Transportation Deployment Casebook/2022/Alberta

Historical Overview
Streetcars were the precursor to modern urban light rail vehicles. They were powered by electricity, ran on rails through urban and suburban area and were predominantly (if not exclusively) used for the transport of paying passengers. The introduction of systems in Canada somewhat mirrored that in the United States in the late 1800’s.

Early Development - A Market Niche to be Filled
The benefits and advantages of streetcars are best demonstrated when contrasted with preceding modes of transport. Ball (1991) notes that horse drawn streetcars began in larger Canadian cities, such as Montreal, in the 1860’s. These involved wooden-wheeled carriages during Spring which tracked through mud, rail cars in Summer and sleighs in Winter (Ball 1991). Similar technologies would have been the only option in Albertan cities.

Horse drawn street cars were uncomfortable for passengers when riding on wooden wheels and these were superseded by metal-wheeled cars riding on rails (Ball 1991). However, these still retained the obvious issues with the disposal and management of animal excrement.

Ball (1991) also notes that during this era, traditional railways were dirtier, generating soot and smoke, in addition to requiring a dedicated right-of-way. The absence of a viable private form of transport (the automobile) at the time meant that a niche existed for a cleaner, more local alternative to service the growing suburbs of Canadian cities.

Cable cars appear to be the first real alternative to horse-drawn street cars. However, (Gregersen and Britannica Educational Publishing 2011) recognises that they are best suited topographies that feature steep hills and where there a no service issues related to the fixed speed of the vehicles that are a result of the continuously moving cable that is gripped in order for the cable car to move. These systems also had high capital costs and were not used at all in Canada (Ball 1991).

The development of expertise in the harnessing of electricity to drive motors, along with the roll-out of associated distribution infrastructure could be argued to be key technological advances that lead to the development of electric street cars. Hardware, such as tracks and cars, would have already been supported as part of the development of railways. Of the multiple system architectures under development in the late 1880’s, that of Frank Sprague, implemented in Richmond, Virginia became the “dominant system” (Ball 1991).

As an evolution of horse-drawn systems, the first electric street cars were similar in terms of size and passenger comforts. The mass that a horse (or horses) could pull dictated the design of those cars. As it was realised these constraints no longer applied, street cars became longer and more luxurious with seats and enclosed passenger accommodation (Ball 1991). Given winter temperatures in Albertan cities this maturing in streetcar technology would have made streetcar systems more desirable to local communities.

Local Adaptations and Growth
Ball (1991) also notes that in Montreal, street cars were fitted with “storm windows” and snow catchers. It is reasonable to assume that such features would have been available to Albertan streetcar operators.

Innovations in ticketing were introduced to support streetcar operations. Ball (1991) describes the introduction of PAYE (Pay As You Enter) ticketing to support faster boarding of passengers. Presumably, prior systems involved the purchase of a ticket prior to boarding the vehicle.

Electric streetcars initially provided a replacement for their horse-drawn predecessors. Particularly relevant for the smaller cities of Alberta, streetcars enabled the implementation of a viable public transport system (Ball 1991). The development of commuter suburbs where residents lived outside away from an urban core that was the focus of employment was enabled by streetcars. This supported other environment changes that included growth of the Canadian economy, increased settlement in areas of lower density along with the development of natural resources (Stelter and Artibise 1982).

The birth of streetcar systems in Alberta was characterised by the role of municipal government. Calgary and Edmonton launched systems after eastern cities and they were owned by local government (Ball 1991). This means it is difficult to identify precursor models in an Albertan context and explicit policies that governed the operation of streetcar systems were most likely adapted from those in eastern Canadian cities or the United States.

One policy that likely contributed to accelerated growth of streetcar systems in Albertan cities was that related to taxation. Prior to World War 1, the prairie cities had an almost total dependence on land taxes (Stelter and Artibise 1982). This meant that the development of land to become more economically useful would generate higher taxes for municipalities. Streetcars were a key to unlocking the value of land further out from urban centres.

System Growth and Prairie Politics
The growth of streetcar systems in Albertan cities was driven through the private and public realms. The larger macro forces that lead to the development of suburban estates required streetcar systems, in the absence of suitable alternative modes, to be viable. In fact, there were middle-class expectation of streetcar service to a new development, in a similar way to electricity, water, sewerage and made roads (Harris 2020). Market demand, felt through private land developers created pressure for the continued expansion of streetcar systems to service ever more distant land subdivisions. Subsequently, there were decreasing returns on further investments in streetcar systems (Stelter and Artibise 1982).

The relationship between the growth of streetcar systems and the cities they served was impacted by the objectives of municipal authorities. Stelter & Artibise (1982) posit that cities in the Canadian Prairie States, including Alberta, had their growth shaped by “prairie urban elites” who spanned the private and municipal sectors within tight social circles. As a result, it was easier to seek and obtain consensus for any new initiative, such as the establishment and extension of streetcar systems. Additionally, there was competition between regions, with a desire for urban growth and consequent status. The result was the continued growth of streetcar systems in areas of low density when compared to eastern Canadian states. For example, in 1921, Edmonton 2.2 people per acre compared to 19.2 for Montreal and 31.5 for Toronto (Stelter and Artibise 1982).

Maturity and the Rise of the Automobile
In some ways, the success of streetcar systems in facilitating the development of suburbs around an employment core provided a niche that could also be served by private automobiles. However, increasing automobile use lead to congestion in urban cores, slowing streetcars (Ball 1991). This furthered the relative advantages of private automobile transport, leading to more congestion. Effectively, a negative feedback loop was in operation during this transition.

