The benefits of bicycling through the fields
Calculating the effects of an increased bicycle commuting scenario in a Swedish rural setting on transport related societal costs
Emiel Driessen
Bachelor thesis – Individual assignment Main field of study: Environmental science Credits: 15.0
Semester/year: Spring semester 2020 Supervisor: Jenny Zimmerman Examiner: Erik Grönlund Course code: MX045G
Degree program: Ecotechnology
Abstract
Bicycling as a form of active transport has been gaining in popularity amongst policy makers and urban planners. In many cities, it is seen as a beneficial and efficient transport alternative to the congesting and air polluting car. Besides that, bicycling has also shown to reduce greenhouse gas emissions, noise and come with a range of health benefits when done regularly. Despite studies showing the individual and societal benefits of bicycling in an urban context, there is a lack of literature on how these benefits translate to a rural setting, characterised by longer distances and lower volumes of travellers. A commuting route between Östersund and Krokom was taken to study this. A cost-benefits analysis was performed for this route in which the transport related societal costs were calculated for a hypothetical increased bicycling scenario on this route and compared to the costs of a status quo scenario with no bicycling. Results show significant bicycle related costs such as travel time, and accidents, but also benefits compared to the car such as health and low operating costs. The environmental benefits are relatively small. Due to the results of the study covering a wide range, no conclusion could be drawn on if the studied rural bicycling scenario would be desirable or not.
Conservative estimates indicate extra societal costs, while bicycle favouring estimates indicate cost savings.
Keywords: bicycling, cost-benefit analysis, rural, Sweden, environment, society
Table of contents
1. Introduction ... 1
1.1 Bicycling as a solution ... 1
1.3 Aim and research question... 3
2. Method ... 3
2.1 The methods, delimitations & unit ... 3
2.1.1 Cost-Benefit Analysis (CBA) ... 3
2.1.2 Health Economic Assessment Tool (HEAT) ... 4
2.1.3 Delimitations ... 4
2.1.4 Unit ... 4
2.2 The Krokom-Östersund commuting scenario ... 4
2.2.1 Geography and demography ... 5
2.2.2 The planned bicycling infrastructure ... 5
2.3 Data collection ... 5
2.3.1 Data hierarchy and main sources ... 5
2.3.2 Communication with Simon Östberg ... 6
2.3.3 Mode share ... 6
2.3.4 Travel distance and speed... 6
2.3.5 Value range ... 6
2.4 Impact categories ... 6
2.4.1 Travel time costs and congestion costs ... 6
2.4.2 Vehicle operating costs ... 7
2.4.3 Health benefits ... 7
2.4.4 Accident cost ... 7
2.4.5 Climate change ... 7
2.4.6 Air pollution ... 8
2.4.7 Noise ... 8
2.4.8 Winter maintenance... 8
2.4.9 Construction costs ... 8
2.5 Data analysis ... 8
3. Results ... 12
3.1 Supporting data ... 12
3.2 Costs per impact category ... 12
3.2.1 Travel time costs and congestion costs ... 12
3.2.2 Vehicle operating costs ... 13
3.2.3 Health benefits ... 13
3.2.4 Accident costs ... 14
3.2.5 Climate change ... 14
3.2.6 Air pollution ... 14
3.2.7 Noise ... 15
3.2.8 Winter maintenance... 15
3.2.9 Construction costs ... 15
3.3 Costs for driving and bicycling ... 15
3.4 Marginal cost per impact category ... 16
3.5 Changing the bicycling mode share ... 17
4. Discussion ... 18
4.1 The major impact categories ... 19
4.1.1 Travel time and congestion costs ... 19
4.1.2 Vehicle operation costs ... 19
4.1.3 Accident costs ... 20
4.1.4 Health ... 21
4.2 The influence of the rural context ... 21
4.3 CBA as an indicator for social and environmental sustainability ... 22
4.4 What was not included ... 23
4.5 Conclusion ... 23
4.6 Future research ... 24
Reference list ... 25
Appendix A: Existing and planned bicycle paths ... 28
Appendix B: Input values for HEAT ... 29
Appendix C: CO2 emission calculations for increased bicycling scenario ... 32
1
1. Introduction
Bicycling has been a popular mode of transport for over a century, but after the Second World War, the number of automobiles around the world increased. New infrastructure was built to support this change which has led to the expansive road networks that we see all around the world today. This increase in car-use has coincided with a decrease in bicycling in many countries (Kurnik, 2018). Bicycling in car-centered environments is often perceived as stressful and dangerous, but a new interest in creating more bicycle friendly public spaces, especially in urban areas, has emerged. This is due to the fact that increasing the rate of bicycling is seen as a solution to problems such as congestion, air, soil, water and noise pollution and greenhouse gas emissions (GHG), all in which cars play a part in creating (de Nazelle et al., 2010; Department for Transport London, 2011; Maibach et al., 2008).
Congestion is estimated to cost around 1% of the total GDP of the EU. In Sweden, the country of interest for this study, this is similar with 0,9%, costing the country 2,6 billion Euros a year (Christidis & Rivas, 2012).
It has also shown that congestion intensifies air, water and soil pollution from cars. These forms of pollution can have significant effects on the health of both people and the environment. It can cause issues such as respiratory infections, asthma and cardiovascular disease, while it can also get pollutants such as carcinogens and heavy metals into drinking water, waterways, food crops and other plants (Goonetilleke, Wijesiri, &
Bandala, 2017; Meyer & Elrahman, 2019). One study estimated the amount of worldwide premature deaths linked to vehicle emission of particle matter to be 385.000 in 2015 (Anenberg et al., 2019). Noise pollution has shown to cause hypertension, coronary heart disease and sleeping problems (Li, Qiao, & Yu, 2016). In the EU, at least 100 million people are experiencing levels of noise that fall outside of the range of comfort set by the EU. The main cause of this noise pollution is road traffic (European Environment Agency, 2017).
