Part II-B - Applications of Tomorrow

11 Sharing in Future Electric Energy Systems

Michael Kane , Elizabeth Allen , Yutong Si , and Jennie C. Stephens

11.1 Introduction

Engineering advances have been opening new possibilities for sharing electric energy. Technological and social innovations in the electric energy sector may allow consumers to become more actively engaged in producing and managing the generation, distribution, and use of their electricity, which could shift the locus of organizational decision making and control away from traditional utilities. These innovations also have potential to diversify and restructure who is included and excluded from energy sector benefits (Stephens, 2020).

The transition toward a “sharing economy” in the transportation and lodging sectors, and in other emerging sharing economy systems, can be understood as a process of separating rights of use from the other rights of ownership for goods that had previously not been as easily divisible in this manner. For example, app-based short-term lease platforms have disrupted the lodging sector by enabling owners to create new value by more easily assigning the rights of use of their dwellings, and consumers now have new options of affordable unique rentals as an alternative to owning a vacation home. In the current traditional deregulated energy system, ownership of energy generation infrastructure is centralized along with allocation of rights of use. Generally, consumers did not own energy assets and had little power to consume the type (for example, traditional vs clean and renewable) or price (for example, fixed rate, time-of-use) of energy that aligned best with their economic and social motives. Sharing economy ideas combined with new technology promise to decentralize energy generation thus increasing ownership and opening broader markets (for example, demand response, community generation, and resilient microgrids).

Sharing economy innovations in the electric grid, including community solar and energy blockchain systems, are expanding the role of assets owned by consumers. These innovations are transitioning the role of generation and management from large corporations and utilities to consumers. This restructuring has potential to democratize energy systems depending on how policy and regulations guide the development of a more distributed renewable-based society (Stephens, 2019). As households and businesses become prosumers, they are producing and sharing surplus energy with the grid and other users, creating what some are referring to as “locavolts” linked in microgrid systems to enhance electric energy resilience and to leverage local financing. Prosumers who can “share” their electricity may also be empowered to change the rules that have governed their relationships with utilities for the past century.

At the same time, technological innovations related to renewable electric energy generation, distribution, and demand-side management may enable new types of energy sharing with the potential to disrupt and transform the current electrical energy industry. In the same way that Uber and Airbnb have disrupted existing transport and hospitality industries, we suggest that the electricity energy sector may also be trending toward large-scale system reconstruction by way of innovations for energy sharing. In recent years, one outcome of these energy sharing innovations has been the rise of a new group of energy platform operators (for example, thermostat companies and other third parties) acting as “energy services providers” that position themselves as the electricity sharing economy platforms facing consumers. Many have identified the “death spiral of the utilities’ business model” in which the traditional fee for energy delivery that utilities charge to customers fails to provide adequate revenue from prosumers who use the grid only for short periods of peak power (Felder and Athawale, 2014). These periods of peak power, for example, in the early evening when the solar resource diminishes and residential demand increases, drive infrastructure costs. Energy platform operators may enter the power sector as new market players that aggregate small amounts of distributed energy resources (DERs) provided by each user into megawatts that can be traded in the traditional regional energy, capacity, and ancillary services markets.

Future energy systems are likely to integrate a regionally appropriate mix of renewable electricity generation that is dispatched, stored, and distributed through platforms that enable sharing at multiple scales, from milliseconds to decades (Stephens, 2019). The potential for sharing in electricity systems is expanding as energy systems around the world are transitioning from centralized large fossil fuel power plants to distributed renewable-based power generation at multiple locations (Stephens, 2020). Such flexibility in time scale and locations is critical, as the success or failure of any organization or individual operating in the sharing economy is determined by its ability to effectively manage an elastic match between resource supply and demand. Sharing in future electric systems has the potential to disrupt relationships governing utilities, energy consumers, and distributed electricity generation at the individual and household levels, at the community and organizational levels, and at the regional, state, national and even international levels.

This chapter reviews the range of institutional models for electricity generation and distribution; considers how electrical energy systems are undergoing a transformation toward more distributed, renewable-based configurations; and explores the multiple evolving mechanisms for energy to be shared among generators, distributors, and consumers. We present a typology of sharing economies in electric energy systems and identify the technological and sociopolitical potential for sharing electrical energy as systems evolve. The chapter also explores policy considerations that flow from such large-scale transformations in the energy sector. The chapter concludes with opportunities for future research and questions about the implications of energy sharing for energy justice.

11.2 Background on the Electric System Structure, Business Models, and Regulatory Frameworks

Over time, a wide range of institutional models, shaped by both technological and sociopolitical factors, have emerged for provisioning energy for households and industry (Stephens et al., 2015). Regulatory frameworks have evolved in parallel with transformations of technology and business models. It is important to understand how these transformations evolved over the last century, and the sociotechnical factors that led to them, to understand the disruptive potential of sharing economy innovations in the electric power sector.

Traditional utilities are structured as regulated monopoly providers of electrical power and grid services for a specific geographic area (Kiesling, 2014). The role of utilities is to deliver power to satisfy consumer demand, which may require the utility to generate, transmit, store, and/or purchase power and regulate its frequency and/or voltage. Early on, utilities faced steep barriers to market entry, because power plants and transmission lines were very expensive. Consequently, when household electricity was introduced in the United States in the early 1880s, it was available only in larger cities for wealthy homeowners and almost 90 percent of rural homes were without electricity in the first decades of the twentieth century (Velaga et al., 2019). Over time, growth in household energy demand drove down the average cost of delivered power. This combination of economies of scale and high barriers to entry led to the three dominant utility business models that operate in the United States today: (1) state-operated not-for-profit consumer-owned electric cooperatives, (2) local publicly run or managed utilities (POUs), and (3) investor-owned utilities (IOUs). Although there are relatively few investor-owned utilities, they tend to be very large, serving over 75 percent of utility customers nationwide.

Across the world, electric power markets operate in a variety of forms that lie on a spectrum from monopolization to democratization. The evolution of energy systems and markets around the world has been very heterogeneous. In some countries, deregulated markets were established through the restructuring of some of the existing utilities, with the intent to disrupt energy generation monopolies by increasing competition. In deregulated electric grids, electric energy, capacity, and reserves are each traded on a wholesale market, which must meet customer demands while satisfying any constraints of the transmission system (for example, line limits). Finally, in energy choice systems, energy users have the opportunity to select among multiple energy providers based on rates and energy generation options (see Figure 11.1).

Figure 11.1 Three different systems for electricity generation, transmission, and distribution.

In the twenty-first century there have been considerable advancements in technologies for household electricity generation with renewables. These affordable distributed generation technologies are beginning to challenge the stability of the centralized electricity system (McKenna, 2018). There is a growing embrace of distributed energy resources enabled by new information and communications technologies such as blockchain technology, which enables trusted peer-to-peer distribution; and internet-of-things (IOT) devices, which enable real-time response to changes in supply and demand. These technological changes have precipitated the rapid growth of sharing economy businesses (Hamari et al., 2016). A primary argument in favor of the sharing economy model is that collaborative effort among participating actors leads to more efficient, sustainable, and resilient outcomes than centralized decision making (for example, by utilities).

Continued global population growth and resource-intensive consumption practices have increased energy demands, contributing to the global climate crisis. In response, a growing number of consumers are beginning to look for opportunities to participate in alternatives to the current electric energy systems that promise improved efficiency, reduced fossil fuel emissions, and more local control over energy system decision making. There is tremendous disruptive potential associated with sharing economy innovations that will enable a broader group of actors to produce, transmit, store, and consume electric power with each other.

11.3 Taxonomy of Electric Energy Sharing

In this section we explore key trends in sharing of electric energy, looking at forms of sharing that have emerged based on market structures and services being provided.

At face value, electric energy may appear to be a perfectly fungible commodity that can be produced by a variety of methods, transmitted across nations, and consumed and billed through meters. In reality, electric energy markets face unique challenges compared to most other sharing economies, in that the product being traded is not discrete (compared with a ride between two points provided by a ride-hailing service or two nights in short-term rental), cannot be easily stored (generation must match consumption at every instant in time, unless expensive batteries or other storage systems are utilized), and transferring the commodity requires specialized infrastructure (for example, wires and transformers) that have limited capacity (for example, circuit breakers on distribution lines can trip if a large load overheats a transformer).

To facilitate the discussion of the sharing economy in the electricity sector, it will be helpful to classify various types of “sharing” by the market structure: Top-down, in the form of a vertically integrated utility, or bottom-up, where individuals govern the market forces. Additionally, it is important to identify what is being bought and sold in different arrangements of energy sharing: Energy (for example, kilowatt-hours), ancillary services (that is, voltage and frequency control, generator dispatch, operating reserves), storage, and/or transmission.

Figure 11.2 represents a framework and taxonomy of sharing economy approaches in the electric grid along two dimensions – market structure and services provided. The services encompassed by a sharing economy may only include energy, or all of the services required to operate an electricity system (including energy, voltage and frequency control, and asset management). Along the other dimension, the prices and availability of services may depend on the market structure.

