PUBLIC TRANSPORT IN SMALL TOWNS AN AREA WITH GREAT POTENTIAL Andreas Persson Department of Technology and Society, Lund Institute of Technology Sweden 1. INTRODUCTION The project Attractive Public Transport in small cities Requirements and possibilities in attracting more passengers started in the year 2000 at the department of Technology and Society at Lund Institute of Technology. Important aims of the project are to find the similarities and differences between the conditions in small and large towns and to find a suitable planning strategy for the small town traffic. The first thing to do is to define what a small town is. There are about 60 Swedish towns with less than 40 000 inhabitants with a public transport system of their own. The Stockholm area is excluded, since the towns here are functionally linked to the urban area. Since there is a gap between the largest town in the less than 40 000 inh. group and the rest of the Swedish towns this is a natural and practical definition. The attractiveness of a public transport system depends on many different factors. These have been studied in detail for example by Sjöstrand (2001), Norheim & Stangeby (1993) and Widlert et al. (1989). Sjöstrand and Norheim & Stangeby show the relative weight for in-vehicle time, walking time, headway and transfers in Göteborg and Oslo (table 1), while Widlert et al. have investigated the willingness to pay for different quality factors (table 2). Relative weight Göteborg (Sjöstrand) Oslo (Norheim & Stangeby) In-vehicle Walk Headway Transfer-time 1.0 1.1 1.1 1.5 1.0 2.0 1.8 1.25 Table 1: Relative time weights for passengers in Göteborg and Oslo
Factor: Willingness to pay (SEK): Better punctuality 1.28 0.99 Transfer -0.93-0.72 Rain shelter at bus stop 0.97 0.70 Real time information at bus stop 0.67 0.48 Not getting a seat on the bus (2 min) -1.69-0.64 Not getting a seat on the bus (10 min) -2.93-1.11 Table 2: The willingness to pay for some of the factors investigated by Wilnert et al. Studies like these provide valuable information about which factors to improve in public transport. Most of these results are based on research projects that have been carried out in large or medium-sized cities. It is therefore interesting to study if these results also can be applied on public transport in small towns, or if the conditions differ so that other factors become more important. There are also differences between the public transport passengers in big and small towns. In this paper a national survey carried out by the Swedish Public Transport Association is used to identify some of these differences. The variables that are investigated are age, gender, travel frequency and general attitudes towards public transport. In Hässleholm and Ängelholm, two small Swedish towns with about 18 000 and 22 000 inhabitants respectively, the bus systems were changed in July 2001. The main idea was to provide traffic with higher frequency. In Hässleholm the routes were also changed and the central bus terminal was moved to a more attractive location. The number of passengers has increased since the changes and this was studied using time series analysis. An inquiry was also carried out among the passengers aiming to find attitudes about the changes and how the changes have influenced their travel behavior. 2. DIFFERENT CONDITIONS IN SMALL TOWNS Comparing geographic and demographic factors in the small towns with other Swedish towns give valuable information about differences concerning planning conditions for the public transport system. It is for example known that high population density makes it easier to provide an effective transport solution. In Swedish small towns the mean population density is 1550 inh/km 2 compared to 2400 inh/km 2 in the 13 largest towns (SCB, 2003). The total
number of inhabitants in the town is also of importance, since more people provide a larger basis for the public transport. The distances within the town are shorter in small town. This makes walking and cycling more attractive compared to going by bus. In Hässleholm, the distance from a peripheral housing area to the town centre is approximately 1.9 km. This distance takes about 6 minutes to travel by bike while a bus trip from the same area takes 12 minutes. This gives a quotient between travelling time with bus and bicycle of 2.0. For a similar housing area in Malmö, Sweden s third largest town, the distance is 5.2 km. The travelling time by bus is 24 minutes and by bike 16 minutes, giving a quotient of 1.5. The headway in Malmö is in the example 8 minutes while it is 30 minutes in Hässleholm. With the waiting time and the hidden waiting time included the difference between bicycle and bus is even bigger in Hässleholm. Walking and cycling are the most environmental friendly means of transport, so the aim should not be to get pedestrians and cyclists to take the bus. Nevertheless it is an important factor that can explain differences between the towns. Other strictly geographic factors that can make an impact on the number of passengers in the public transport system are topography, the shape of the town and the presence of barriers such as rivers or big traffic routes. These factors should not differ much between small and large towns. 3. PASSENGERS OPINIONS ABOUT PUBLIC TRANSPORT There are also differences between the passengers and their attitudes towards public transport in small and big towns. The Swedish Public Transport Association carries out a national survey every year (SLTF 2003). 