The transportation sector has been the only sector in the EU to increase its emissions since 1990. It is responsible for 30% of the total GHG emissions in the EU. Of these emissions, 60,7% comes from cars, making them the major polluter in the transport sector. With an average of 1,7 people per car in Europe, the car is often not used efficiently which is one of the reasons for this high share in emissions (European Parliament, 2019).
1.1 Bicycling as a solution
This wide range of negative impacts influences the well-being of the individual, the environment and society as a whole. It is thus not surprising that policy makers are looking towards alternatives such as bicycling to reduce these impacts (Kallas, 2011). Bicycling can address environmental, health and transportation benefits.
Research shows that bicycling can reduce greenhouse gas emissions as well as air pollution such as NOx
2
particle (Blondel, Mispelon, & Ferguson, 2011; Frank et al., 2010; Johansson et al., 2017; Mason, Fulton, &
McDonald, 2015). A study done in Stockholm estimated that there is a potential for 111.000 new bicycle commuters in the city, which would reduce the exposure to NOx by 7%, saving 449 years of life annually.
This is twice as effective as the congestion tax is estimated to be (Johansson et al., 2017). When it comes to GHG emission reductions, a scenario with a high share of bicycling (23%) would be responsible for a 11%
reduction for urban transport by the year 2050, compared to a scenario with no focus on high amounts of bicycling (Mason et al., 2015). There are also significant health benefits that come with changing from driving a car to riding a bicycle (Chertok et al., 2004; De Hartog et al., 2010; Johansson et al., 2017; Macmillan et al., 2014). In the Netherlands, a country in which about 27% of all trips are done by bicycle, an estimated 6500 deaths are prevented each year and residents live around half a year longer due to the positive health effects of bicycling. These benefits are said to account for more than 3% of the total GDP of the Netherlands (Fishman, Schepers, & Kamphuis, 2015). Bicycling has also shown to reduce congestion during rush hour (Hamilton & Wichman, 2018; Wang & Zhou, 2017).
While planning large transportation projects such as a road construction, it is common to do an analysis on what range of benefits this new piece of infrastructure will bring to its users (e.g., time savings) and compare them to the different costs (Van Wee & Rietveld, 2013). This is called a Cost-Benefit Analyis (CBA). The information retrieved from a CBA can aid policymakers in deciding the desirability of a particular project (Nas, 2006). When it comes to bicycling and bicycling infrastructure, these CBA’s are less common. The most common version of studies found that add up the different benefits and costs of bicycling are done on a city, regional or country level. These studies have looked at how the societal costs of bicycling compared to the costs of car-use in Copenhagen, Calgary and the European Union (COWI and City of Copenhagen, 2009;
Dekker, 2016; Gössling et al., 2019; Gössling & Choi, 2015). Other studies focus on one or a few aspects such as the health benefits of active transport (i.e., bicycling and walking) compared to car-use (Pérez et al., 2017;
Tainio et al., 2016).
There has also been some critique on having too much of a focus on the benefits that are not directly related to transport (e.g., health and GHG emissions). The authors of this critical study claim that this implies a view of bicycling as a way of getting people out of cars, instead of seeing bicycling as a legitimate form of transport by itself (Börjesson & Eliasson, 2012).
Most studies done on bicycling benefits are done in an urban context where distances are short and congestion caused by cars is the largest problems. For this study, no articles could be found that confirm if these bicycling benefits are still applicable in a rural context with longer distances and a lower population density. They also tend to look at the large scale combined societal benefits (e.g., city-level) instead of looking
3
at the benefits of single piece of infrastructure (e.g., a bicycle path), of which only one study could be found (COWI and City of Copenhagen, 2009).
1.3 Aim and research question
The aim of this research will be to look at a hypothetical bicycling scenario on a local commuting route and in a rural context to determine the benefits and costs associated with this scenario. The aim of this study is to create a better insight into the desirability of bicycling in a non-urban context by looking at a hypothetical bicycling scenario on a local and rural commuting route and calculate the benefits and costs associated with this scenario. Having this information could aid policymakers in deciding on investments in bicycling infrastructure in this specific case, or in similar cases in other Swedish municipalities. It could also help other actors push for certain policies or investments that would increase bicycling.
The rural route in question connects the urban areas of Krokom and Östersund. The current absence of a complete bicycle connection and the relatively long distance makes this an interesting case study to help answer the following research question:
- In what way does increased bicycling in a rural setting affect the transport related societal costs?
This study will try to answer this question by applying a CBA framework on this case study and then to organise, analyse and discuss the results that were calculated from applying this method.
2. Method
2.1 The methods, delimitations & unit
2.1.1 Cost-Benefit Analysis (CBA)
The goal of this study was to find to calculate the potential social benefits, which also includes environmental benefits, of a bicycle connection between Östersund and Krokom. The best suited method for achieving this goal was assumed to be a Cost-Benefit Analysis (CBA), since it is widely used for assessing the social benefits of public projects (Boardman et al., 2017; Hanley & Splash, 1993). It does so by converting all the societal gains (benefits) and losses (costs) of a project, including many externalities, into a monetary value and add them together. This will give a more complete picture of the true negative or positive effects the project might have on society and does so in a way that can aid policy makers in making decisions regarding. the desirability of a project (Nas, 2006). The same was done for this research.
A specific CBA framework for the appraisal of cycling was used in this study. This framework was developed in Copenhagen and takes into account the different impact categories that were used in this study too (COWI
4
and City of Copenhagen, 2009). This was combined with a more recent case study done on the societal benefits of cycling in Calgary (Dekker, 2016). Combining these two studies, the following impact categories were used: vehicle operating costs, travel time costs, health benefits, accident costs, air pollution, climate change, noise, congestion and winter maintenance. These are the same impact categories that are used in the Dekker study (2016).
2.1.2 Health Economic Assessment Tool (HEAT)
HEAT is a tool that was used within the CBA framework to help calculate the value of some of the impact categories. It was created by the World Health Organization (WHO) to help interested parties calculate the health benefits of bicycling. The tool combined user input with information from a database to calculate the reduced mortality that a bicycling scenario could bring (Kahlmeier et al., 2017). This is displayed in premature death avoided and in euros saved. It was also used for calculating the savings on air pollution and CO2
emissions. Air pollution results are displayed in lives and money saved while the CO2 emissions were displayed in tons of emissions saved and money saved (THE PEP, 2019).
