indicative evidence on the overall effect of the full and partial lockdowns due to the pandemic (including the effects of teleworking, teleshopping, psychological effect of fear of virus spreading in public transport, etc.), as well as other factors relating to public transport operators' decisions on service levels, schedule and safety measures.

FIGURE 4.

Price of monthly public transport travel card in Lisbon metropolitan area before PART (left) and after PART (right). Source: IMT (2020)

Figures 5 and 6 show the monthly change in passenger number by public transport operator and mean daily traffic count between 2018 and 2021, respectively. The values are reported as a monthly fixed base index for 2018, measuring the percentage change in relation to the respective month in 2018. The figures for the absolute number of passenger and traffic counts are reported in Appendix C1, C2.

FIGURE 5.

Public transport demand in Lisbon, suburban rail & ferry connections (2018 = 100). Source: Author, based on data collected from INE and directly obtained from public transport operators. Note: Passenger numbers for the rail operator CP refer to the three main urban agglomerations in Portugal: Lisbon, Porto and Coimbra

FIGURE 6.

Mean daily traffic in urban motorways and arterial roads serving Lisbon (2018 = 100). Source: Author, based on data collected from IMT quarterly traffic reports

Figure 5 shows a continuous year‐on‐year increase in public transport demand until February 2020, after which there is a sudden drop in passenger numbers due to the implementation of the first full lockdown to reduce the spread of COVID‐19. Public transport demand between April 2019 and February 2020 shows relatively stronger growth rates, which might partially reflect the reduction in the user cost of travel cards following the introduction of the PART programme in April 2019. Interestingly, Figure 6 shows that traffic demand between April 2019 and February 2020, compared with the same period in 2018, decreased in some parts of the trunk road network, notably traffic crossing the bridge 25 de Abril (which is also served by the Fertagus rail operator). These travel trends are in line with the conclusions of the aforementioned study of the impact of PART carried out by IMT (2020).

After March 2020, following the introduction of the first lockdown period due to the COVID‐19 pandemic, we observe a drastic and immediate drop in car traffic and public transport demand. In the case of public transport, the reduction in monthly passengers in relation to March 2019 ranged between 50% (CARRIS) and 74% (Fertagus). Traffic volumes in March 2020 dropped to about two‐thirds the values recorded in March 2019. April 2020 was the first full month of lockdown and compulsory home working, recording the lowest values of travel mobility since the start of the pandemic. Travel mobility recovered gradually from May to August/September 2020, after which it reduced again because of the new constraints imposed on travel owing to the increase in the number of COVID‐19 infections and deaths. The declaration of the second lockdown state of emergency in November 2020, lasting until mid‐March 2021, led to another fall in travel demand, with monthly figures typically below 50% of 2019 pre‐pandemic values. The first 3 months of 2021 registered the lowest values of travel mobility, but since then we can observe a steady recovery. The recovery is especially notable for car use, with traffic figures recorded across the main urban motorways and arterial roads serving Lisbon in the third quarter of 2021 at nearly the level of 2018 and 2019 (Figure 6). The recovery is more heterogeneous for public transport (Figure 5): whereas passenger numbers for rail and bus services exceeded 70% (80–90%) of 2019 (2018) pre‐pandemic levels by August–September 2021, metro demand remained below 60% of 2018 and 2019 pre‐pandemic levels for the same period.