Figure 11.2 Taxonomy of sharing economy approaches in the electric grid.

In traditional utility-control approaches to energy sharing, consumers do not generate electricity; instead, they provide value through their flexibility to change demand. This is realized through demand response programs that ask users to modify behavior, or automatically adjust electric loads, for example, by turning off large energy loads such as air conditioners during times when the grid might overload. Similarly, time-of-use pricing enables users to shift their energy use to the times when low-cost, sometimes renewable, energy is available and can be easily transmitted to the consumer. However, such demand-side management approaches have limited capabilities since electricity demand is relatively inelastic to price.

Recent advances in regulatory structures and technology allow consumer choice in which consumers can choose which wholesale provider they buy their power from or generate power themselves. They then sell the excess energy (calculated over hourly, daily, or monthly periods) back to the grid through the wholesale power agreement of their distribution utility. Energy choice legislation in the form of the Energy Policy Act of 1992 enabled consumers to choose who they purchased their energy from, even specifying that only renewable energy be provided. As far back as 1983, consumers with the ability to produce electricity, primarily through rooftop solar, could share any excess, and receive the market value of that energy through net metering.

More recently, peer-to-peer energy sharing uses decentralized ledgers (energy blockchains, such as LO3 Energy) that track energy consumption, storage, and generation through a digital platform and enable users to choose who produces or consumes energy, and when, how and for how much the electricity is produced or consumed. The challenge in the future of the electric energy sharing economy is how to share all aspects of operations, safety, and management of the electric energy system.

Community microgrids promise to realize this future through breakthroughs in automation, machine learning, and digital platforms with both digital and physical aspects that enable sharing of distribution systems, or a fully shared grid. Community microgrids rely heavily not just on shared generation, including large amounts of renewables, but also on shared services through distributed energy resources (DERs). These DERs may include energy storage in parked electric vehicles, smart thermostats that run only when energy is available, and smart solar inverters that can help stabilize voltage and frequency on the shared wires. This sharing of the responsibility to balance supply and demand reduces costly peaks and transmission costs and creates a more resilient electric energy supply that can continue operating even when the larger grid suffers a blackout (Poudel & Dubey 2018).

The following sections describe in more detail these four approaches to sharing economies in electric energy systems, moving clockwise around Figure 11.2 from the status quo utility control to future shared grids.

11.4 Utility Control – Demand Response and Time-Of-Use Pricing

Utility control describes top-down (centralized) management of grid services and reserves. In this arrangement, utilities are responsible for all grid services, including managing capacity of transmission lines and ancillary services that may include energy storage or controlling the readiness of power plants to begin producing energy to match demand.

Demand response innovations give some responsibility to energy consumers to help manage the electric grid. For example, during a time of high energy demand that might historically have led to a temporary partial shut off the grid (brownout), consumers can instead respond to signals from the energy provider and suspend operation of nonessential appliances, such as air conditioning units or pool pumps. This enables energy service to be maintained for essential functions and reduces total peak electrical load, reducing the need for costly increases in electric grid capacity. Such demand response (DR) may be fully automated with smart appliances (such as Google Nest) that adjust operation schedules based on signals from the utility or can simply take the form of utilities sending customers texts or other communications to encourage reducing energy use at specific times to avoid overloading the grid (for example, “Shave the Peak!”).

Time-of-use pricing involves changing the electricity rate based on current demand, more closely reflecting real-time energy prices on the wholesale market. For example, customers might be charged a much higher rate to use energy during times of the day when aggregate demand is at its peak. This time-of-use pricing incentivizes consumers to adjust their energy consumption for more efficient operation of the grid. Time-of-use pricing enables users to shift their energy use to the times when low-cost, sometimes renewable, energy is available and can be easily transmitted to the consumer.

Since electricity has become an integral and necessary part of modern life, electricity demand is largely inelastic, therefore limiting the impact of pricing policies on demand (Lee and Chiu, 2011). Furthermore, price elasticity is nonlinear and asymmetric, complicating the efficient implementation of such dynamic pricing approaches (Haas and Schipper, 1998).

11.4.1 Role of the Users

In these examples of utility control energy sharing, innovations enable consumers to share the value of their flexibility to change aggregate demand in real time. Both demand response and time-of-use pricing can work automatically or by active participation of consumers. Consumers can potentially benefit from sharing by avoiding situations where electrical power would otherwise be interrupted and by reducing their electrical energy bill by adjusting their behavior based on signals from the utility service provider. These utility control sharing mechanisms can be a considerable benefit to customers who have the flexibility and information access necessary to vary their energy usage in response to market signals. However, since the average household spends only 2 percent of its income on electric energy, and various consumption-specific social and behavioral factors make household-level flexibility difficult, the elasticity of energy consumption is low (Drehobl and Ross, 2016; Wilson and Dowlatabadi, 2007). Further, as low-income households may spend up to 30 percent of their income on electric energy, demand response and time-of-use pricing may place an inequitable burden on the already disadvantaged if they lack technology or flexibility necessary to adjust energy usage in response to market signals (Drehobl and Ross, 2016).

11.4.2 Role of the Platform

For utility control of energy sharing, the utility, whether a private company, consumer-owned cooperative, or state-owned public company, functions as the platform for coordinating all sharing activities. Alternatively, demand response providers may be third-party commercial entities that aggregate the load flexibility of their customers into bids for the wholesale market. This motivates these third parties to help customers with strategies or technology to adjust their electricity consumption in response to market signals.

11.4.3 Notable Examples

Most US utilities offer both commercial and industrial customers options for centralized energy sharing in the form of demand response. Each of the nation’s independent system operators/regional transmission organizations (ISO/RTOs) sponsor demand response programs (Department of Energy [DOE], 2019). The Federal Energy Management Program (FEMP) helps federal agencies and other organizations identify opportunities for energy project incentives and demand response programs, execute these opportunities, and fully capture their benefits (DOE, 2019). Most utilities offer time-of-use pricing options as well: These program structures include simple time-of-use rates, where prices change at set times through a 24-hour cycle to reflect afternoon peak electricity use, overnight off-peak hours, and the shoulder periods in the morning and evening hours. Other variations of time-variable pricing include real-time pricing in which customers’ rates reflect the wholesale electricity market or the utility’s cost of production; day-ahead hourly pricing, where the utility sets prices to reflect the cost of acquisition or production for the coming day; and block-and-index pricing, in which the customer can lock in set energy prices for part of their energy use and pay current market price for additional usage.

11.5 Consumer Choice – Energy Choice and Net Metering

In the context of top-down (centralized) management of energy, without the complexity of real-time control of the grid, various forms of consumer choice energy sharing are possible on hourly, daily, or monthly timescales. These are distinct from what we have described in the previous section in that they engage the consumer in financing distributed energy resources (DERs), either on-site or remotely. DERs are decentralized because they rely on energy sources other than the utility to provide energy, and such sources can and often do include consumers themselves. DERs are not an innovation whose primary purpose is to manage constraints on the electrical grid. Instead, their primary benefits are to allow prosumers to buy energy from their preferred source (for example, renewables), reduce their bills through efficiency upgrades, and/or sell excess renewable energy they generate back to the grid at favorable prices.

Energy choice programs enable consumers to choose who they buy their energy from. For example, they might pay a premium for renewable electricity from wind turbines in another state. While this choice doesn’t directly affect the “source of electrons” transmitted to the household user, the arrangement directly finances clean energy production. In these wholesale markets, consumers are buying from a corporate producer of energy, such as a large wind farm aggregator or financial entity specializing in trading energy.

As far back as 1983, consumers with the ability to produce electricity, primarily through rooftop solar panels, could share any excess, and receive the market value of that energy through net metering. In net metering, consumers have the opportunity to produce energy locally and use their energy first, selling any excess back to grid, typically averaged over the day or the month. Sometimes energy sales back to the grid are managed on a yearly basis, and very rarely on an hourly basis. These systems give consumers the option to choose renewable energy for their own consumption and reduce dependence on the utility provider, thus leading a transition toward renewable energy generation. Electricity consumers who generate their own energy on-site, sell a portion of that energy back to the grid, and store energy or otherwise control their load, are called prosumers. Figure 11.3 depicts a generalized scheme for how energy choice and net metering systems interface with traditional utility businesses.

Figure 11.3 How energy choice and net metering systems interface with traditional utility businesses.

Adapted from Potter, 2019.

11.5.1 Role of the Users

The act of producing energy locally and sharing it back to the grid transforms “energy consumers” to “energy prosumers.” A growing body of research explores prosumers’ behavior and interactions with the electrical energy system. A survey of prosumer perceptions found that prosumers are motivated to collaborate with utilities to contribute to the societal goal of creating more renewable energy (Silva et al., 2012). Information sharing and communication innovations are key to the successful management of energy sharing in prosumer engagement systems (Zafar et al., 2018). However, since prosumer systems that require the individual to own an asset (for example, rooftop solar) are in general limited to those individuals that own their home, renters and other low-income citizens find it difficult to benefit from prosumer engagement, aside from energy choice.