2.1.3 Delimitations
To simplify conducting this CBA, only commuters between the Krokom area and Östersund and the other way around were included. This group is likely to be the biggest user of the bicycle infrastructure in this scenario. They are also the most consistent which somewhat simplified the calculations. Lastly, data on this group was available, making the results more realistic and reliable.
To simplify the calculation and reduce the amount of assumptions, no long term picture was created. The starting point was a comparison between the current scenario (reference scenario) and one in which the infrastructure was instantly created and the benefits and costs instantly at full effect (comparison scenario). All the scenarios were created as if they were taking place over the same year
The area in which the commuting takes places consists of the city of Östersund in the municipality of Östersund, and the areas of Rödön, Ås and Åspås. This area was chosen because the retrieved commuting data also covered this area (S. Östberg, personal communication, March 23, 2020).
2.1.4 Unit
To be able to compare and add together all the different impact categories, one universal unit had to be used.
For a CBA, this is often the cost per kilometre travelled by a certain mode of transport. This unit (i.e., SEK/km) was also used for this study.
2.2 The Krokom-Östersund commuting scenario
For this research, the CBA methodology for bicycling was used to calculate the benefits and costs associated with bicycling infrastructure between Östersund and Krokom
5 2.2.1 Geography and demography
Krokom and Östersund are two municipalities in Jämtland, a county located in the centre of Sweden. Krokom lies to the North of Östersund. With about 50.000 inhabitants, Östersund is the largest city in the region.
Krokom is more sparsely populated, with about 15.000 inhabitants in the entire municipality. Most people live in the Southern part of the municipality, which is closer to Östersund. The South is also where the biggest town in the municipality is located. It is also named Krokom and has a bit more than two thousand inhabitants. Other towns near Krokom and Östersund are Ås, Nälden, Dvärsätt and Aspå, having a combined population of almost 2800 (Krokoms Kommun, 2014). Both Krokom and Östersund are municipalities that are growing in population size and between which a significant amount of commuting takes place (S. Östberg, personal communication, March 6, 2020; S. Östberg, personal communication, March 23, 2020).
2.2.2 The planned bicycling infrastructure
Because of the population growth in municipalities, new housing and infrastructure to support the new inhabitants is being planned. Although not formalized yet, this has also led to the planning of a new bicycling route between Ås and Dvärsätt, completing the connection between Krokom and Östersund (S. Östberg, personal communication, March 23, 2020). This is supposed to be a bicycle highway, allowing for higher average speeds due wider paths and fewer interruptions such as traffic lights. This is mainly aimed at increasing the convenience for commuters using the path (Interreg North-West Europe, 2020; Löwing, Koucky, & Kleberg, 2012). The planned route can be seen in Appendix A.
2.3 Data collection
2.3.1 Data hierarchy and main sources
The data needed for this study was collected from various sources. As a general rule, local data was preferred over general data. If data was not found on a regional or municipality level, then Swedish data was used. If this wasn’t available, data from other Nordic countries or Europe was used.
The main local source for this research came from correspondence with Simon Östberg (S. Östberg, personal communication, March 23, 2020; S. Östberg, personal communication, March 6, 2020) in combination with a geographic information system, in this case Google Maps (Google, n.d.). The main source on a national level is the Swedish report that includes many of the values that can be used for CBA’s of transport projects (Bångman & Nordlöf, 2018). It does often lack data on bicycling, which was sometimes supplemented with data from Copenhagen, which has more information specifically for bicycling (Transport DTU & COWI, 2019). Lastly, information from the EU was also used to fill in the data gaps where necessary. This is mostly statistical data.
6 2.3.2 Communication with Simon Östberg
An interview with Simon Östberg was conducted. He works at the municipality of Krokom as a
“planarkitekt” (spatial planner). He was able to share plans that the municipality has when it comes to building a bicycle path in the area of interest for this case study. He could share the length of the path and how much it would cost to construct. He also shared some local commuter data as well as the cost of winter maintenance.
2.3.3 Mode share
There is little useful regional data on the mode shares of commuters or even the general population.
Therefore, the used mode shares had to be estimated based on other data of Sweden. Commuting data on public transport was based on a report made on North Sweden by Trafikanalys (Trafikanalys, 2013). The bicycling mode share when a bicycle path would be put in place cannot be predicted in a reliable way without any survey. All the commuting that is not done by public transport or bicycle was assumed to be done by car.
2.3.4 Travel distance and speed
The average travel distance was estimated by looking at the distances from the different towns that cover the commuting area to Östersund, while also keeping in mind the amount of people living in these areas. The same distance for bicycling, driving and taking public transport was used to simplify the calculations. The travel speeds were also based on the travel time from Krokom to Östersund by all three forms of transport used in this study (Google, n.d.). For bicycling it was estimated slightly higher based on the average speed on bicycle highways (Interreg North-West Europe, 2020).
2.3.5 Value range
To account for the uncertainties or discrepancy in data, the results were presented in a range. This means that when different valid sources were found that gave different impact category values, these different values were represented in a range from a conservative estimate to a high estimate instead of using one value only.
2.4 Impact categories
In the following part, every impact category is explained. The collection and conversion of data for this category is elaborated upon further in figures 2.1, 2.2 and 2.3.
2.4.1 Travel time costs and congestion costs
In the appraisal of infrastructural projects, travel time costs are often one of the most influential factors driving societal benefits. The travel time cost describes the willingness to pay to save a certain amount of time by the traveller (Athira et al., 2016). The amount that people are willing to pay per unit of time can change with the mode of travel, but also with the circumstances of the travel or the purpose of the journey (Bångman
& Nordlöf, 2018). The more people are willing to pay to save time, the more valuable their time is in that situation, indicating unfavourable conditions of travel.