Although it is not possible to make any causal claims about the effect of home working on travel demand, the limited information on home working available for Portugal is indicative of a large effect. Data from EUROSTAT's Labour Force Survey shows that the percentage of employed persons working from home in Portugal jumped from 6% to 7% between 2011 and 2019, and to 14% in 2020, which can only be explained by the COVID‐19 lockdowns. 5 The only other source of information on home working, particularly during the pandemic, is the bespoke rapid survey Inquérito Rápido e Excecional às Empresas COVID‐19 (COVID‐IREE), carried out jointly by the Bank of Portugal and Statistics Portugal to assess the effects of the pandemic on the economy. 6 The COVID‐IREE collects information on various economic dimensions of private sector businesses (excluding the financial sector), asking specific questions about home working. It was implemented between the second half of April 2020 and July 2020, after which it was suspended given the positive evolution of the pandemic situation. The worsening of the COVID‐19 pandemic in the fourth quarter of 2020 led to the implementation of additional, occasional, editions (i.e., November 2020 and February 2021). According to the COVID‐IREE, the percentage of firms with more than 75%, 50% and 25% of the workforce on teleworking ranged between 10% (July 2020) and 16% (April 2020), 12% (July 2020) and 20% (April 2020), and 16% (July 2020) and 26% (April 2020), respectively. 7 It is difficult to predict future trends in teleworking for Portugal, but the recent advances in its legal framework, with effect from 1 January 2022, may facilitate its adoption. 8

The discussion of Figures 5 and 6 suggests the following tentative conclusions. The first conclusion is that driving has recovered faster than public transport, and in some cases already reached (or even surpassed) pre‐pandemic traffic and congestion levels. 9 The second conclusion is that the increase in traffic levels is likely to result partly from the substitution of public transport trips. In fact, given that teleworking is still a recommended practice adopted by many employers, the nearly full recovery of driving can only be understood in the presence of modal shifting from public transport to driving.

Although there is no information on which public transport trips are being substituted for, the low(er) patronage levels for the metro operator suggest that these journeys may be particularly affected. Metro journeys are mostly limited to Lisbon municipality (there are only a few stations outside Lisbon) and tend to be shorter than other public transport journeys (Table 4). This means that shorter, urban, trips previously done by metro may now be done using the car or active travel. The observed rise in micromobility systems in Lisbon, including personal and shared e‐bikes and shared e‐scooters (e.g., Félix et al., 2020; Moura et al., 2021), may account for part of the substitution of metro trips (and public transport more generally).

The implementation of temporary cycling infrastructure during the pandemic may also help explain the rise in the use of active travel in the city (e.g., Moura et al., 2021). The best comparable data available consist of manual counting campaigns carried out in May and October since 2017 in the same locations and over the same time periods, summarized in the CML‐Ativos report by Moura et al. (2021), which shows that the number of cyclists observed on the same 19 spots and over the same time periods increased by 87% between 2017 and 2018, 10% between 2018 and 2019, and 30% between 2019 and 2020, whilst decreasing by 5% between 2020 and 2021. However, the data for 2021 refer only to the May campaign, preventing any definite conclusions for that year. Furthermore, the observed reduction is not likely to be representative of total cycling in Lisbon in 2021 because the expansion of the cycling network created alternative routes, increasing cycling on new parts of the cycling and road networks, many of which were not in the original 19 spots (see Figure 1 of Moura et al., 2021). Moreover, some new segments of active travel increased considerably, notably e‐bikers and e‐scooters (Moura et al., 2021). 10 The creation of a financial ‘mobility fund’ in June 2020 by the municipality, in addition to the national mobility fund, to support the acquisition of conventional and electric bicycles by residents, also increased the demand for new bicycles. The increase in bicycle sales registered by sports and/or bike shops may partly result from the public financial support provided by the local and national governments, and is also indicative of a strong increase in cycling in Lisbon. 11 Both trends – that is, increase in bicycle sales and use during the pandemic – have been observed in other cities worldwide in tandem with the provision of temporary cycle lanes (Moreno et al., 2021).

In a nutshell, travel behaviour as we move out of COVID‐19 (or live with it in a vaccinated world), in the more immediate short term, appears to mean a business‐as‐usual scenario in terms of car use. The main reason seems to be that public transport is being substituted by solo driving and active travel. This scenario presents a real concern that congestion and pollution will return to pre‐COVID‐19 levels. Reversing new solo driving habits will be very difficult in the future because, once people have a car, they feel that they must justify the past investment cost even if it implies spending more money than if they opted for an alternative transport mode (i.e., sunk cost fallacy irrational decision). Although car sales fell during the lockdown periods, the effect was asymmetric in that it affected especially the sales (and imports) of new cars, whereas the sales of used cars actually increased as the economy re‐opened. There are still many incentives to drive, particularly through the provision of free or under‐priced parking in the municipality. 12 According to the IMob 2017 travel survey, 41% of resident household cars in Lisbon park for free on the curbside, while 28% pay a heavily subsidized resident parking permit for up to three cars: no cost for the first car, 30€ per annum for the second car and 120€ per annum for the third car. Parking at the workplace is also heavily under‐priced: 50% (27%) of workers report having free parking at their workplace (public space).