11.5.2 Role of the Platform

In the case of the prosumer systems based on energy choice markets and net metering, traditional utility companies maintain all physical hardware and assets of the energy distribution system and real-time controls, while prosumers begin increasing their share of ownership or investment in energy generation assets. Energy choice and net metering programs do not specifically address the challenge of managing the electrical load on the grid to match supply with consumer demand. A growing body of research focuses on the problem of peak shaving, that is, reducing the maximum points of electricity demand. An index termed the sharing contribution rate (SCR) quantifies users’ contributions to energy sharing and peak shaving (Wang et al., 2019).

11.5.3 Notable Examples

Globally, a broad array of programs for energy choice and net metering are emerging. In the United States, over a dozen states have deregulated electricity markets, meaning that electricity providers can compete to sell energy to consumers. One example of a platform that provides individual energy consumers with the opportunity to invest in solar energy projects is Mosaic, a solar lending platform. Traditionally, only individuals with sufficient starting capital could invest in household solar panels. However, new technologies have improved the efficiency and lowered the cost of manufacturing, installing, and permitting solar energy. Solar lending companies enable consumers to become prosumers by financing their initial solar panel installation, and then collecting back the loans over time with interest as these prosumers sell energy back to the utility. New platforms have emerged to connect investors with borrowers seeking financing for solar power projects.

11.6 Peer-to-Peer – Community Solar and Blockchain

Recent advances in regulatory structures and technology have enabled community solar programs in which solar panels or other DERs are installed on shared land, while dividends for the energy produced are paid out to the community. Even more recently, energy blockchains have emerged, functioning as decentralized ledgers that track energy consumption, storage, and generation through a digital platform and enable energy consumers and producers to choose when, how, by whom and at what price their electricity is purchased and sold.

Community solar and blockchain-based energy sharing are examples of peer-to-peer electricity sharing that integrates decentralized management of energy systems and does not directly address the technical considerations associated with real-time control of the electrical grid. In these systems, transmission lines and distribution infrastructure are still managed by the traditional utility. Because there are substantial costs to operating the grid, such as controlling voltage and managing storage, electricity consumers in these types of peer-to-peer arrangements continue to pay utilities for use of their infrastructure (but not for the generation of electricity).

11.6.1 Role of the Users

When prosumers and consumers trade self-produced energy in a peer-to-peer manner, they can both profit, and this could provide incentives for different kinds of investments. Previous research indicates that compared to traditional peer-to-grid models of energy sharing, peer-to-peer electricity sharing may confer economic benefits to consumers and environmental benefits in the form of carbon footprint reduction (Secchi and Barchi, 2019).

11.6.2 Role of the Platform

In peer-to-peer sharing, the utility still manages (although, may not own) the physical assets to satisfy real-time needs. Recent research explores the potential for blockchain-based microgrid energy markets that do not require central intermediaries to facilitate transactions between producers and consumers, thus reducing costs associated with these traditionally labor-intensive markets.

11.6.3 Notable Examples

Examples of peer-to-peer sharing include community choice aggregation programs, as well as more sophisticated electricity trading platforms. One example of an electricity trading platform is Vandebron, a startup in the Netherlands that allows people to buy energy from independent producers (Vandebron, 2020). The mission of this company is to connect people who have surplus renewable energy with energy consumers who are not producing their own electricity. Utilities are not involved in the transaction. Functioning as an “Airbnb for electricity,” this platform allows energy consumers to search for producers. For example, a consumer could share their information on the platform to arrange for the purchase of their power from a farmer in a neighboring community who has wind turbines or solar panels in their fields. Such business models may be easier to implement in deregulated energy markets, such as the Netherlands, in contrast to other countries and many states in the United States (Schiller, 2014).

While the case has been made that blockchains are an effective technology to decentralize the energy system, to date there are only limited examples of blockchain-based, local, peer-to-peer energy sharing operating in practice (Mengelkamp et al., 2018; Noor et al., 2018). The Brooklyn Microgrid project (BMG) is one such example (Brooklyn Microgrid, 2020). Note that this example differs from true shared grid systems described in detail later, because while consumers are able to trade and share energy with fellow consumers in the BMG project, the energy distribution infrastructure is owned and managed by a centralized utility. The network connects prosumers and consumers by enabling people to buy and sell locally generated renewable energy through a Brooklyn Microgrid mobile app, which gathers and records energy data for users. Through blockchain technology, BMG developed Exergy, a data platform that creates localized energy marketplaces for transacting energy across existing grid infrastructure. The BMG also acts as an educational and community engagement initiative: For example, the organization facilitates workshops in partnership with public schools in New York City to teach community members about the technology behind electric microgrid systems and encourage broader participation in energy sharing.

11.7 Shared Grid – Community Microgrids

When shared ownership of electricity generation infrastructure is combined with shared ownership of distribution and energy management infrastructure, these systems can be classified as a shared grid. Shared grid systems are characterized by decentralized management of real-time energy generation and advanced grid controls, made possible by breakthroughs in automation and machine learning. They promise improved physical reliability and resilience to cyber threats, opportunities to improve energy system sustainability, reduce reliance on fossil fuels, and improved economic efficiency. These innovations help scale and reduce the costs associated with the hundreds of people required to run a regional grid, to a handful of contractors to operate a microgrid.

Community microgrids rely heavily not just on shared generation or renewable energy, but also on shared grid management services provided by distributed energy resources (DERs). These DERs may include energy storage in a parked electric vehicle, smart thermostats that run only when energy is available, and smart solar inverters that can help stabilize voltage and frequency on the shared wires. Thus, DERs represent a pathway for sharing the responsibility to balance supply and demand among all participants in the system. When integrated into a community microgrid system, these technologies have the potential to reduce costly peaks in energy demand as well as transmission costs. Many researchers also argue that community microgrids are a viable solution for more resilient local electric energy systems that can continue operating even in instances when the larger national or regional electric grid suffers a blackout (Jiménez-Estévez et al., 2017; Marnay et al., 2015; Wu et al., 2016).

11.7.1 Role of the Users

In a shared grid system, all users contribute to the cost of grid management. There are various ways that such cost sharing might be structured in real-world shared grid systems. For example, third-party platforms, decentralized algorithms, and/or self-governed user decisions can determine how the resources are managed.

11.7.2 Role of the Platform

In shared grid systems, the role of a traditional utility significantly diminishes. When energy generation, assets, and management of the grid are controlled by third-party platforms and decentralized algorithms that integrate a system of DERs, electrical energy systems can potentially function without any involvement of a centralized utility. A market of third-party platforms may even operate simultaneously in the same way that Uber and Lyft both operate in the same city. This future vision for shared grid systems requires, however, new technologies for information sharing between users in the system. For example, the algorithms needed to control millions of household-scale renewable energy generators and the load at millions of homes is very different than the algorithms that control load and generation in the traditional centralized generation system.

In addition to developing information sharing technologies to facilitate operation of shared grid systems, some scholars regard resource optimization technologies to be critical for continued innovation of smart grids. For example, Zafar et al. (2018) discuss linear and nonlinear optimization programming in the context of prosumer-based energy management and sharing. Other researchers present the optimization technique in a two-stage approach. For instance, Cui et al. (2019) argue that the bi-level optimization problem could be transformed into a single-level mix integer linear programming problem through proper linearization techniques. In the second stage, an online optimization model is proposed for each prosumer to make the energy schedule according to the latest system situation and prediction error (Cui et al., 2019). Alternatively, Long et al. (2018) present a nonlinear programming optimization in the first stage with a rule-based control to update the control set-points through real-time measurement in the second stage.

11.7.3 Notable Examples: Community Microgrids

To date, most examples of shared energy generation and grid management can be found in institutional microgrids. The best examples of shared grid systems can be found at US military bases, where there are systems in place for coordinated management to govern when and how system users produce and consume energy (Hossain et al., 2014; Prehoda et al., 2017). Shared grid examples also exist in hospitals and corporate landscapes. For example, East Boston has a microgrid network made up of institutions that share renewable electricity generation and backup generators among the buildings (Sheehan, 2015).

Looking ahead toward the possibility of more shared grid systems being implemented throughout the US and globally, there is considerable potential for increased production of renewable energy and efficient management of supply and demand enabled by DERs and algorithms for real-time grid management. However, while these innovations suggest potential benefits of the sharing economy for the electric energy sector, there are important access and equity considerations to be addressed. In a future where corporations and wealthy neighborhoods shift toward community microgrids, less wealthy and powerful energy consumers could suffer negative consequences. If a subset of energy consumers invests in their own microgrid’s reliable distribution and generation infrastructure, they are no longer effectively investing in the reliability and efficiency of the whole electrical energy system. Such a trend in consumer behavior would effectively create islands of energy resilience in a sea of energy vulnerability.

11.8 Policy Considerations

In the future, in much the same way that sharing economy platforms such as Airbnb have simplified the process of contracting with an independent short-term lodging provider, electricity sharing has the potential to radically change how electricity consumers interact and behave. Indications of the future impacts can already be observed in the form of smart devices such as thermostats, operated by third-party platforms, that optimize their demand based upon real-time energy markets.