7
Congestion can be seen as a form of travel time costs. It is the time spent stuck in traffic and how much people are willing to pay to save this time. It is made up of two types: Time stuck in congestion and the time lost due to the congestion which both have their own values (Bångman & Nordlöf, 2018).
2.4.2 Vehicle operating costs
The operating cost is another factor that can vary between different modes of transport. For the bicycle these costs are represented in the Swedish report (Bångman & Nordlöf, 2018). The operating costs of a car are more complex due to the many factors that play a role as well as taxes. There is also no average cost to be found in the Swedish report. In the Danish database there is an average number, but this might be different than a Swedish one due to different taxes on fuel and car-use (Transport DTU & COWI, 2019). Using this Danish value and several Swedish websites that have tried to estimate the cost of owning a car, a range for the operating costs for car-use was created.
2.4.3 Health benefits
The health benefits were calculated using HEAT , created by the World Health (THE PEP, 2019). Using this tool has simplified the process of doing these calculations. The most important input needed was the current mode share for bicycling and car use and that same mode share after the bicycling infrastructure is put into use. Also the average length of the trip and the population that can possibly be affected by this new infrastructure were inputs needed. Many of the other values are standardized and compiled from research by the creators of this tool (Kahlmeier et al., 2017). When necessary and possible, the standards were adjusted.
All the input necessary can be found in more detail in Appendix B.
The output from HEAT is in Euros saved per year, this has to be exchanged to Swedish Crowns and then be divided by the total amount of kilometre bicycled on a yearly basis to get the health benefits of bicycling per kilometre travelled.
2.4.4 Accident cost
The accident costs include both the material and health costs that are related to accidents and collisions (Bångman & Nordlöf, 2018).
2.4.5 Climate change
In this study, the benefits of reduction in CO2 due to more bicycling were also taken into account. These benefits were calculated in two ways. The first way was by using local data on the average CO2 emissions per kilometre of a car and using the Swedish carbon tax price in combination with the avoided car travel to calculate the benefits for the bicycling scenario (Government Offices of Sweden, 2020; Miljöbarometern, 2019). The precise calculations done can be found in Appendix C. The other way is by using HEAT to do the calculation (THE PEP, 2019). The input used can be found in Appendix B.
8 2.4.6 Air pollution
In the same way as was done for CO2 emissions, air pollution was calculated with two methods. The first method used the values from the Swedish report, using values from both rural and urban environments and combining them (Bångman & Nordlöf, 2018).
The second method used HEAT to calculate the costs of air pollution. The specific input can be found in Appendix B. As input value of air quality, a concentration of particulate matter smaller than 2.5 micrometer (PM2.5) was used (SMHI, 2018). The output was given in money saved per year by bicycling, which then had to be translated to a cost for car use by using yearly kilometres travelled for both forms of transport.
2.4.7 Noise
The noise pollution of bicycling is deemed negligible while the cost of noise for an average car can be found in the Swedish report (Bångman & Nordlöf, 2018). The same as with air pollution, values of noise in both rural and urban environments were used, based on what part of the journey took place in those environments.
2.4.8 Winter maintenance
The winter maintenance costs were provided by Simon Östberg from the municipality of Krokom and involve snow removal as well as anti-skid measures (S. Östberg, personal communication, March 23, 2020).
2.4.9 Construction costs
The bicycle path construction costs were provided in Swedish Crowns per square meter by Simon Östberg from the municipality of Krokom (S. Östberg, personal communication, March 23, 2020).
2.5 Data analysis
All the data collected represents values in different units and needs to be converted to Swedish Crowns per travelled kilometre (SEK/km) and Swedish Crowns per year (SEK/year).
In the following figures, it can be seen how the calculations to get to the impact category values were done and what sources the data was retrieved from. Figure 2.1 shows the calculations needed for retrieving commuting data on trips per year and distance commuted per year. Figure 2.1 also shows how the area of the bicycle path was calculated.
These numbers were needed to calculate the values of the different impact categories for the car and the bicycle and to calculate the total yearly transportation costs in the reference case and the comparison scenario.
These calculations and the sources used can be seen in figure 2.2 for the reference case and in figure 2.3 for the comparison scenario.
9 Figure 2.1
Calculations and sources needed to get the total amount of commuting trips done per year, total distance commuted per year for all commuters or by mode of transport and the area of the bicycle path
10 Figure 2.2
Calculations and sources needed to get the values (in SEK/km) of every car-related impact category and to get the total societal costs of the reference scenario (in SEK).
11 Figure 2.3
Calculations and sources needed to get the values (in SEK/km) of every bicycle-related impact category and to get the total societal costs of the comparison scenario (in SEK).
12
3. Results
Firstly, all the impact category data and supporting data will be presented, after which they will be combined and analysed.
3.1 Supporting data
In table 3.1, the values needed for performing the CBA are presented.
Table 3.1
Supporting variables used for calculating impact category values represented in value, unit and reference
Category Variable Value Unit Reference
Bicycle path
Length 7,8 kilometres S. Östberg, 2020
Width 3 metres S. Östberg, 2020
Area 23.400 square metres Figure 2.1
Life span 20 years (Transport DTU & COWI, 2019)
The commuters
From Krokom to Östersund 2.297 commuters S. Östberg, 2020 From Östersund to Krokom 1.169 commuters S. Östberg, 2020
Total 3.466 commuters Figure 2.1
The commute
Distance 20 kilometres (Google, n.d.)
Working days a year 220 days (European Commission, 2017)
Average trips per day 4.178 trips/day Figure 2.1 Trips per year 1.525.040 trips/year Figure 2.1
Speed (car - publ. tr. - bicycle) 60 - 40 - 20 km/h (Google, n.d.; Interreg North-West Europe, 2020)
Urban vs. rural 20 - 80 Per cent (Google, n.d.) The mode shares
(car - public transport - bicycle)
Reference scenario 92 - 8 – 0 Per cent (Trafikanalys, 2013)
Increase bicycling scenario 87 - 8 – 5 Per cent (Trafikanalys, 2013)
3.2 Costs per impact category
3.2.1 Travel time costs and congestion costs
According to the guidelines from the Swedish Transport Administration, the cost assigned to bicycling on a separated bicycle path is 129 SEK/h (Bångman & Nordlöf, 2018). This equals to 6,45 SEK/km if you divide it by the average speed of 20 km/h.