Some streetcar operators in Canada began to run buses (Roberts, Meadowcroft et al.). It is unclear to what extent this occurred in Albertan cities and to what extent these were attempts to supplement or replace streetcar systems. In any case, it showed that there was a viable alternative that was not reliant on dedicated infrastructure, only what shared with cars. Additionally, buses overcame some of the limitations of streetcars in colder Canadian climates. In Calgary, streetcar accidents were attributed to reduced grip of wheels on rails during cold weather (Stark 2015).

A Future for Streetcars in Albertan Cities
Any re-establishment of streetcars in the form of a system based on modern light rail vehicles would best be served by consideration of the factors that lead to their decline. Where possible, dedicated lanes or rights of way could be utilised to reduce the impact of urban road congestion on average speeds. This is additionally important because of the  regulatory environment has developed around car use (Roberts, Meadowcroft et al.). Risks of accidents, assuming modern technology has no resolution for cold weather grip issues, would also be reduced in a more isolated operating environment. Particularly relevant in the cold-climate cities of Alberta, private automobiles offer residents a high degree of comfort door-to-door when weather-protected parking facilities are available. For modern streetcar to be competitive they should provide sheltered connections to controlled indoor environments, such as Calgary’s “Plus 15” network.

Source Data
Total historical length for each system was sourced from Street Railway Investments (1894-1910), McGraw Electric Railway Manual: The Red Book of American Street Railway Investment (1911-1914) and McGraw Electric Railway Directory (1917-1920). The Albertan cities identified as having streetcar systems were Calgary, Edmonton, Lethbridge and Strathcona. There was a single entry for Strathcona in 1909, as the system was taken under the administration of Edmonton in the following year. For this reason, Strathcona was not included in the analysis. Observe from Table 1 that data was not available for the years of 1915 and 1916.

Method
Curves were fitted using a three-parameter logistic function.

S(t) = Smax/[1+exp(-b(t-ti)]

Where:


 * S(t) - Modelled system length in a given year, t.
 * Smax – Maximum length that each system reached
 * ti – Time period at inflection point (where half of the maximum system length was reached).  This was estimated using initial plots of the base data.
 * b – This was estimated by using the MS Excel Goal Seek function across a calculated set of system lengths at various values for t. Goal Seek maximised the value of R-squared, which had been calculated by using the MS Excel RSQ function.

The output of the three-parameter logistic function was calculated and charted for each of the Calgary, Edmonton and Lethbridge streetcar systems. Additionally, an Alberta-wide analysis was performed using the sum of total length for each system across all years.

A regression was then performed using the MS Excel Regression function to determine the t-Statistic and R-squared values. Key parameters and statistics are summarised below. Plots show actual and modelled system lengths were generated in MS Excel.

Calgary
It can be observed from the plot that the birth phase of the Calgary streetcar system occurred approximately in the years 1909 and 1910. Over a slightly longer period (1911 to 1913) growth occurred. The model provides a plausible projection of the final years as the system matured and maximum length was reached. The model makes no allowance for the start of the decline apparent in 1920. While the model appears to be reliable at predicting streetcar system length in Calgary from a visual inspection of the chart and also considering the high R-squared value of ~0.98, the low t statistic of ~0.02 at indicates no confidence in the value of b at the 95% level.

Edmonton
In contrast to Calgary, the Edmonton streetcar system experience a longer period of birthing (4 years) where system length fluctuated at around 20 miles from 1908 to 1912. The growth period is extremely short, with system length increasing to the maximum of nearly 80 miles in less than a year. While system length dropped from 1914 to 1917 it is difficult to determine if this was part of a continuous period of decline or a one-off adjustment from the maximum system length. When compared to Calgary, the R-squared value (~0.87) is lower and the t statistic (~-1.24), while larger than that for Calgary is not significant at the 95% confidence level. This reinforced by the obvious variance between the actual and model predictions early in the system life from 1909 to 1911, along with the drop been 1914 and 1917.

Lethbridge
The Lethbridge system is unusual in that it never increased in length further than the initial 10 miles. In this context there is no relevance in commenting on the birth, growth or maturity of the system.

Interestingly, this results in the model generating an R-squared value extremely close to 1, however still with a t-statistic well under 2 and therefore not statistically significant.

Alberta (combined)
The length of systems in the state of Alberta is primarily affected in equal parts by those in Calgary and Edmonton that both reached maximum lengths of between seventy and eighty miles. This is reflected in each growth phase. Birthing appears to have occurred between 1909 and 1911, followed by growth over the next two years. Maturity lasted at least from 1914 to 1913. Given the available data it is not possible to determine if the drop to 2020 was a significant point in the decline of systems across the state. Still with a relatively high R-squared value of 0.97 the model is a good fit for the actual state-wide system length but does have a t statistic of -2.75 indicating significance at the 95% confidence level.

Conclusion
Growth in the length of individual streetcar systems appears to be sporadic, perhaps affecting the ease of generating models at the 95% confidence level. Consolidation of data produced a statistically significant model, but not one that is useful for predicting all stages in the lifecycle of a streetcar systems in Alberta as a whole. The state-wide model does not generate points that are related to an obvious birthing phase, with only a handful of useful points potentially related to a growth phase. An extension to this analysis could be the exploration of different models that yield more useful results.