In earlier chapters, authors have discussed how the rise of the sharing economy in some sectors may increase consumption without improving sustainability. However, in the case of energy sharing, we anticipate a different trajectory that does not lead to increased consumption of fossil fuels. In part, this is because energy sharing is connected with a larger transformation toward renewable energy and a reduced carbon footprint.

The other benefits of utility control, consumer choice, and peer-to-peer energy system innovations primarily involve the reduced cost of energy generation and delivery. These reduced costs and impacts could theoretically result in an increase in energy consumption, but due to the price inelasticity mentioned earlier, we do not expect energy sharing to significantly impact energy consumption rates. That is, while reducing the price of lodging (for example, with the growth of Airbnb) may lead to more travel, energy consumption is unlikely to increase appreciably as the cost of energy goes down, particularly in developed economies (US Energy Information Administration, 2021). Any future increases in electricity consumption will instead likely be driven by increased adoption of electric vehicles and heat pumps as opposed to the specific influence of the sharing economy on energy. Finally, even with the rise of smart metering, demand response, and time-of-use pricing, there is relatively limited transparency in energy pricing. For example, the average consumer will most likely not be aware of the cost of running their dishwasher. This is due to the traditional monthly electricity billing cycle, temporally separating the saliency of the energy costs from the energy consumption occasions. New real-time IoT devices bring the potential to better synchronize consumption and costs; however, care must be taken to not overload users with too much information in real-time.

The potential societal benefits of a transition toward peer-to-peer and shared grid energy systems include improved efficiency, an increased share of renewable energy production, and more consumer engagement. In addition, these new mechanisms for local control and management provide individuals, households, and communities with more economic and political power as well as electrical power. Thus, increased energy sharing can be understood as a pathway toward realizing the goals of the energy democracy movement. Originating from trade union activism, energy democracy focuses on restructuring the political, economic, and social makeup of the energy system by transitioning to renewables while establishing democratic energy decision-making processes, equitable access to energy and energy-ownership for marginalized groups, and widely distributed and renewable energy resources (Sweeney 2012; Burke and Stephens 2018).

However, while sharing economy innovations enable consumers to have more say in where and how their electricity is produced, there are possible negative consequences associated with a transition to a shared grid and associated “energy democratization” (Stephens et al 2015). Increased development of community microgrid systems could potentially lead to greater inequality in access to reliable and affordable electric energy. The traditional electric grid doesn’t just provide electric energy, it provides electric energy reliability. That reliability is critical to economic productivity and, in some cases, social wellbeing. If the transformation underway in the electric energy sector does in fact represent a death spiral of the utility business model, the result could be that only wealthy energy consumers could afford to have reliable community microgrids, and fewer resources would go toward maintaining the grid. This could lead to a decrease in energy reliability for consumers who do not have access or cannot afford to participate in microgrid systems. One possible solution would be to increase public investment in shared grids for communities who may not otherwise have access to emerging energy sharing systems. This may in practice resemble the move in recent decades toward public–private partnerships that support charter schools in low-income communities (Koirala et al., 2016).

Another way that energy sharing is tied to disruption and transformation in the electric energy sector is through altering the geographic distribution of energy generation sites (Stephens 2019). Historically, power generation occurred near coal reserves or at hydroelectric facilities. With increased local solar- and wind-powered electricity generation, preferred sites for power generation are changing and power lines built for transmission over long distances are no longer optimally placed, leading to increased potential for congestion. However, building new transmission lines is exceedingly expensive. There is a market opportunity to invest in reducing energy consumption demand. For example, it is cheaper to give away smart thermostats and selectively adjust them automatically during peak demand times than it is to build a new transmission line. This example points to just one of the ways that new technologies and changing social and economic priorities may transform the electrical energy sector.

11.9 Conclusions

As the pace of transformation in electrical energy generation and distribution systems continues to accelerate, the possibilities for innovative ways to “share” electricity are rapidly evolving. This chapter has reviewed this dynamic landscape and provided a taxonomy to characterize different modes of energy sharing. There is a transition underway away from legacy systems with high levels of utility control and toward new opportunities for prosumer engagement, peer-to-peer sharing, and ultimately in the direction of fully integrated sharing of all aspects of electricity generation and grid management. In this transition from utility control, to prosumer engagement, to peer-to-peer sharing, toward a decentralized market structure and integrated end-to-end management energy and assets, we recognize a possible end of the traditional utility business model. As third-party platforms for energy sharing and technologies enabling local renewable energy generation, storage, and supply and demand management systems emerge, we anticipate that the traditional integrated power supply and grid management roles of utilities may shrink across many regions.

In addition to a rise in prosumers, who are both energy users and energy generators, future energy sharing could also lead to a growing number of passive electricity consumers who are integrated into community microgrids. We anticipate that prosumers, currently primarily homeowners who own private renewable energy generation infrastructure, will increasingly be outnumbered by energy consumers whose energy purchasing decisions are regulated by smart devices. For example, a customer who wants to charge their electric vehicle at their home will have their default charging algorithm set to only charge when electricity is most readily available or cheapest.

Previous research on energy sharing has focused on technological innovation and prosumer behavior, including both empirical case studies and theoretical analyses. This research provides a valuable foundation for future studies. There is a need for more research on policy mechanisms to promote and regulate sharing in real-world energy systems. Barriers to redesigning the current energy system include current regulations and policies, public awareness and acceptance of new technologies, and established corporations’ resistance to change, which can lead to manipulation of markets and disinformation campaigns that question the legitimacy and reliability of alternative energy systems. There is also need for more social science research on how sharing could contribute to more inclusive energy systems including opening up opportunities for women, people of color, and indigenous people who have been historically excluded from energy sector jobs and economic empowerment through energy systems (Allen et al., 2019).

Government and regulatory bodies have an important role to play in facilitating a transition to increased electrical energy sharing. Policy makers could, for example, establish incentive programs in the form of subsidies for new energy sharing enterprises entering the market, and punitive measures such as a carbon tax or pollution charges for fossil fuel-based energy systems. By enacting policy measures such as these and providing an environment that supports fair competition, government regulatory bodies can facilitate increased energy sharing systems that balance the goals of energy efficiency, sustainability, resilience, and equity.

12 Shared Last Mile Delivery Current Trends and Future Opportunities

Necati Oguz Duman , Ozlem Ergun , and Mehdi Behroozi

12.1 Introduction

Last mile delivery is the movement of products from a warehouse or store to their final delivery points, most commonly residential locations. One of the largest and growing components of last mile delivery operations is parcel delivery. The total volume of the parcel delivery market in the United States was 14.8 billion packages in 2018 [1], with the United Parcel Service (UPS) having the largest market share at 5.2 billion shipped packages [2].

E-commerce is a major contributor to the global parcel delivery market and is one of the fastest growing business sectors in the world. Retail e-commerce revenue is expected to rise to $6.5 trillion in the world and to $565 billion in the United States by 2023 [3, 4]. The number of packages shipped increases along with revenue growth. Amazon alone, which accounts for about 49 percent of all e-commerce in the United States [5], shipped more than 5 billion items to its Prime customers in 2017 [6] and delivered more than 3.5 billion packages through its own delivery network in 2019, almost half of its total number of shipped packages [2]. The size of the last mile delivery market increases rapidly as the share of e-commerce in the whole retail industry continues to grow and new contributors to last mile logistics activities, such as food and grocery deliveries, that rely on the same last mile infrastructure emerge. It is thus a vital matter for many businesses to manage last mile operations efficiently.

Last mile delivery related costs are estimated to reach up to 53 percent of total delivery costs [7]. There are several factors that make last mile delivery operations challenging and more expensive than the other parts of the delivery journey. One factor is that last mile delivery generally takes place within urban areas where the speed limit is lower and there is more traffic congestion. Hence more vehicles are required to deliver parcels by their due time. Furthermore, while multiple and cheaper transportation modes, such as those using railroads and waterways, are usually available for intercity logistics, mostly the only available mode for last mile delivery in an urban area is road transportation.

Meanwhile, the size and number of urban areas also keep growing. By 2030 the number of megacities, which are cities with a population of 10 million or more, is expected to rise to 41 [8]. High population density in cities leads to increased transportation activity and high emissions, as more people and packages travel. In 2017, the transportation sector was the largest contributor to greenhouse gas emissions production by generating 28.9 percent of greenhouse gas emissions in the United States [9]; Figure 12.1 compares sectors with high emissions.

Figure 12.1 U.S. greenhouse gas emissions allocated to economic sectors (MMT CO₂ Eq.) [9].