13
For a car trip, which is generally assumed to be more comfortable, travellers were willing to pay 93 SEK/h (Bångman & Nordlöf, 2018; Björklund & Isacsson, 2013). This would equal to 1,55 SEK/km if you simply divide it by the average car speed of 60 km/h. But this value changes when time spent in congested traffic is also considered. The hourly value goes up to 141 SEK and the cost of delay caused by congestion is estimated at 327 SEK/h (Bångman & Nordlöf, 2018). Knowing that the average Swede spends 21,56 hours in congestion each year and that every hour spent in congestion for this case study is estimated at 10 minutes of delay, the total time cost for a commuter can be estimated at 15.343 SEK/year (European Commission, 2017;
TomTom International BV, 2013). This is equal to 1,74 SEK/km.
3.2.2 Vehicle operating costs
For the bicycle, the operating costs are estimated to be 0.68 SEK/km, including the initial cost of the bicycle itself, insurance, repairing and maintenance (Bångman & Nordlöf, 2018).
The Transport Ministry of Denmark has released a data set in which the operating costs, including taxes, is estimated to be 2,98 Danish Crowns (DKK) per kilometre which would be equal to 4,36 SEK when taking an exchange rate from DKK to SEK of 1,46 as of April the 16th (Boliga Finans ApS, 2020; Transport DTU &
COWI, 2019). This was combined with data from the following online articles, as given by Tom Petersen from Trafikanalys (T. Petersen, personal communication, March 16, 2020). In table 4.2, these numbers are shown. They represent a range from 4,25 SEK/km to 6,65 EK/km.
Table 3.2
Values and sources of the different costs per kilometre used for calculating operating costs
Value (SEK/km) Explanation Source
4,25 Volvo V70 Diesel (Eklund, 2018)
4,36 Based on Danish situation (Transport DTU & COWI, 2019)
5,04 Average of the 3 price classes (Swedbank, 2012)
5,65 Based on 240.000 SEK new car (Dagens PS, 2014)
6,65 Based on 315.000 SEK new car (diesel and gasoline)
(Tornvall, 2018)
3.2.3 Health benefits
The health benefits that were calculated using HEAT represent a range, mainly because of the change in amount of physical activity that is expected to be substituted for the commuting by bicycle. If there is a lot of physical activity being substituted, as is the case in the conservative estimate (40%), than the health benefits will be lower (Kahlmeier et al., 2017). For the conservative estimate, the savings were calculated at 365.000
14
EUR/year or 3973031 SEK/year (Boliga Finans ApS, 2020; THE PEP, 2019). Using the total amount of bicycle kilometres commuted of 1.525.040 km/year, this is calculated to be a benefit of 2,61 SEK/km. For the high estimate (0% substitution) the benefits are valued at 609.000 EUR/year or 6.628.975 SEK/year, which equates to 4,34 SEK/km (Boliga Finans ApS, 2020). These benefits can also be seen as negative costs.
3.2.4 Accident costs
The Swedish report with standard values for CBA’s estimates the accident cost to be 4.5 SEK/km for the bike. This number was calculated by using the average of 1.5 accidents per million km ridden and the average cost of an accident being 3,000,000 SEK and was taken as the conservative estimate (Bångman & Nordlöf, 2018). For the high estimate, a value on accident cost for bicycling from the Danish database was used. This is set at 1,68 SEK/km (Transport DTU & COWI, 2019).
The marginal accident cost of driving a car was estimated to be lower at 0.16 SEK/km (Bångman & Nordlöf, 2018).
3.2.5 Climate change
The climate change benefits of bicycling differ depending on the method used to calculate them. For the conservative side of the range, the value of 1190 Swedish Crowns per tonne or 1,19 Swedish Crowns per kilogram of CO2 emitted was used (Government Offices of Sweden, 2020). The average emissions of a car in Krokom (174 g CO2/km) and in Östersund (168 g CO2/km) were used in combination with the amount of commuters from each area to come to an average emission of 172 g CO2/km as can be seen in Appendix C (Miljöbarometern, 2019). This 0,172 kg of CO2 was then multiplied by the 1,19 SEK/kg to get a saving of 0,2 SEK/km of commuting by bicycle.
For the high estimate, HEAT was used to calculate the benefits of bicycling for reducing emissions.
According to these results, 475 tonnes of emissions are avoided every year. With CO2 cost of 171 EUR/tonne or 1.861 SEK/tonne, this equates to 879.509 SEK/year (Boliga Finans ApS, 2020). If you divide this by the total amount of commuting done by bicycle of 1.525.040 km, the benefit equates to 0,58 SEK/km saved by bicycling (THE PEP, 2019).
3.2.6 Air pollution
Air pollution was taken at 0,02 SEK/km in a rural environment and at 0,06 SEK/km in an urban environment (Bångman & Nordlöf, 2018). Since it was assumed that 20% of the trips takes place in an urban area and 80% in a rural area, the average used for the conservative estimate was 0,03 SEK/km (Google, n.d.).
For the high estimate, the results from HEAT were used (THE PEP, 2019). This calculated the benefits of bicycling to be 9.210 EUR or 100.251 SEK/year (Boliga Finans ApS, 2020). Divided by the total amount of bicycling done over the year, this equates to 0,07 SEK/km (THE PEP, 2019).
15 3.2.7 Noise
The costs of noise are estimated to be 0.03 SEK/km in rural areas and 0.18 SEK/km in an average urban area (Bångman & Nordlöf, 2018; Transport DTU & COWI, 2019). With the 20/80 division of urban versus rural areas, the average noise costs was taken at 0,06 SEK/km.