5. FUTURE TRAVEL MOBILITIES IN POST‐COVID‐19 VACCINATED LISBON

The COVID‐19 health crisis created two major disruptions with potential long‐lasting effects on travel behaviour. The first disruption refers to the change in working practices; in particular, the mainstreaming of working from home has become an accepted practice in many occupations and industries. The second major disruption refers to the negative shock affecting public transport systems owing to the perceived ‘bio‐security’ risk of virus contamination. These two disruptions have created new challenges for public transport operators and may hinder the transition to a more sustainable urban mobility system.

Homeworking and more flexible working practices may not reduce the demand for travel, especially in the medium term, as individuals and households adjust residential location, with subsequent impacts on commuting distances (even if fewer trips are made) and travel mode choice. The net effect on travel demand, including the effect on congestion and air pollution, will depend on the characteristics of the built environment of the new residential locations, notably the availability and quality of public transport. If households favour larger homes and areas with more greenery, it is possible that teleworking will favour relocation to peripheral areas that are commutable, which in turn can reinforce urban sprawl and car use. This outcome is very much in line with the evidence from the literature on teleworking and travel demand, as discussed in section 2.1.

Even in the shorter term, that is, assuming fixed residential location, people may re‐allocate the saved commuting time and money due to homeworking to other activities that may also require travel. Travel patterns may become more heterogeneous – in both space and time – because on teleworking days people may centre their activities and travel around their homes, whereas on commuting days they may centre their activities around the workplace. One consequence put forward by the literature on teleworking and travel demand, as discussed in section 2.1, is that the extent of trip chaining may decrease with a resulting increase in the number of trips and distances travelled. Likewise with residential relocation, the effects on travel mode choice are difficult to predict and will depend on the characteristics of the built environment of both the residential and workplace locations.

The main message is that working from home is not going to be the panacea for shifting towards more sustainable mobility patterns. In fact, as put forward by the literature on teleworking and travel demand, it may promote an overall increase in distance travelled even if there are fewer commuting trips. The fear of virus contamination in public transport may also contribute to greater car use because travelling in crowded buses or trains has become riskier than before the pandemic. To avoid crowding, public transport operators need to provide higher service frequency and capacity to ensure comfortable levels of social distancing. The fact that patronage remains considerably below pre‐pandemic levels, but a more than proportional increase in operating capacity is required, leads to widening operating deficits and increases the need for subsidies, putting at risk the financial sustainability of public transport. Mainstream vaccination has helped mitigate the negative impact of fear of infection on patronage levels, but it may be insufficient, especially as new virus variants continue to emerge. One long‐lasting consequence of the pandemic may be the change in individuals' preferences from shared transport modes – notably public transport – to personal transport modes, including solo driving but also active micromobility. To avoid going back to pre‐pandemic congested and polluted urban environments, local governments need to incentivize the latter over the former.