In general, objectives regarding increasing efficiency and reducing negative externalities may contradict each other. Furthermore, unexpected externalities, such as due to changing consumer behavior, may result from policy and/or operational changes leading to outcomes that are the exact opposites of the originally intended ones. For example, traffic congestion and emissions, two negative externalities of last mile delivery, are among the major concerns of city administrations, because they reduce residents’ quality of life. Shared last mile delivery attempts to make the logistics operations in urban areas more efficient and also is thought to have the potential to reduce negative externalities such as emissions by reducing the number of vehicles in traffic and by converting deliveries performed by trucks to crowdsourced cars since cars produce less emissions than trucks. However detailed analysis of some of these sharing-based last mile logistics systems have shown that opposite impacts might occur under certain city and service system characteristics. It is observed that in a subset of last mile logistics systems using crowdsourced drivers increases congestion and emissions because the small carriage capacity of cars ends up increasing the number of vehicles required and the deadheaded distance to deliver all packages [10]. Therefore, these systems and the characteristics of the environment they exist in must be analyzed comprehensively and designed and operationalized in a way that creates a win-win-win outcome for businesses, customers, and the society.

There are different applications of crowdsourcing and sharing economy practices in last mile delivery. Since they differ from conventional business models they introduce new system characteristics to be considered in both practice and research. In this chapter, we briefly introduce the existing and potential sharing economy business models in last mile delivery, and discuss the aspects that distinguish them and make their operations more complex to manage.

12.1.1 History of E-Commerce

Home shopping existed before the Internet was widely used by the public. People could buy products via phone calls. They could choose a product they saw in catalogs or TV or newspaper advertisements and order it by calling the number given. In 1994, around 98 million consumers made $60 billion worth of home shopping purchases, and almost all of the purchases was ordered through phone calls [11].

The history of e-commerce transactions and online bookstores goes back to as early as 1991 and 1992 when Computer Literacy Bookstores and Book Stacks Unlimited started selling books online through email and their websites [12]. The first item sold online through a secure and encrypted system using a credit card was made possible by Dan Kohn, a 21-year-old entrepreneur. It was the compact disc “Ten Summoners’ Tales” by the rock musician Sting and was sold through the website of Net Market Company of Nashua [13]. The transaction took place on August 11, 1994, and it was important enough to be featured in the New York Times the next day [14]. In 1995, Amazon started as an online bookstore [15] and eBay (starting as AuctionWeb) was established as an online auction and shopping website [16].

Since 1995, many online shopping websites have opened. The revenue from e-commerce continues to grow as the number of people who use the Internet grows as shown in Figures 12.2(a) and 12.2(b). The only period e-commerce retail revenue did not grow significantly was 2008–2009 due to the 2008 financial crisis (see the GDP growth in Figure 12.2(c)). It is clear, by comparison of Figures 12.2(a) and 12.2(c), that even a recession could not bring a sustained halt to the growth of e-commerce sales; in fact, in terms of percentage of total US retail, online retail grew during the recession [17].

Figure 12.2 (a) E-commerce retail sales in the United States between 2002 and 2018 [17], (b) percentage of adults who use the Internet in the United States between 2000 and 2018 [18], (c) GDP growth in the United States between 2000 and 2018 [19], and (d) the number of mobile device owners in the United States between 2012 and 2018 [20] (d)

12.1.2 History of Algorithmic Research for Optimizing Last Mile Delivery Operations

One of the earliest challenges of last mile delivery operations was faced by traveling salesmen, as they had to determine the order of visitation of potential customers’ homes or towns, often taking into account the products that they carried and the expected demand at each location. Peddlers have existed for more than 2000 years [21]. After loading products, such as toiletries, clothing, and accessories, on their pack animals, or in vehicles in recent times, they travelled in one trip to several towns and villages to sell them. Both for the peddlers and the modern travelling salesman, efficiency often is equal to minimizing the cost of travel, which usually incorporates multiple components, such as energy and vehicle depreciation costs that are all a function of the total distance travelled.

The famous Traveling Salesman Problem (TSP) aims to optimize a traveling salesman’s route. The objective of TSP is to find the sequence of customers or towns to be visited which has the minimum travel cost. The sequence must begin and end at the same location and each location must be visited exactly once. The earliest known mention of the problem is in a handbook titled “The Traveling Salesman – how he should be and what he has to do to get orders and be sure of happy success in his business” published in 1832 [22]. The first known work on the problem by a mathematician goes back to Menger’s work in 1930 [23], after which the attention of many mathematicians was drawn to the problem [24], especially after the 1954 seminal work of Dantzig, Fulkerson, and Johnson [25], as discussed in [26–31]. Many variations of the problem have also been developed [32]. Nowadays, TSP is a well-studied problem among researchers and continues to attract the energies of both mathematicians and practitioners alike due to its complex and interesting theoretical nature and its applicability in a wide variety of real-world problems. The methods developed to solve TSP are the fundamental methods for solving many combinatorial optimization problems that aim to find the optimal solution in a countable set of solutions.

TSP was later generalized into the Vehicle Routing Problem (VRP), which first appeared in a paper by Dantzig and Ramser in 1959 [33] and is the most common form of modeling last mile delivery optimization problems. The objective of the VRP is to find the minimum cost assignment of orders to a fleet of homogeneous vehicles and delivery routes for each vehicle starting from the depot and ending at the depot. Since 1959, an immensely rich literature has been established for VRP models and the development of exact, approximation, and heuristic algorithms to solve them (see [34–41] for some of the early major works).

Many variants of VRP have been introduced over the years (see [42] for a detailed history) to reflect the changing and complex characteristics of real-world delivery operations. The most important VRP variants are obtained by changing or relaxing some of the assumptions such as considering multiple depots, a heterogeneous fleet of vehicles, allowing both pickup and delivery at the destination locations, adding time-dependent costs, or removing the requirement for a vehicle to return to the depot. Furthermore, new constraints such as delivery (or pickup) time windows and load-balancing for all vehicles can be added to the problem. Finally, one can use different objective functions, such as minimizing fuel costs, minimizing emissions, and a combination of different goals to produce variants of the VRP. Mostly, these new variants of the VRP are proposed and then studied by researchers as a result of novel needs emerging from the markets by the development of new technologies or changing customer requests. For example, time windows, which are the time frame when a pickup or delivery operation must take place, are important in same day delivery; in crowdsourced delivery models, heterogenous vehicles must be considered, as drivers may have different cars; and time-dependent costs are common in urban mobility where the time to travel from one point to another point may drastically change depending on the time of day.

12.2 Sharing Economy Models in Last Mile Delivery

Digital platforms enable the interaction among different participants who supply or demand a set of services or products. Generally, sharing platforms themselves do not offer the services or products, but only match the available supply and demand. Matching as a service has been around for more than three millennia [43]. However, large scale and rapid matching of supply and demand became available with the pervasiveness of the Internet and smartphones. The Internet enabled the exponential growth of digital platform economy businesses, since it lowered the entry barrier by not requiring a large-scale initial investment in hard assets. Especially after the 2008 financial crisis, a wide variety of online platforms started services such as DoorDash, Uber, TaskRabbit, and AirBnB. The growth of online platforms was also accelerated as smartphones became widespread and the development of mobile technologies and applications accelerated (Figure 12.2d).

Currently in shared last mile delivery, many businesses operate with different sharing models. In the following, we categorize different business models that currently exist or are being discussed as options for the near future.

12.2.1 Crowdsourced Delivery

Crowdsourced delivery platforms do not perform shipments with their own dedicated vehicle fleets or professional drivers. Individual drivers join these platforms as independent contractors, commonly after a security background check, and deliver the packages to customers using their own vehicles. In many platforms, crowdsourced drivers have different levels of flexibility in choosing when to start and stop working.

The landscape of last mile delivery platforms evolved over time. Different platforms, past and present, such as Deliv, GoShare, Instacart, GrubHub, and Postmates, provide last mile delivery using crowdsourced drivers for different types of products such as parcels, groceries, and meals. Furthermore, while some platforms focus on delivery of one type of product, others deliver multiple types. The type of product delivered impacts the constraints of service. For example, some groceries must be carried in cold boxes to prevent spoilage and must be delivered rapidly, while parcel delivery may have a longer delivery time period and can be heavier or bulkier.

Sharing economy delivery platforms also differ in terms of how they match delivery orders and drivers. Some platforms post the delivery orders with origin, destination, and earnings information. Drivers login to these platforms, check the posted orders, and can select the orders that they would like to complete. Other platforms perform the matching themselves given the availability of the drivers and time windows of the orders. They assign the orders to drivers so as to maximize the platform’s profit. Afterwards, drivers can choose to accept the order or refuse it. Furthermore, some of these platforms only make the assignment and leave the delivery sequence up to drivers, while others generate efficient delivery routes and ask drivers to follow these routes. All of these supply–demand matching strategies, complemented with different pricing mechanisms, create digital marketplaces with different characteristics and externalities.

It is commonly believed that crowdsourced drivers provide greater flexibility for delivery companies compared to a dedicated fleet of vehicles, because the driver pool can be adjusted to meet service demand dynamically and with lower costs. However, since crowdsourced drivers are currently not full-time employees, they usually are less stable and predictable, which introduces a high level of uncertainty in workforce and capacity planning. Additionally, crowdsourced driver performance is generally observed to be worse compared to delivery employees of the platform, as the latter have more experience with the business and its customers.¹ Reputation² mechanisms and other incentives are designed to encourage crowdsourced drivers to display more stable behavior and deliver services with an eye towards higher customer satisfaction.