3.2.8 Winter maintenance
The cost of winter maintenance in the form of snow removal and anti-slip measures are estimated at 0,34 SEK/m2 or 1,02 SEK/m on a three meter wide bicycle path (S. Östberg, personal communication, March 23, 2020). This is the cost per trip. The total amount of trips needed per year to clear the bicycle path was estimated to be fifty. If 1,02 is multiplied by the total length of 7.800 meters and by 50 turns the yearly cost comes to 397.800 SEK. If this is divided by the total amount of bicycling distance covered in a year at 1.525.040 km, which brings the cost per kilometre of winter maintenance for bicycling (5% mode share) at 0,26 SEK/km.
3.2.9 Construction costs
The construction cost of a bicycle road in Krokom is estimated to be 1.000 SEK/m2 or 3.000 SEK/m for when the bicycle path is three meters wide (S. Östberg, personal communication, March 23, 2020). The total cost of the 7.800 meter long path would thus be 23.400.000 SEK over the twenty year life span of the path.
Divided by the total amount of bicycling distance covered over a year, this comes at 0,77 SEK/km with a 5%
bicycle mode share.
3.3 Costs for driving and bicycling
In table 3.3 all the results from the CBA are shown. For driving by car, the cost per kilometre is 6,24 to 8,68 SEK. For bicycling, this is 4,91 to 9,84 SEK per kilometre. In the “status quo scenario”, in which no extra bicycle infrastructure is built, the cost of driving is estimated to be between 175 and 244 million SEK. In the increased bicycling scenario in which infrastructure is built and 5% of the commuters take the bicycle to work, the cost of cars is estimated to between 165 and 230 million SEK, but with the extra cost of bicycling at between 7,5 to 15 million SEK.
16 Table 3.3
A comparison of all the impact category values between the status quo scenario (0% bicycling) and the increased bicycling scenario (5% bicycling)
Status quo scenario (d = 92%, pt = 8% b = 0%)
Increased bicycling scenario (d = 87%, pt = 8% b = 5%)
Driving Driving Bicycling
SEK/km Million SEK/year Million SEK/year SEK/year Million SEK/year
Travel time costs 1,74 49 46 6,45 10
Vehicle operating costs 4,25/6,65 119/187 113/176 0,68 1,0
Health benefits 0,00 0 0 -4,34/-2,61 -6,6/-4,0
Accident cost 0,16 4,5 4,2 1,68/4,50 2,6/6,9
Climate change benefits 0,00 0 0 -0,58/-0,20 -0,9/-0,3
Air pollution 0,03/0,07 0,8/2,0 0,7/1,9 0,00 0
Noise 0,06 1,7 1,6 0,00 0
Extra winter maintenance 0,00 0 0 0,26 0,4
Extra construction cost 0,00 0 0 0,77 1.2
Total 6,24/8,68 175/244 165/230 4,91/9,84 7,5/15
Note: The marginal costs of driving in SEK/km are the same for the same for the status quo as for the increased bicycling scenario, and thus are only shown once (status quo scenario). When there is a conservative and high estimate for certain impact categories, they are separated by a “/”.
3.4 Marginal cost per impact category
In figure 3.1, the values in SEK/km for the different impact categories can be compared. It is apparent that some categories contribute to a large extent to the total costs, while others contribute little to none and how some impact categories have a large range of possible values.
17 Figure 3.1
Comparison between the marginal costs and benefits of driving and bicycling in the increased bicycling scenario
Note: The value range for some of the impact categories is represented by the darker colour.
3.5 Changing the bicycling mode share
In figure 3.2, it can be seen how, when the bicycling mode share increases, the cost difference of the total costs between the status quo scenario and increased bicycling scenario is increasing or decreasing with it. The starting point at 0% bicycle mode share is above zero due to the initial cost of constructing and maintaining the bicycle path, regardless of how many people use it. When the most conservative estimation is taken, the added extra costs of an increased bicycling scenario continue to rise, meaning that it is costing society more to have these extra bicyclists. In the high and average estimate, these extra costs decrease as more people start to commute by bicycle. This means that it quickly becomes beneficial to society to have bicyclists use the new infrastructure to commute compared to having them take the car.
18 Figure 3.2
Bicycle mode share against the conservative-, high- and average extra yearly costs in the increased bicycling scenario compared to the status quo scenario without any bicycling
4. Discussion
As can be seen from the results from this case study, it is not apparent that an investment in bicycle infrastructure will surely be a benefit for society. The range is wide so that increased bicycling covers both a positive and negative societal outcome. When using the conservative estimate, it will cost society to have a higher mode share of bicyclists due to a higher marginal cost of bicycling compared to car-use. When looking at the high estimate of the range, it becomes beneficial to society to work towards an increased bicycling scenario, which means the marginal costs for bicycling are lower than for car use. As can be seen in figure 3.2, the average of these to extremes in the range of outcomes seems to indicate that an increased bicycling scenario seems to become beneficial for society when a bicycle mode share of around 5% or higher is reached.
Below, these results are discussed by looking at the different aspects specific to this case study that could responsible for this outcome, while the method itself is also critically looked at.
19 4.1 The major impact categories
Travel time, vehicle operation, accidents and health are the four major impact categories. Together they are responsible for by far the largest part of the total costs associated with both mode of transport.
4.1.1 Travel time and congestion costs
Travel time is such an important factor when doing a cost-benefit analysis because our time is seen as valuable and thus it becomes more and more costly, the longer we spent that time commuting. Not only the time spent in a commute, but also the circumstances are important in determining the cost. The easier and more comfortable the commute, the less costly it is perceived to be. While bicycling can be seen as the more uncomfortable option, compared to sitting in a car, it is not hard to imagine being more stressed stuck in traffic than it would be to bicycle through a park to get to work. These changing costs depending on the circumstances are reflected in the significantly higher hourly costs of cars stuck in traffic and bikes sharing the road with cars instead of being separated (Bångman & Nordlöf, 2018; Björklund & Isacsson, 2013). These environmental factors do not take away the fact that cars are faster than bicycles. This is especially true for this case study that largely takes place outside of urban areas and over a large distance that requires cars to take the highway. Where in urban areas, bicycles can often compete with cars or public transport due to their lower average speeds over shorter distances; this cannot be said for this case study. The average speed of cars was estimated to be three times as high as the bicyclist’s (60 km/h versus 20 km/h), making the car journey three times as short. Ignoring the circumstances, this only, reduces the costs per kilometre of car travel by a factor of three, compared to bicycle commuting. Had the commute been a short journey of a few kilometres in an urban area, the commuting time could have been the same for both forms of transport which would result in similar marginal travel time costs too.