To understand how the pandemic is affecting travel behaviour in the present, and consider the possible future trends, we need better monitoring tools. Since there are no regular travel surveys in Lisbon (nor in Portugal), it is practically impossible to measure the effect of public policies on travel demand and use the evidence to guide future policy‐making. Unsurprisingly, a major limitation of this study is thus its ability to measure the effects of the pandemic on travel mobility. It is therefore helpful to consider the existing evidence for other cities. The international evidence from recently published studies using bespoke stated preference and/or revealed preference surveys to study the effects of the COVID‐19 and working from home on travel behaviour seems to be mixed. This is not surprising given that cities differ not only in terms of the characteristics of their built environment and transport systems, but also in relation to the measures adopted by local governments in response to the pandemic. Some authors are optimistic and conclude that working from home can reduce congestion levels (e.g., Hensher et al., 2021), whereas other scholars suggest that future mobility patterns can include a greater use of the private car (e.g., Currie et al., 2021; Schaefer et al., 2021). For example, Currie et al. (2021) conclude that public transport demand will remain at about 80% of its pre‐pandemic level. Given the uncertainty about future outcomes, most studies point to the importance of carrying out ongoing monitoring of individuals' travel behaviour (e.g., Beck et al., 2020; Wang et al., 2021).

6. CONCLUSION

This study provided a discussion of the likely effects of two major COVID‐19‐related disruptions on travel patterns in Lisbon, namely the adoption of widespread working from home and the ‘bio‐security’ fear affecting public transport. Drawing on evidence from the literature on teleworking and travel demand and the determinants of public transport, together with a descriptive empirical analysis of travel mobility data before and during the pandemic, the study concludes that the pandemic may hinder, more than aid, the transition to a sustainable urban mobility paradigm in Lisbon and its metropolitan region.

Concerning working practices, although full‐week working from home is most likely not viable in the future, weekly working from home and related flexible working practices (e.g., staggered working hours) are likely to stay common practice in post‐COVID‐19 economies. In the short term, the increase in household income budget and time budget can be used to engage in new activities which may or may not involve travel. Furthermore, the perceived health‐related ‘bio‐security’ fear of virus contamination, even in extensively vaccinated economies, may contribute further to the decline of public transport use in the short term. In the longer term, households may adjust their residential location, by relocating to more peripheral areas further away from their workplaces. Since people tend to prefer more space to less space, and space is more costly in urban areas, the increased disposable income and reduced commuting frequency can facilitate residential relocation outside urban areas and city centres where the quality of public transport is weaker. Consequently, a possible outcome may be a permanent rise in car use, congestion and its associated health and environmental negative externalities.

The empirical analysis of monthly data for travel demand by transport modes in Lisbon showed that car use recovered at a faster rate than public transport demand, reaching pre‐pandemic levels in parts of the road network. This increase in car use is likely to result partly from a substitution of public transport trips by solo driving, although it is not possible to test this hypothesis. Among public transport modes, the recovery has been more sluggish for Lisbon's metro system. Since metro trips tend to be shorter (i.e., the mean travelled distance is 9 km) and are essentially limited to the city of Lisbon, they may have been replaced not only by driving but also by active travel, namely micromobility services such as shared e‐bikes and e‐scooters. The main takeaway is that we seem to be observing a growing polarization of travel mobility, consisting of a rise in driving and active travel and a commensurate decline in public transport. Public transport operators are facing unforeseen challenges. The pandemic increased operating deficits and the need for additional public subsidies, raising concerns about the financial sustainability of public transport in the future.

Neither technological innovation (e.g., EVs, MaaS) nor social innovation (e.g., working from home and other flexible working regimes) will be sufficient to change the current car‐based mobility paradigm. Innovative proximity‐based urban planning solutions implemented during the pandemic (e.g., temporary cycle lanes, provision of proximity‐based services) appear to have had some success in permitting people to travel while abiding by health and safety protocols and contributing to sustainability goals by avoiding car use (Moreno et al., 2021). Urban planning policies need to be re‐defined in ways that favour public transport and active travel. This can be achieved by giving more space to these modes and less space to cars (including parking), planning for denser, more mixed land uses, and for proximity‐based services. In addition to rethinking urban planning, local governments also need to eliminate existing economic incentives to drive, notably the under‐pricing of car parking in urban areas. Without serious commitment by local and national governments, the pandemic is likely to be another missed opportunity to transition to a better, more sustainable, mobility system.

APPENDIX A.

Severity of restrictions on travel mobility and social meetings.