From an optimization perspective, the crowdsourced delivery problem builds upon the rich body of VRP literature; see [44] for an overview. In addition, recent studies have focused on the optimization of crowdsourced delivery systems, the challenge of finding the right balance with respect to the above-mentioned criteria, and the design of appropriate incentives [45, 46–48]. However, many questions related to market design and comprehensive optimization of crowdsourced delivery platforms, including externalities, still remain open.

12.2.2 Shared Urban Distribution Centers

Urban distribution centers are facilities that are designed and built close to city centers to efficiently handle warehousing operations, such as loading and unloading trucks and sorting packages, as well as cross-docking operations. Their objective is to reduce last mile delivery mileage and fuel consumption and benefit from economies of scale by pooling deliveries into larger vehicles at the urban distribution centers. While it is generally believed that delivery operations using urban distribution centers benefit both delivery companies by reducing their costs and the general public by reducing pollution and traffic congestion, comprehensive analyses of these systems sometimes point to a net negative impact depending on the characteristics of the demand and the urban area.³

In operational models that use a shared urban distribution center, larger trucks bring packages from different companies’ warehouses to an urban distribution center. After sorting the packages according to their destinations, packages are loaded to usually smaller vehicles and delivered to customers. At this stage, as an additional benefit of shared distribution centers, orders from multiple companies can be delivered by the same vehicle. Urban distribution centers have multiple docks to load and unload the delivery trucks which reduces the truck queue as the operations can be completed in parallel. Also, these docks complete loading and unloading operations faster than most brick-and-mortar stores, as they are designed for this purpose.

Furthermore, sharing might occur at different levels among companies in the urban distribution centers. They can share only the floor of the building which would increase the space utilization and could reduce the energy consumption, as heating, air conditioning, and lighting could be achieved more efficiently. The participating companies can also share some or all of the operations mentioned earlier. This would further increase the benefits, as the idleness of the shared machinery and equipment would be reduced. Additionally, a collaboration could be formed among the delivery companies, so that the deliveries of different companies could be pooled together, and better routes could be formed in terms of cost efficiency, traffic congestion, and emission production.

On the other hand, convincing e-commerce and/or delivery companies to participate in an urban distribution center is a challenge. The competition among these companies may prevent them from joining a collaborative system. They must be assured that benefits will be greater than costs. Several related problems regarding stable sharing of distribution centers, such as costing, space allocation, and delivery scheduling are discussed in [49–51].

12.2.3 Pickup Lockers

Pickup lockers are secure delivery locations similar to post office (PO) boxes. Generally, however PO boxes are rented for a long time period, whereas lockers are assigned to deliveries for a short time, mostly up to a few days. For this reason, lockers are typically shared by customers. Companies, such as GoLocker, Neopost, and Amazon Locker, currently utilize pickup or product return lockers in their last mile delivery operations.

To use locker delivery service, a customer would choose a pickup location as the delivery address. The delivery company ships the package to this pickup location. When the package is placed into a locker, the customer receives a one-time passcode that is used to open the locker which contains the order. After the customer picks up the packages, the locker is assigned to other deliveries.

Pickup lockers provide benefits for both delivery companies and customers. For customers, a locker station is a secure place to receive their packages in case they do not have a safe area, such as a lobby, in their building. Also, customers do not have to wait at home for delivery since they can pick up their packages from the locker anytime they want. For delivery companies, the locker delivery may reduce costs, since multiple deliveries are expected to be consolidated if their destination is the same locker location. Additionally, drivers avoid delivery problems, such as a wrong address, since the lockers are at known locations.

Pickup locker stations must be designed so that they are an attractive option for consumers. For example, the station must be user friendly and close to areas that customers visit frequently. Additionally, incentivizing customers, for example by providing discounts to encourage consumers to use the service, may increase both the cost and environmental benefits. Designing incentive schemes that consider several factors, such as demand and geographical area characteristics, for each customer dynamically may amplify these benefits. Hence system-wide analysis of all benefits and costs for these shared locker model must be made carefully to inform its design and operations. Comparatively, this area in last mile delivery systems has been studied less completely (see, for example, [52–54]).

12.2.4 Autonomous Vehicles

Autonomous vehicles (AVs) including self-driving cars, automated guided vehicles (AGVs), and unmanned aerial vehicles (drones) may release a big potential in last mile delivery operations. There are several AV brands that are working to produce autonomous vehicles that can follow a given route without human input along the way or be guided by an operator in a semi-automated manner. Currently no AVs are fully autonomous, because such vehicles still require human supervision to prevent accidents when unexpected events occur.

Development and commercial availability of fully autonomous vehicles in the future will open up the possibility for owners of AVs both to make personal use of them and share them with delivery companies. A future scenario could include people going to work with their AVs and then, while they are at work, allowing those AVs to be used by last mile delivery platforms. In a more advanced scenario, AVs could communicate and choose their paths so that traffic congestion could be reduced. Such future sharing of AVs can significantly reduce the idle hours of vehicles sitting in parking lots, while also benefiting the public and environment, as fewer vehicles would be required to achieve both urban mobility and freight transport.

Similar to AV sharing, and in the much nearer future, drones could be shared between their owners and delivery companies. When owners are not using their drones, they can lend them to make deliveries. Currently, one of the limitations of personal drones is their small size, which constrains delivery capacity and allows only small parcels to delivered. Also, personal drones currently have batteries that are much smaller compared to commercial drones, thus reducing their delivery radius. This poses a significant problem especially in cities with tall buildings, where drones cannot fly in a straight line.

Autonomous vehicles can transport packages between locations, but a system to load and unload packages to and from AVs is also required in a last mile logistics operation. An automated system to handle loading to different brands of vehicles, as might be the case in a shared AV system, might be hard to develop in the near future. Additionally, customers might need to be present for delivery, as there will be no one to pick up a package from the AV and deliver it to them. A customer picking up a wrong package can cause significant problems. Therefore, the safest approach currently could be distributing packages one at a time on an AV, which would increase inefficiency and potentially the resulting energy use and pollution of the operations. Incorporating both system optimality and energy usage in operational models may help mitigate the side effects of using autonomous vehicles (see [55].

12.2.5 Public Transportation for Package Delivery

In most cities public transportation network capacities are designed to handle a significant portion of the rush hour commuter needs. However, this causes the public transportation network to have a high idle capacity during long stretches of the day. This idle capacity during low volume hours could be used for last mile delivery, which in turn would reduce the number of vehicles in traffic, especially in the most congested parts of the cities. For example, instead of sending additional vehicles to downtown areas to pick up or deliver packages, last mile delivery operations could utilize the existing public transportation networks in similar ways as described in several studies [56–59]. Synergistic usage of public transport for package delivery could also produce extra revenue for the public transportation authority of an urban area, which in turn can be invested in infrastructure, benefitting the general population.

Public transportation vehicles, such as buses, trams and subways, are designed to carry people and not packages. To use them to deliver packages, operational strategies must be developed and required tools must be deployed at the stops for loading and unloading operations. This would require a high initial installation cost. Furthermore, loading and unloading operations would increase time spent at stops, thereby extending travel times. A small tail car designated for carrying packages that can be automatically and quickly removed and replaced at specific stations is a possibility for carrying packages in and out of highly congested areas using public transportation. Additionally, not all customers might be willing to come to pick up their orders. Therefore, a system to distribute packages from the stops to customers’ houses could complement such a system. An alternative approach, that may be ideal for certain types of customers and packages, could be to combine public transport of packages with delivery lockers at metro stations. This way customers could pick up their packages at their local stations, perhaps at the end of the day on their way home.

12.3 Sharing Economy Platforms vs. Conventional Business Models

Sharing platforms and conventional last mile delivery businesses differ in several ways, including workforce, competition, structure of their corresponding markets, and the externalities that may arise as a result of their operations. In this section we discuss these differences.

12.3.1 Workforce

Conventional delivery companies have dedicated and professional drivers to deliver packages. Since these drivers are traditionally also full-time employees, they have certain benefits which may include health insurance, paid sick leave, and retirement plans. In crowdsourced delivery businesses, generally drivers do not have such benefits, as they are categorized as independent contractors and not as employees, which continues to be a subject of many heated debates and regulatory initiatives from a labor perspective [60].⁴

From an operational perspective, although professional drivers have responsibilities regarding the use of the instruments and vehicles given to them, their employers take care of the cost and planning of insurance and maintenance of these instruments and vehicles. However, crowdsourced drivers have to manage these issues because they use their own equipment and vehicles.