4.1.2 Vehicle operation costs
Vehicle operation is where most costs are made when driving a car. As is the case with travel time, these are internal costs, borne by the commuter themselves and not by externalized to society. Although the case could be made that GHG emissions, air and noise pollution should all fall under vehicle operation costs and be fully borne by the commuter, this is currently not the case. Regardless of that, owning a car is expensive compared to owning a bicycle. Even though you can easily cover more ground by car than by bicycle, the marginal costs of car travel are still many times larger than that of bicycling. With cars, there are several high costs involved such as the purchase, taxes, insurance, maintenance and fuel. For bicycles, these costs are generally many times lower. Vehicle operation costs are one of the best arguments against car-use on an individual level.
Even though the vehicle operation costs are undoubtedly higher for cars, it has been hard to estimate the average costs. No existing value exists for the average Swedish car. With the help of Trafikanalys, that provided some sources that have tried the estimate the average costs of owning a car, a value was established,
20
but it must be noted that this value does not take into account second-hand cars (T. Petersen, personal communication, March 16, 2020).
4.1.3 Accident costs
Accidents induce a high cost for bicycle commuting while they are quite low for car commuting. This is another factor that is highly debatable because of how different car and bicycle accidents are. In general bicyclist is seen as more vulnerable than car users. Even though there might be fewer total bicycle incidents than car accidents, per travelled kilometre, the chance of an accident happening are higher as well as more costly due to higher vulnerability (Dekker, 2016; Ommeren et al., 2012).
For the high estimate, numbers from a Danish database were used that significantly reduced the associated cost from 4,5 to 1,68 SEK/km, although still 10 times higher than the costs related to car accidents.
The uncertainty of the costs related to bicycle accidents is mirrored in how it differs from study to study. In the study for from Dekker, done in Canada, it is estimated at around 9 SEK/km, while in the study from Gössling et al. that looks at Copenhagen, it is just over 1 SEK/km and in another study by Gössling et al., that looks at the whole of the European Union, it is even lower at around 0,7 SEK/km (2016; 2019; 2015).
This wide range might partly be explained by geographical and demographical differences. Sometimes, safety comes in numbers. The more people choose the bicycle, the saver it becomes (Ommeren et al., 2012). This might not be the case in the beginning, when the levels of bicycling are so low that more bicycling causes more accidents due to the lack of infrastructure or other road users that are unaccustomed to bicycles on the road, but in places where bicycling is already established an increase in bicycle mode share will increase safety (Olde Kalter, 2007). For example, in the Netherlands, the bicycle mode share is generally very high compared to most other countries in the world. This has led the estimated accident costs per kilometre with cars or bicycles to be close to being virtually equal. In urban areas this is around 0,7 SEK/km for both cars and bicycles, while it is 0,27 SEK/km for both modes of transport outside of urban areas (Ommeren et al., 2012).
When looking at the demographical differences, the age the person matters too. In the Netherlands, younger people who have just gotten their driver’s license, changing their mode of transport from the car to the bicycle, will have a positive impact on traffic safety, while the opposite is true for people over fifty years old.
They tend to drive relatively safely, but fall more often while riding a bicycle (Ommeren et al., 2012).
When taking these differences in consideration for the case study laid out in this research, it can be concluded that the cost of accidents might be on the lower side of the range discussed above. This is mostly due to the fact that the proposed infrastructure is fully separated from the roads, in a quiet rural area and over a longer distance which is likely to discourage older accident prone commuters to choose the bicycle. Due to these
21
factors, the 4,5 SEK/km used, even though a value based on Swedish data, might be too high for this case study.
4.1.4 Health
The health benefits associated with switching from a car to a bicycle is among the most difficult impact categories to calculate, due to the many unknown or uncertain variables. For this reason, the tool created by the WHO was used. When available, local data was used in this tool. When there was not local data available, the averages or standard values from the tool were used.
One particularly important value that has a large influence on the benefits associated with health is the Value of a Statistical Life (VSL). The EU value of almost 4 million euros was used since no up to date Swedish value could be found.
The biggest problem with using HEAT is that the calculations that are done by the program are not shared with the user. After the input is given and some standard values are changed according to the circumstances, the results are given. On the other hand, in an elaborate guide on how HEAT works, the inner workings of the program, including the formulas used, are explained (Kahlmeier et al., 2017). On top of that, other studies that include CBA’s of active travel have used the HEAT tool to help them do calculations on health benefits.
This means that the tool is trusted in science (Dekker, 2016; Pérez et al., 2017).
The estimated health benefits from two Copenhagen studies were estimated around 4,3 SEK/km, which falls within to the range used in this study, although close to the high estimate (COWI and City of Copenhagen, 2009; Gössling & Choi, 2015). The CBA done for bicycling in Calgary in Canada came with a higher benefit of 6,9 SEK/km, while the one done for all of the European Union estimated the health benefits to be worth 3,6 SEK/km (Gössling et al., 2019). Except for the Canadian study, the benefits per kilometre are all fairly close. This can indicate less uncertainty about the health benefits of bicycling and makes the value calculated for this study more reliable.