Mar 2020	1st lockdown
Apr 2020
May 2020
Jun 2020
Jul 2020
Aug 2020
Sep 2020
Oct 2020
Nov 2020	2nd lockdown
Dec 2020
Jan 2021
Feb 2021
Mar 2021
Apr 2021
May 2021
Jun 2021
Jul 2021
Aug 2021
Sep 2021
Oct 2021
Nov 2021
Dec 2021

Legend

Very high (state of emergency)

High (state of emergency with phased re‐opening plan)

Medium (calamity status)

Low or very low (contingency or alert status)

Notes:

• The announcement and renewal of COVID‐19 status and restrictions was implemented on a fortnight or weekly basis, so the monthly summary is only indicative and may not reflect accurately changes in status conditions within the same month.
• Temporary travel mobility restrictions were implemented between Christmas and New Year.

APPENDIX B.

FIGURE B1.

Lisbon metropolitan area and its 18 municipalities. Source: Inquérito à Mobilidade nas Áreas Metropolitanas do Porto e de Lisboa 2017 (INE, 2018)

APPENDIX C.

TABLE C1.

Monthly passengers by public transport operator

	CARRIS (103)				Lisbon metro (103)				Ferry boat
	2018	2019	2020	2021	2018	2019	2020	2021	2018	2019	2020	2021
Jan	10,480	10,660	12,300	5,320	13,377	14,165	16,072	4,825	1,437,378	1,532,408	1,663,046	641,349
Feb	9,650	10,410	11,850	3,750	12,483	13,691	15,223	2,900	1,300,320	1,476,127	1,599,214	451,498
Mar	10,800	11,460	5,680	5,210	14,520	15,096	8,565	3,928	1,425,691	1,677,956	850,814	609,266
April	10,560	11,280	30	6,530	14,181	15,076	2,662	5,423	1,387,130	1,562,287	306	739,206
May	11,540	13,020	3,510	7,910	15,372	16,794	3,300	7,020	1,576,633	1,705,146	531,267
Jun	10,480	10,880	5,210	8,320	13,790	14,550	4,800	6,823	1,467,059	1,511,525	741,747
Jul	10,350	11,940	6,170	7,820	13,575	15,466	6,096	6,840	1,514,957	1,620,525	893,479
Aug	8,790	10,100	6,000	7,230	11,959	12,958	6,080	6,719	1,434,028	1,573,942	914,023
Sep	10,510	12,180	7,530	8,990	14,803	15,487	7,265	8,635	1,382,619	1,661,275	1,021,823
Oct	11,880	13,880	7,930		16,520	18,220	7,852		1,675,784	1,823,372	1,022,860
Nov	10,980	12,350	6,900		14,882	15,874	6,531		1,552,962	1,642,946	804,320
Dec	9,650	11,330	6,180		13,706	15,698	6,113		1,481,629	1,575,705	765,108

	Train: Fertagus (103)				Train: CP (103)
	2018	2019	2020	2021	2018	2019	2020	2021
Jan	1,803	1,938	2,521	961	10,968	11,634	14,469	7,429
Feb	1,763	1,926	2,224	655	9,921	10,734	13,533	4,350
Mar	1,761	1,846	1,369	891	10,629	10,966	7,559	5,726
April	1,849	1,884	304	1,151	11,017	12,966	1,254	7,582
May	1,860	2,246	685	1,433	11,688	14,015	4,352	9,240
Jun	1,710	1,861	975	1,329	10,672	12,528	6,109	8,881
Jul	1,506	2,160	1,127	1,339	10,942	14,291	8,228	9,566
Aug	1,315	1,764	1,080	1,197	9,766	13,234	7,703	9,241
Sep	1,933	2,176	1,360	1,566	11,528	15,221	9,242
Oct	1,988	2,690	1,489		12,115	17,128	9,676
Nov	1,958	2,444	1,279		11,371	14,644	8,543
Dec	1,686	2,261	1,151		10,660	14,284	7,896

TABLE C2.