13 000 randomly chosen people aged between 15 and 70 are telephone interviewed about their opinion on public transport in their town. Background data about the respondent is also collected. In this paper data from three surveys (2000-2002) have been used and divided into two groups, one with people living in small towns and one with people living in the 13 biggest towns. In the latter group the interviews from Stockholm have been excluded since both the public transport system and the answers in the survey differs significantly from the others in the group. This gives 11 466 interviews in the first group and 14 881 in the second. There is a big difference in how big share of the respondents that actually use the public transport. In the big town group 36 % use public transport at least a few times a month while in the small town group only 21 % say they do. The share that has access to a car differs between the groups. In small towns it is 88 % but in big towns only 76 %. In big towns 59 % of the respondents that have access to a car use the public transport at least a few times a month. In small towns this share is only 35 %. This indicates that public transport have better possibilities to compete with the car in the big towns.
100% 80% 60% 40% 20% 0% 88% Small towns 76% Big towns Figure 1: Percentage with access to car 100% 80% 60% 40% 20% 0% 35% Small towns 59% Big towns Figure 2: Percentage with access to car, but using public transport at least a few times a month There is also a big difference in how attractive the public transport is for work trips. In small towns only 34 % of the employed use public transport at least a few times a month. In big towns that share is 55 %. 100% 80% 60% 40% 20% 0% 34% Small towns 55% Big towns Figure 3: Percentage of employed using public transport at least a few times a month In the next part of the survey the respondents were asked to what extent they agree about certain statements. The first statement was I am pleased with the public transport all in all. In the small town group 54 % answered that they are pleased or very pleased. The corresponding figure for the big town group
was 57 %. There were many other interesting differences between the answers from the two groups. These are shown in table 3. Statements that people in small towns agree to a higher extent - The bus stops and the stations are well-managed - The vehicles are neat and tidy - The driver s behaviour is nice - It is comfortable to travel - It feels safe to travel - Information about delays and temporary changes is good. Statements that people in big towns agree to a higher extent - It is easy to get information about times of departures - It is easy to buy tickets - The bus stops and stations are near by - The times of departures are suitable for my needs - The bus routes go the nearest way - The information about changes of timetables or routes is good. - It is easy to travel Table 3: Statement to which people in small and big towns agree to a different extent It is obvious that public transport in small towns is better than in big town in certain aspects. Service and comfort factors generally get higher marks, while factors connected to travel speed and the design of bus routes get higher marks in the big towns. The fact that the big town group to a higher degree is pleased with the public transport indicates that these factors are of great importance. One way to calculate how important the different factors are is to study the correlation coefficients between the answers on the different statements and the answer on the question how pleased they are all in all. The higher correlated the factors are, the more important they are for the all in all mark. The five most important factors in small town public transport according to this method are shown in table 4. The correlation coefficient and the percentage of the respondents who are pleased or very pleased with the factor are also shown.
Factor Percentage pleased or very pleased Correlation coefficient Simplicity 65 % 0,64 Times of departures 39 % 0,51 Sensitivity to complaints 28 % 0,49 Design of bus routes 46 % 0,48 Travel speed 55 % 0,47 Table 4: The five most important factors in small town public transport according to this analysis One problem with this method is that it calculates which factors that are most important for the marks in the survey. It is not obvious that these correspond with the most important factors to achieve a successful public transport system or to increase the number of passengers. 4. PUBLIC TRANSPORT IN SMALL TOWNS TODAY In all the 60 Swedish small towns in this study the public transport system consists of a bus system, often with focus on short walking distances and good area coverage, naturally at the expense of travel times and frequency of departures. The typical frequency is one bus an hour or in some cases half an hour. On weekends the traffic in most of the towns is limited to the minimum if any at all. In many of the towns the bus routes are designed as loops, something that in most cases lead to long travel times. If the loop always is operated in the same direction the passengers living in the beginning or the end of the loop have to travel all the way around in one of the directions. Figure 4 shows three typical small town bus systems from the towns Mora, Ludvika and Lidköping.
