The Four Horsemen of the Production Apocalypse
Gary Becker begins his “A Theory of the Allocation of Time” (Becker 1965) with a discussion of time spent at work relative to other activities, highlighting the need for more robust analyses of uncompensated time use in economics. Little did he know that his theory and model structure would strongly influence travel behaviour and economic analysis over subsequent decades. In the 1970s, DeSerpa (1971) and Evans (1972) provided among the first connections to transportation analysis. Building from DeSerpa and Evans, Jara-Díaz (2003) completed fundamental work on the value of time in transportation economics. However, there is a wealth of other insights relevant to transportation analysis to be gained from Becker’s theory. I will focus on the topic of transportation service provision.
Being a founder of the neoclassical “Chicago School of Economics”, Becker would certainly have strong opinions on the extension of his theory to include other non-market means of production. Nonetheless, one can consider four basic forms of production: market, home, government, and commons. Market production is the exchange of a good or service for money. Home production, as defined by Becker, is uncompensated production by the household (e.g., making dinner, cleaning the bathroom, etc.). The latter two forms of production were not discussed by Becker in his 1965 paper. In the first form, government produces a good or service for use by the public, financed by tax revenue or similar sources. For example, most healthcare services in the United States are privatized and subject to market forces, whereas many other nations provide healthcare as a public service. The final form of production is in the commons. Neoclassical economic theory suggests that common resources will be exploited if not managed by state intervention or market forces via monetization. However, both Mariana Mazzucato (2018) and Elinor Ostrom (1990) provide compelling arguments for the efficient use of common production. As one example, Wikipedia is managed in the commons as most contributors provide their services at no cost.
The reader may wonder: what does any of this have to do with travel behaviour? I argue that the above economic production framework can be mapped onto the transportation mode choice problem and provides valuable insights for behavioural theory and modelling methods. Home production of transportation takes the form of walking, cycling, or using one’s own vehicle. The vehicle and fuel must be purchased on the market, but the transportation service is provided by the household using personal time as the only input to production. In contrast, taxi or ridehail modes are purchased in the market from private firms (e.g., Uber, Lyft, Didi, etc.). Government provides transportation as a public service through transit. Transportation provided in the commons is less common (a nice little pun) but takes the form of carsharing and other systems wherein the service cost is shared among a pool of otherwise unaffiliated households. Each of the above transportation alternatives could manifest as the same physical mode – i.e., a small passenger vehicle. However, the behavioural and operational models differ. For example, a person may be unable to afford market production (ridehail). They may not use public production (transit) to avoid sharing a vehicle with strangers. They may be unwilling to participate in the commons (carshare) because they don’t want someone’s muddy boots in their car.
Being trained as a civil engineer, I am always keen to find concrete (two puns in one article is the limit) applications for theory. The modelling of autonomous vehicles (AVs) provides an application of the above theory to transportation. Many scholars make the distinction between personal AVs, shared AVs (or SAVs) provided as a public service, and SAVs provided in the market. However, in my reading, there is minimal thought given to how these operating models would function from an economic production theory perspective. Let us first focus on the first alternative. It is unclear whether to classify a personal AV as home production or market consumption – Is sitting comfortably in a vehicle while it performs the driving task pure leisure? Are you purchasing driving services in the market as a software package? The latter two alternatives provide an interesting contrast. In both cases, SAVs are provided to the household as a service. However, if we consider transportation as a public service, another economic concept may be brought to bear, termed private provision of a public good by Bergstrom et al. (1986) . Public transit provides a subsidized travel option for low-income households and those who choose not to own a personal vehicle. The widescale deployment of SAV fleets would therefore require ridehailing operators to take on this role. The risk is that each operator may rely on other operators to serve low-income households, with the result being a Nash equilibrium that does not maximize total social utility. Government intervention is necessary to ensure an equitable distribution of mobility through incentives or regulations (Sieg 2020). Existing simulations often ignore the profit-seeking motives of private operators, assuming they match trips to minimize detour time and are agnostic to traveller income and profit maximization motives on the part of operators.
As a final thought on the historical distinction between private and public travel modes. The marginal private automobile trip tends to add travel time for existing travellers through congestion that slows travel speed and increases its variability. In contrast, the marginal transit trip tends to improve service for existing travellers through reduced vehicle headways and the use of higher-order transit (i.e., trains rather than buses). The latter outcome is termed as an anti-congestion effect (Mohring 1972). The trend in SAV research is toward small, automobile-style, vehicles that may be shared or used alone. This shift adds shades of gray to the public-private dichotomy. Will shared mobility exhibit the congestion effect of fthe private automobile or the anti-congestion effect of transit? Importantly, in which contexts do these new forms of mobility tend to exhibit each of these effects (i.e., there is likely a spectrum of possible impacts depending on the application of technology, local built form, and transportation infrastructure conditions)?