Furthermore, traditional delivery companies have well-established procedures for making deliveries. The decisions on package sorting, loading, and delivery routing are mostly made by specialized professionals or software packages following established procedures. In these traditional business models, drivers have little freedom or flexibility on how to perform these operational processes. Their work shifts may also not be as flexible. In contrast, crowdsourced drivers working as independent contractors are thought to have more freedom of choice on these operational issues. Depending on the platform, they can decide when to work, which deliveries to accept, and how to route their journey. While there appears to be more freedom for an independent contractor, in many situations these drivers’ actions are also restricted through rules or incentive mechanisms. For example, drivers might be held responsible for late or missed deliveries if they occur because of drivers’ actions. In some cases, the flexibilities given to crowdsourced drivers are increasing as platform companies are reacting to recent regulatory efforts and trying to fortify the often-blurred boundary between the definition of a worker and an independent contractor [61].

In traditional delivery companies, the fluctuation in the workforce on a given day is very limited, as drivers must inform their employers in advance if they cannot work on a day and this does not occur frequently. As a result, the delivery capacity on a given day is known in advance, which allows for better operational planning of deliveries. However, crowdsourced drivers may choose not to work and do not even have to inform the company in advance. These conditions create a high degree of uncertainty, which leads to higher operational costs.

12.3.2 Competition

Traditional businesses mostly compete with others in the same business sector. For example, parcel delivery companies compete with other parcel delivery businesses and not with taxi companies. However, crowdsourced last mile delivery companies must compete with crowdsourced urban mobility companies as well as other crowdsourced delivery companies, because drivers can choose to work for any of these businesses as independent contractors. Drivers can check multiple platforms at the same time while sitting in their cars and pick the offer with the best payment and/or work conditions from all the platforms they have joined. Drivers can simply switch from a parcel delivery platform to a ride-sharing platform as soon as they complete an order.

Although payment is an important factor for drivers in choosing among platforms, there are other factors to be considered. A driver might be indifferent to small changes in earnings depending on other features of the platforms. For example, if a driver prefers to socialize while driving, it is more likely that they will choose a mobility platform even though a delivery platform might be offering higher earnings. Level of freedom of choice (orders being assigned by the platform vs ability to choose orders), pressure from the platform (receiving frequent calls/texts to induce fast delivery), and incentives offered by the platform are among possible factors that might impact drivers’ decisions.

The factors that impact drivers’ choice of platform and their performance must be explored. These factors are important as they impact the loyalty of a driver to a platform and their dedication to providing quality service to the customers of the platform. Additionally, incentive schemes for the drivers can be developed. The incentives can be customized for each driver according to their individual preferences. Individualized incentives can improve drivers’ job performance and happiness.

Most platforms do not offer employment benefits to crowdsourced drivers. However, recently Deliv started to offer full time benefits, such as retirement plans and health coverage, to its drivers in California [62]. While this decision by Deliv was taken partly in anticipation of CA AB5 regulation [60], Deliv also believed that converting to a full-time employee model might improve service quality and reduce operational costs as well as attract and keep high-performing drivers since offering these benefits provides a bargaining leverage for drivers against other platforms. In a similar recent move, Uber announced that it will consider paying healthcare benefits to workers proportionate to the number of hours they work [63].

Alternatively, multiple platforms might collaborate instead of competing with each other. For example, a mobility and a delivery platform can collaborate to share their driver pool, and the platform with low demand can transfer some of its drivers to the other one with high demand. However, formation of such collaborations must be regulated and watched carefully in order to prevent the emergence of monopolies or trusts and to ensure drivers are not harmed as a result.

12.3.3 Market Structure

In order to survive in the harshly competitive environment, platforms need to have a certain level of demand and supply in their systems and a continuous dynamic balance between the two must be kept. Dynamic matching of supply and demand in sharing platforms has been a very active research area in recent years [45, 64–66]. For crowdsourced delivery platforms, if there are not enough drivers at any given time, the orders in the systems will end up not being delivered on time or get canceled, increasing the risk of customers switching to another platform. On the other hand, if there are not enough orders in the system, drivers will not have enough work to do resulting in low earnings that incentivize them to switch to another platform.

Combined with the dynamic competition for labor, having to match supply to demand dynamically leads platforms to develop complex pricing policies. Reducing service prices is an effective method for competing for demand. However, this contributes to the major problem of unprofitability experienced by most platform-based businesses [67, 68]. Moreover, if the platform lowers prices too much, its drivers might switch to other platforms. This might in turn result in unsatisfied customers who might also leave. On the other hand, surge pricing, in which prices are increased at times of high demand and low supply, could instigate gaming behavior by the drivers (such as turning their apps off to generate “low supply”). Setting the prices to achieve an equilibrium for both supply and demand is a very challenging problem. While the gold standard in two-sided markets is to develop a dynamic pricing policy that balances supply and demand and drives desired customer and driver behaviors, the adaptive behavior of stakeholders and the continuously changing environment make this a moving target. Additionally, in principle, establishing good and clear communication with both sides of the market about the nature of the pricing policies might provide further benefits, as behaviors of customers and drivers can be shifted to a desired setting.

12.3.4 Social and Economic Externalities

Shared delivery platforms have the potential to reduce negative externalities, such as pollution and congestion. They can reduce the number of vehicles operating or the distance travelled. For example, the use of urban distribution centers can reduce distance travelled by increasing truck capacity utilization and integrating public transportation. Similarly, freight transportation networks can reduce the number of vehicles in traffic, especially in congested areas of the city.

However, simple implementations that do not consider all impacts and that lack necessary adjustments to the local needs of urban areas have led shared delivery systems to produce negative externalities in recent years [69]. For example, as discussed previously, crowdsourced delivery cars could be expected to reduce pollution and congestion because they produce less emissions than delivery trucks and delivery trucks are required to return to company parking lots, thereby increasing the distances driven. However, cars are smaller and cannot deliver as many items as trucks. Therefore, more cars are required to complete all deliveries than trucks, and this in turn may produce just as much or more congestion and pollution.

Innovative solutions and strategies are required to reduce the potential negative externalities of shared delivery systems (see [69, 70] for the literature on such strategies). For instance, self-driving cars (that might also be electric) and drones can be used for deliveries (or mobility) when they are not used by their owners. This might reduce the number of vehicles in a city which can lighten the traffic congestion and pollution. However, some people might buy autonomous cars only to produce revenue for themselves. Such a pattern may increase both congestion and pollution in a city.⁵

12.4 Conclusion

Crowdsourcing and shared delivery platforms are among possible solutions which might help to reduce the externalities. They can reduce traffic congestion and pollution by decreasing miles travelled and the number of vehicles in the traffic. Moreover, they can provide additional income opportunities and improve the sense of community. On the other hand, large-scale sharing platforms are relatively new, growing with the speed of internet and mobile technologies. They have many differences with classical business models because they require a level of collaboration of participants to provide benefits for all involved parties. As a result, many questions arise on how to implement and operate these platforms.

To achieve significant benefits from the shared last mile delivery applications for the society, environment, and participating companies, each business model must be analyzed as a system, considering all of its unique set of challenges and impacts. The analyses should take all stakeholders, such as employees, city administrations, and residents, and the interactions among them into account. Additionally, multiple business models should be assessed together. For example, public transportation can be used to bring packages to urban distribution centers from a company’s warehouse, and after the sorting operations, they can be delivered to customers by crowdsourced drivers. Although creating and maintaining such a large collaboration among multiple companies would be very complex, this kind of synergy can amplify the overall benefits in the last mile delivery ecosystem.

13 Future Themes in the Sharing Economy

Babak Heydari , Ozlem Ergun , Rashmi Dyal-Chand , and Yakov Bart

One of our goals in this volume has been to demonstrate that interdisciplinary, convergent research and analysis are indispensable to the ideal of optimizing for a more equitable, democratic, sustainable, and just sharing economy of the future. The contributions to this volume are thus central to our demonstration because they epitomize this ideal by drawing scholars from different disciplines into conversations about the most fundamental questions and challenges related to reengineering the sharing economy. But these contributions have also provided important information about some of the answers to key questions that must be addressed as we move forward. In this concluding chapter, we draw from the rich analyses undertaken by our contributors to outline important substantive lessons that can contribute to a framework for reengineering the sharing economy.

In particular, we focus on five core dimensions that are central to optimizing for a just sharing economy: understanding socioeconomic externalities; pursuing resilience; charting more just and systems-oriented business directions; defining the future of work; and prioritizing access and equity. Our effort in this chapter is more modest than to provide detailed conclusions about the relevance of any of these dimensions. Rather, it is to highlight the multiple ways in which the analyses throughout this volume intersect with these dimensions. As we describe, we believe that the centrality of each of these dimensions is itself an important lesson about the future of the sharing economy. Additionally, these dimensions convey significant information about the values that must be prioritized in the next generation of sharing economy platforms. Finally, and crucially, they help to highlight key questions that remain for future research and exploration.

13.1 Socioeconomic Externalities

Digital platforms are becoming more integrated into our daily lives, collectively adding tens of millions of new users every year. As multiple chapters in this volume have discussed, however, the effects of these platforms go far beyond their users. Most platforms indirectly impact the socioeconomic wellbeing of people in many ways. These indirect effects are often referred to as externalities, and given their broad scale and scope, it has been a major research effort to understand, measure, and regulate them. Such externalities have also fueled public debate since the early days of sharing economy platforms. In important respects, the analyses in this volume push beyond the current frontiers of research about socioeconomic externalities.