Figure 4: The bus systems in Mora, Ludvika and Lidköping One way of measuring the success of public transport is the number of trips per inhabitant and year. For the Swedish small towns there are big differences between the towns. The town with the highest number of trips have more than ten times as many trips as the ones with the lowest numbers. Table 5 shows number of inhabitants, population density, number of bus routes, average headway and number of trips per inhabitant and year for ten Swedish small towns. Even though Karlskrona is the biggest town and has the highest number of trips, population and population density don t seem to be the most important factors. It is instead a combination of many bus routes and high frequency that gives a high number of trips.
Town Inhabitants Population density Number of routes Average headway Trips per inh. and year Karlskrona 32 000 1518 10 30 min 79 Hudiksvall 15 000 1544 5 30 min 43 Bollnäs 12 700 996 3 30 min 27 Ludvika 14 400 1319 5 60 min 23 Mora 10 800 886 3 45 min 21 Ängelholm 21 700 1784 2 30 min 13 Avesta 14 800 1121 3 45 min 10 Hässleholm 17 300 1492 2 30 min 8 Lidköping 24 400 1661 4 45 min 7 Allingsås 22 300 2044 5 60 min 4 Table 5: Data from 10 of the studied towns Fredriksson et al. (2000) showed examples of small towns in Europe with successful and attractive bus systems. Common factors for these are few and simple bus routes, high frequency and a well designed central terminal where passengers easily can change to other routes. These towns have a considerable higher number of bus trips per inhabitant and year than the Swedish towns of the same size. This indicates that factors that they have concentrated on are of great importance for the attractiveness and consequently the number of passengers using the public transport system. In figure 5 the bus systems in the German towns Lindau and Lemgo are shown. Table 6 shows the same data for the German towns as table 5 does for the Swedish towns. Figure 5: The bus system in Lindau and Lemgo
Town Inhabitants Population density Number of routes Average headway Trips per inh. and year Lindau 24 100 1944 4 30 min 112 Lemgo 31 400 3204 3 20 min 75 Table 6: Data from two German towns with a high number of passengers The population density is higher in Lemgo than in any of the Swedish towns, and in Lindau it is higher than most of the Swedish town. This might of course be one important reason why the number of trips is higher, but there is also a difference in average headway between the groups. The travel speed is not visible in the tables but the differences in the design of the bus routes indicates that the routes are faster in Lemgo and Lindau than in the Swedish towns. 5. EVALUATING IMPROVEMENTS IN TWO SWEDISH TOWNS In the Swedish towns Hässleholm and Ängelholm the public transport company recently changed the bus system to make it more attractive and to increase the number of passengers. In Ängelholm the headway was changed from 60 to 30 minutes and in Hässleholm the headway was set to 30 minutes on all of the routes and the routes through the central parts of the town was changed. Also the central bus terminal was moved to a more attractive place in front of the railway station. With the changes the two towns improved the travel speed for the passengers, principally by improving the frequency of departures. In Hässleholm simplicity and travel speed on board the bus also was improved. As mentioned before these are important factors to the passengers. Figure 6: The old bus system in Hässleholm to the left and the new to the right In Ängelholm the number of passengers was increasing already before the changes were made. This is rather unusual for Swedish towns of the same
size, and it is not completely clear why the trend is so positive. The public transport company is working to increase the quality factors and this might be one reason. Another reason might be the increasing number of passengers travelling with the regional trains. Since the same ticket is used on the local bus, the bus-trip is free when you have the train ticket. In Hässleholm there was no obvious trend before the changes were made. To find out how much of the increase of passengers that depend on the changes made in the traffic, a time series analysis was made. Factors that are known to influence the demand for public transport were collected for a period of 55 months (35 months before and 20 months after the changes took place). To calculate the number of trips per inhabitant and month a multiple linear regression was made (Blom, 1970): y = k 0 + S k i * x i + e where y = number of trips per inhabitant k 0 = constant for the model k i = constant for the factor i x i = factor i e = error term The factors that were tested in the model were: Calendar data (number of working days per month) Temperature (SMHI, 2003) Ticket price Gasoline price (SCB, 2003) Industrial production (SCB, 2003) Dummy factor for the month