Monthly mean daily traffic counts

	Vasco da Gama bridge (A12)				25 de Abril bridge (A2)				VFX‐Sacavem (A1)				V. Duarte Pacheco – Cascais (A5)
	2018	2019	2020	2021	2018	2019	2020	2021	2018	2019	2020	2021	2018	2019	2020	2021
Jan	58,835	62,385	64,652	37,254	132,548	129,988	130,797	83,488	79,537	80,731	81,915	52,732	86,962	89,291	90,472	52,824
Feb	59,677	65,535	65,282	34,055	135,770	138,800	135,853	79,551	81,534	83,820	84,070	51,333	87,591	92,975	91,283	46,965
Mar	60,350	66,862	42,031	42,080	133,048	138,324	91,849	92,315	81,379	85,472	57,255	61,345	87,291	92,480	56,024	58,338
April	62,350	65,522	21,707	50,947	139,857	135,896	49,846	116,335	83,689	84,781	33,061	71,030	89,000	89,587	27,547	72,362
May	66,164	71,190	37,227	62,282	149,030	146,212	94,767	140,403	86,635	89,078	53,882	81,949	95,258	98,456	50,289	85,169
Jun	67,374	69,723	53,610	62,380	149,236	143,386	122,002	137,139	84,732	87,323	69,042	78,796	93,624	92,305	67,749	83,129
Jul	71,822	75,345	63,043	65,329	157,399	153,255	141,708	142,321	85,910	89,859	76,800	80,503	94,030	96,247	77,521	82,087
Aug	70,967	71,936	64,430	67,407	156,684	148,666	142,774	147,795	85,036	86,526	78,348	84,538	85,411	82,723	70,210	76,496
Sep	70,774	72,832	64,054	68,817	154,719	148,194	141,850	146,577	87,773	88,452	80,020	87,142	95,305	95,716	82,930	91,131
Oct	66,952	68,901	58,837		142,590	140,492	127,660		85,361	87,790	75,712		93,589	95,997	78,801
Nov	63,785	65,641	47,286		131,780	134,224	105,699		82,172	85,265	64,327		90,134	91,298	67,741
Dec	64,771	67,789	49,998		134,645	134,094	104,308		86,308	86,897	66,299		90,518	91,372	68,862

	CRL‐CREL (A8)				CRIL – V. Duarte Pacheco (IP7)				IC17
	2018	2019	2020	2021	2018	2019	2020	2021	2018	2019	2020	2021
Jan	61,843	63,688	65,067	40,803	75,339	74,359	75,256	48,324	71,721	65,179	68,044	49,273
Feb	63,353	65,763	66,588	38,499	75,737	78,198	76,814	38,625	72,437	67,589	68,689	42,651
Mar	62,511	66,252	43,078	45,836	74,804	75,875	47,550	51,412	71,414	67,418	43,990	49,762
April	64,039	64,245	25,327	53,926	77,748	75,602	28,946	65,187	73,053	66,831	23,571	58,007
May	66,781	69,478	40,277	62,510	82,229	82,777	49,688	71,126	76,806	70,834	46,386	63,593
Jun	66,080	66,194	52,169	59,677	80,516	78,978	63,676	69,604	75,943	68,247	56,321	60,262
Jul	67,757	70,432	58,931	60,491	81,909	80,461	67,265	69,919	77,767	70,148	56,385	58,238
Aug	62,988	63,121	56,717	59,859	76,803	75,111	60,398	68,000	72,447	64,984	53,576	57,731
Sep	67,405	68,783	62,371	66,384	81,932	79,943	71,263	75,146	76,380	67,852	62,885	61,553
Oct	67,142	69,658	59,626		78,414	82,636	71,769		74,818	71,747	63,383
Nov	65,083	66,470	51,398		75,801	77,659	61,590		72,207	69,445	57,270
Dec	65,916	67,199	51,671		73,539	76,768	57,889		68,248	69,911	57,082

Melo, P. C. (2022). Will COVID‐19 hinder or aid the transition to sustainable urban mobility? Spotlight on Portugal's largest urban agglomeration. Regional Science Policy & Practice, 1–27. 10.1111/rsp3.12518