4.2 The influence of the rural context
This case study and more specifically, the rural and long-distance character of this case study, have add a real influence on the results. This rural context is mostly manifested in the long distance and small pool of potential users. The influence of the rural context can most clearly be seen in the travel time costs being high due to large difference in average speeds between the two modes of transport looked at for this study. Other factors with a smaller influence are the air and noise pollution of cars being significantly lower in rural areas, compared to urban areas. Congestion might also be less of a car-related issue in rural areas even though that was not compared to other cases in this study. There seem to be indications that accident costs are lower in rural areas, compared to urban areas (Ommeren et al., 2012), which would further shed doubt on the
22
surprisingly high accident costs given by the Swedish report (Bångman & Nordlöf, 2018). Even though a rural context might not have strength in numbers, due to a limited amount of people willing to travel long distances by bicycle, the ones that choose to bicycle 40 km a day contribute to a larger extent individually to benefits such as avoiding GHG emissions and being healthy than urban commuters travelling only short distances.
This might thus compensate some of the drawbacks associated with bicycling in a rural context.
4.3 CBA as an indicator for social and environmental sustainability
A cost-benefit analysis is one way of getting an insight into the sustainability of a project. It is practical, because it attempts to include the most important externalities, both environmental and social, and move them into the economic realm to help give decision makers a view on the long term sustainability of the proposed project. On the other hand, the sustainability related conclusions that can be drawn based on a CBA are limited and partial by nature (Hickman & Dean, 2018).
They are limited because the CBA method can be seen as a form of weak sustainability. This means that it is set up in such a way that the environmental costs can be compensated for by social or economic benefits. The whole method is based on the fact that natural capital can be exchanged for economic capital that is deemed of “equal” value (Bell & Morse, 2013). Strong sustainability, on the other hand, treats the environment with its natural capital as something that can’t be exploited and replaced with manufactured capital.
The CBA method is inherently partial because of the subjective approach that comes with putting a monetary value on the environment and other aspects of society, but also what is included and valued in the first place (Bell & Morse, 2013). One of the pitfalls this creates is the fact that certain social or environmental impacts can be overlooked (Hickman & Dean, 2018).
Looking at this case study, an argument for the usefulness of a CBA can still be made. Bicycling as a transport mode is already widely accepted as beneficial for both society and the environment, being regarded as a generally more sustainable option than car-use. With this information in mind, what is done in this case study is looking at the economic costs that come with bicycling and bicycle infrastructure, and see if they weight up against the societal and environmental benefits. When taking in mind that this method is partial and should be used in tandem with other methods such as environmental impact assessments and multi-criteria analysis, it can be of help in the decision making process.
While a CBA does not replace an environmental impact assessment, it can serve as a tool for policy makers on deciding on if a certain project plan can have an overall positive effect on the municipality.
23 4.4 What was not included
There are several factors that were left of this CBA. Some environmental factors were left out such as land and water pollution associated with the different scenarios as well as the effects of different modes of transport on other ecosystem services, the extra emissions related to the extra food consumption when bicycling and the emissions related to congestion, the effect the seasons have on the amount of bicycling done and the general maintenance of both bicycling paths and roads. These were mostly factors that were too complicated to implement in the CBA, but sometimes, as is the case with maintenance, are shown to have a very low cost related to them (Haraldsson, 2007). These are all factors that could be included for a fuller picture although they might thus sometimes turn out to not have a large influence on the outcome.
4.5 Conclusion
The results from this case study have shown that the desirability of the construction of this bicycle path is highly dependent on what values are being worked with. While it might not seem beneficial for society to build looking at the conservative estimates, it shows promise when looking at the high estimate or even when taking the average of the range.
This discrepancy becomes more problematic the higher the share of the cost or benefit of an impact category is on the outcome of the CBA. Some of the major impact categories used in this study such as operating costs, health benefits and accident costs, have shown to vary significantly, based on the source used or the input given. To a great extent, these impact categories have caused the needle to move from beneficial to society, to non-beneficial, depending what end of the range was taken.
One other major impact category, the travel time cost, is more reliable and it is this impact category that also shows us the possible weakness of this proposed bicycle path in a rural setting. Because of the large distances worked with in this case study; it is hard to make the case for the bicycle as an efficient form of transport. The time loss and the associated costs, compared to the other transport alternatives, especially the car, makes the bicycle an unlikely competitor when it comes to the efficiency and speed of movement. Since this impact category is also internal, meaning that it is a cost directly incurred by the commuter, it can dissuade people to choose the bicycle. The commuters that do choose to change their mode of transport from a car to a bicycle do in many cases have a larger positive societal and environmental impact than their urban counterparts due to the many kilometres of car travel avoided.
Although no winner when it comes to time efficiency, the largest benefits of bicycling in this case study are in the low operating costs and in the health benefits. Both of these benefits, although not always fully comprehended, are to some extent internal and thus can persuade people to choose the bicycle over the car.
What is clear is that the more personal (internal) benefits or costs play the most important role in determining
24
the desirability of the bicycle path. Although health benefits and also accident costs can both, at least partly, be seen as societal (external) costs or benefits, most other impact categories that relate to society such as maintenance and construction or that are environmental impacts such as the climate change benefits or air pollution, only play a small role in the outcome. When more environmental impacts that are now overlooked are taken into account for this CBA, the influence of environmental impacts on the costs and benefits might rise accordingly. This does not change that this analysis is a form of weak sustainability, in that it enables for compensation of natural capital by manufactured capital. To ensure that the environmental damage of a project is minimized, some form of Environmental Impact Assessment needs to be performed in addition to a CBA.
4.6 Future research
To get a better grasp on what desirable societal effect a bicycle path between Östersund and Krokom could have, a continuation of this research is necessary. This could be done in different ways. One way is to do more research on the true value of certain impact categories in this case study. Another aspect that could be researched further is the use of the e-bike to shorten the time spent commuting and thereby making bicycling a more attractive option. A better view of the willingness of people to change their mode of transport could also help in getting an estimate on how much the new bicycle path will be used. This can be done by a survey.
Most literature is focused on bicycling over shorter distances in densely populated urban setting. The suggestions for further research done here go in the other direction and are aimed at adding knowledge on bicycling over longer distances, in predominantly rural settings and with a relatively low user volume.