June Dummy factor for the changes A regression was then made over the 55 months to find a model that describes the number of monthly trips as accurate as possible. To decide which factors to use in the model the adjusted R-square was studied. This shows how well the model describes the variation in data with a figure between 0 and 1. The closer to 1 the R-square is, the better the model is. The adjusted R-square is used because the R-square value overestimates the model when many factors are used. Another measure of the quality of the model is the t-value for every factor. If a factor is to be decided with a confidence level of 95 % the t-value should be higher than 1.96. Factors with a lower t-value than 1.96 and factors that made the adjusted R-square lower when added to the model were excluded from the final model. As mentioned before there was an increasing number of passengers already before the changes took place in Ängelholm. To deal with this increase a
factor called general increase was introduced in the model. In the final model these factors are used: Adjusted R-square: 0.918 Model parameter t-value k 0 3.862 6.48 Number of working days 0.019 7.95 Average temperature -0.013-6.15 General increase 0.208 7.44 Dummy for June -0.264-5.36 Dummy for the changes 0.156 3.41 Table 7: The model for monthly trips per inhabitant in Ängelholm In figure 7 the actual number of trips and the calculated number of trips are shown over the period. Number of trips per inh. and month in Ängelholm 2 1,8 1,6 1,4 1,2 1 Calculated Actual trips 0,8 0,6 0,4 0 10 20 30 40 50 60 Figure 7: Number of trips calculated by the model compared to actual trips over the period July -98 to February -03. In Hässleholm the number of passengers is calculated in two different ways over the period. During the first period, from 1999 to 2000, some of the regional bus routes were added to the passenger statistics. To deal with this, a dummy called method change has been introduced in the model in January 2001, giving the final model:
Adjusted R-square: 0.8380 Model parameter t-value k 0 7.1496 6.50 Number of working days 0.030 6.56 Average temperature -0.023-5.94 Dummy for June -0.211-2.42 Dummy for method change -0.811-11.17 Dummy for the changes 0.408 2.42 Table 8: The model for monthly trips per inhabitant in Hässleholm Number of trips per inh. and month in Hässleholm 1,8 1,6 1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 0 10 20 30 40 50 Calculated Actual trips Figure 8: Number of trips calculated by the model compared to actual trips over the period January -99 to February -03. The model parameter for the changes shows how many new passengers there are because of the changes. In Ängelholm the parameter is 0.156, meaning that the number of trips per inhabitant and month has increased by 0.156. This gives about 4 000 new passengers a month or a 13 % increase. For Hässleholm the parameter was 0.408, the increase in number of passenger a month was 7 000 or 37 %. The number of trips per inhabitant and year, which were shown in table 5, has increased from 13 to 15 in Ängelholm and from 8 to 13 in Hässleholm. To find out more about the passengers opinions an inquiry was made on board the buses. This was done on a weekday from 12.00 to 18.00 so that most of the passengers were given the chance to answer. The inquiry was designed to be answered on the bus during the trip, but it was also possible to
send it in by mail. This led to a high response rate; over 90 % of the questionnaires that were handed out came back. Approximately 15-20 % of the passengers refused to take a questionnaire when entering the bus, so the total response rate was about 70 %. First some background data such as age, gender and errand for the trip was collected. Figure 9 shows the age of the passengers in the two towns. The share of people under 18 is remarkable in Hässleholm, with 50 % being under 18. In Ängelholm the passengers are more scattered between the age groups. Percentage of passengers in age groups in Hässleholm Percentage of passengers in age groups in Ängelholm 60 50 40 30 20 10 0 0-18 19-30 31-50 51-65 66-80 81- Age 40 30 20 10 0 0-18 19-30 31-50 51-65 66-80 81- Age Figure 9: Age of passengers in the two towns In Hässleholm 64 % of the passengers were women and in Ängelholm 75 %. This is not anything unusual, but it shows that it is harder to get men to take the bus, especially when they have a driver s license. In the under 18 group the distribution according to gender is much more balanced. Most of the trips are made from the outskirts to the central parts of the town, or the other way around. Around 80 % of the passengers get on or off at one of the central stops. In Hässleholm the new central bus stop is highly frequented with about 75 % of the passengers getting on or off. Passengers that stated that they travel more by bus than before the changes were asked how they travelled before. The result is shown in figure 10.
Hässleholm Ängelholm Car 19% Bus 27% Taxi 4% Walk or bike 50% Mobilty service 6% Car 17% Bus 17% Taxi 5% Walk or bike 55% Figure 10: How the new passengers travelled before the changes In Ängelholm, the share of older passengers is bigger than in Hässleholm. That explains why the new passengers to a higher extent travelled by taxi or mobility service before. In both towns about one fifth of the new passengers are former car users. The inquiry was done during a period of cold weather in December. This is probably one reason why the share that usually goes by foot or bike is so big. 14 % of the passengers in Ängelholm and 21 % in Hässleholm stated that they could have used the car for this particular trip. 6. DISCUSSION Many things indicate that it is harder to design an effective public transport system in a small town. The population and population density have been mentioned and the fact that the average trip length is shorter in small towns makes walking and cycling more attractive. Another important difference between small and big towns is that in the small town the needs from different costumer groups must be met in one single system. In bigger town you can often provide a separate service net for people that need shorter walking distances and maybe express routes during rush hours. The small town bus systems that are used in many Swedish towns are not attractive to commuters and other passengers that prioritize short travel time. Important incentives for public transport investments are to deal with congestion and pollution. These problems don t appear to such an extent in the small towns. It is therefore important to show other advantages such as providing mobility for people that don t have access to a car or who are unable to walk or cycle. Even if the pollution problem is smaller in small towns it would be good for the environment to reduce car traffic.
The results from the two Swedish towns in this study show that it is possible to increase the number of passengers using public transport. The fact that some towns in other parts of Europe have a much higher number of passengers and that the number of passengers in the Swedish towns differ so much also indicates a big potential. 24 of the 60 Swedish towns have less than 10 trips per inhabitant and year. It ought not to take too much effort to increase the number of passengers in these towns. Results from the inquiry, the high number of passengers in the other European towns and the increase of passengers in the two Swedish towns indicate that travel speed, simplicity and headway are important factors to improve in the small towns. 7. REFERENCES Blom, G. (1970) Statistikteori med tillämpningar, Studentlitteratur, Lund, Sweden. Fredriksson, L., Wendle, B. and Möller, J. (2000) Attraktiv kollektivtrafik i små städer Förutsättningar och möjligheter för ett ökat resande. Förstudie, Bulletin 194, Lund University, Lund Institute of Technology, Department of Technology and Society, Traffic planning. Lund, Sweden. Norheim, B. and Stangeby, I. (1993) Bedre kollektivtransport Oslotrafikanternas versettning av høyere standard, TØI Rapport 167/1993, Oslo, Norway SCB, (2003) Sweden s statistical databases, www.scb.se, Sweden Sjöstrand, H. (2001) Passenger assessments of quality in local public transport measurement, variability and planning implications, Bulletin 202, Lund University, Lund Institute of Technology, Department of Technology and Society, Traffic planning. Lund, Sweden. SLTF, (2003) Kollektivtrafikbarometern, report generated from the database, Sweden SMHI, (2003) Statistics from the Swedish Meteorological and Hydrological Institute, Sweden Widlert, S., Gärling, T. and Uhlin, S. (1989) Värdering av kollektivtrafikens standard, TFB-rapport 1989:2, Transportforskningsberedningen, Stockholm, Sweden.