For example, the two chapters on urban mobility companies investigate the socioeconomic externalities of sharing platforms in urban contexts, and in doing so, they provide important clues to solving complicated puzzles about the hidden effects of the shared mobility industry. As Behroozi’s chapter shows, despite initial promises that ride-sharing services could reduce urban traffic congestion, this is not always the case. In practice, ridesharing can even increase congestion for a number of reasons, including by substituting for public transportation in some cities. Evidence suggests that some of these concerns can be addressed if the industry moves from car-hailing to ride-pooling. However, moving to ride-pooling often requires an array of incentive mechanisms and technical design considerations that have heretofore been less well-charted, as Koutsopoulos, Ma, and Zahedi discuss in Chapter 9.

Chapter 10, by O’Brien, Heydari, and Ke, is similarly illuminating in discussing lodging, where debates over the consequences of short-term rental platforms on the quality of urban neighborhood life are especially vociferous. As the authors argue, strong penetration of short-term rentals enabled by platforms such as Airbnb can have an array of socioeconomic consequences at neighborhood levels, since such penetration increases the influx of nonlocal people to the neighborhoods and results in different – positive and negative – social and economic consequences. Such consequences could mean higher local rents due to decreased real estate supply or higher quality of local services caused by increased local competition. At the same time these platforms can have longer-term effects, because a high level of short-term rental penetration can poke holes in the social fabric of a neighborhood and disrupt its social organization over time.

The analyses in these and other chapters contribute to two broader insights about socioeconomic externalities in the sharing economy. While discussion of externalities has always been integrated into the literature on sharing economy platforms and multisided markets, most of the focus has been on network externalities and on some economic externalities such as the effect of these platforms on employment and traditional businesses. One of the insights gained from the analyses in this book is that we must define externalities more broadly and integrate this broader definition into designing the technical and regulatory elements of these systems. Indeed, this is part of the agenda for reengineering the sharing economy. As Heydari argues in Chapter 2, the broader definition of externalities expands the range of stakeholders who are affected by sharing platforms to include local residents and businesses, potential second- and third-tier businesses that could emerge in the ecosystem created by a sharing platform, and even other sharing platforms, given the possibility of interplatform interaction across different platforms that provide similar or complementary services.

In addition, much of the debate about socioeconomic externalities has been shaped by anecdotes and opinions that are often rooted in too much optimism or pessimism towards sharing platforms. While there are cases where a particular positive or negative externality of a platform outweighs the rest of the consequences, the reality about most types of socioeconomic externalities is more complex than what these anecdote-based debates suggest. This book highlights that assessing the overall impact of sharing platforms on a given socioeconomic factor (such as traffic congestion, the environment and carbon emissions, and neighborhood economic and criminal activities) depends on understanding the tradeoffs among competing factors through which platforms can either benefit or harm that factor. The relative weight of these competing factors depends on certain design and regulatory parameters on the one hand and the time horizon of the analysis (short-term versus long-term) on the other. Further, evaluating trade-offs requires us to learn the causal mechanisms by which platform parameters are associated with socioeconomic externalities.

As Heydari’s chapter discusses, such a methodology requires steps such as quantifying the effects of these platforms in the short term and long term, determining different stakeholders and soliciting direct or indirect inputs from them, and establishing methods to aggregate inputs from different stakeholders. Moreover, given the importance of identifying different causal mechanisms, the methodology can benefit from combining empirical studies with analytical modeling. Ultimately, outputs of these models can contribute to designing externally imposed regulations, as described by Dyal-Chand in Chapter 7, as well as internal governance mechanisms designed by platform companies.

The chapters in this volume inspire future research in socioeconomic externalities on three vital areas. An initial step in identifying the impacts and mechanisms of socioeconomic externalities will be to improve sociotechnical modeling methodologies, which will allow empirical identification to be integrated with system-level simulation. Second, even when we can model and quantify various types of externalities, design and policy decisions are influenced by how we weigh and rank them. Considering platforms’ algorithmic nature, this can be challenging, especially since rankings and weights must be updated dynamically. Last but not least, these models need to identify lever points that platform designers and regulators can utilize in order to govern socioeconomic externalities.

13.2 Resilience

As several chapters have observed, some digital platforms serve the function of modern critical infrastructure in many parts of the developed world. This fact, laid bare by the COVID-19 pandemic, is a startling indication of the extent to which the sharing economy has transformed modern living for many of us. It is a fact that requires us to comprehensively reevaluate the forms, functions, and values that inhere in the sharing economy today. This reality also means that it is necessary to examine the resilience of sharing platforms, just as we do with other critical infrastructures. The resilience considerations of sharing platforms require us to ask two overarching questions. As is standard practice in considering the resilience of traditional infrastructures, we must examine how resilient sharing platforms are in response to unexpected disruptions. In addition and moving beyond standard practice for other forms of critical infrastructure, we must consider how these platforms can affect the resilience of other socioeconomic activities.

The COVID-19 pandemic served as a giant stress test for the resilience of many industries, social and economic institutions, and sociotechnical systems. Digital platforms can be credited for contributing positively to the resilience of pandemic life in much of the developed world by facilitating quick transitions to working at home, online shopping, and virtual socializing. Such quick transitions were enabled by a number of factors, such as preestablished logistics infrastructures for companies such as Amazon and Wayfair. Another enabler was the quick repurposing of platform capacities. For example, Uber quickly moved resources from Uber Ride Sharing, for which demand was plummeting, to Uber Eat for which demand was skyrocketing.

As the pandemic revealed, several inherent characteristics of sharing economy platforms make it possible for such platforms to respond quickly to sharp changes in demand level, thus contributing to the overall resilience of the broader ecosystem for the type of services they provide. Consider the mobility industry as an example. First, because mobility platforms do not own the underlying assets (namely cars), they are nimbler in changing the supply level by updating the participation rate on the supply side of the platform. These changes are possible within a feasible range, determined in the short term by the existing pool of agents on the supply side (namely, active drivers), but can grow or shrink in the longer term depending on the overall conditions of the platform ecosystem. Second and as a mechanism to reap the benefits of the first factor, sharing platforms can use dynamic incentives, often in the form of dynamic pricing, to close possible gaps that emerge between the supply and demand levels. Finally, the digital and on-demand nature of many of these platforms means that these platforms can quickly estimate sudden changes on their different sides and buy more time to react to those changes. In Chapter 12 Duman, Ergun, and Behroozi discuss some of these factors in the context of the last mile delivery problem, which is considered a major logistical bottleneck in implementing resilient and sustainable e-commerce systems.

Despite the positive contributions of sharing economy platforms to the overall resilience of essential services, several chapters in this volume raise important concerns about the potential negative impacts of these platforms on infrastructure resilience, especially in the future as we become more dependent on them. As the chapters on mobility discuss, mobility platforms can shift some of the demands from public transportation systems to ride-sharing services, resulting in further reductions of available investment budgets in public infrastructures. From a resilience perspective, this is not necessarily concerning as long as the platform-based systems can provide a continuation of widespread affordable service, especially in the aftermath of a major disruption. However, such access is not guaranteed, given the asset-free nature of many sharing platforms on the one hand, and on the other hand their relationship with their workers, as Schor and Vallas discuss in Chapter 6. Both these factors put much of the supply-side management at the mercy of short-term incentives offered by the platforms, which might fail under extreme circumstances.

Relatedly, it is important to understand that, like most resilient systems, the mechanisms that help these platforms to be adaptive to sudden changes can only contribute to resilience up to certain levels. These same mechanisms can become ineffective, and even counterproductive, once the changes in supply and demand go above a certain level. For example, dynamic pricing can result in unacceptable surge prices in the face of a sudden rise in demand. Importantly, too much decrease in the level of supply, in the case of a drop in demand, means that the platform reduces its geospatial coverage and consequently its on-demand nature. For ride-sharing platforms, this means that the average wait time for each passenger will increase because of the low number of drivers, resulting in dissatisfaction on the passenger side. Such dissatisfaction can lead passengers to pursue other options and lower the demand, which in turn lowers the supply, as discussed by Koutsopoulos, Ma, and Zahedi in Chapter 9). This downward spiral, similar to what is often known as the Wild Goose Chase phenomenon, can make platform systems nonresilient. By contrast, public transportation demonstrates a more linear resilience behavior – at least in the short and medium term – in response to demand changes. As Heydari argues in Chapter 2, public–private partnerships between sharing platforms and public infrastructures can address some of these concerns by including resilience considerations in their agreements governing the provision of services to passengers. These resilience-oriented partnerships can go beyond transportation infrastructure and extend to energy systems, as Kane, Allen, Si, and Stephens show in Chapter 11, focused on future energy systems.

Finally, any discussion about resilience and economic externalities is not complete without considering environmental sustainability considerations. As Eckelman and Kalmykova discuss in Chapter 3, despite initial promises about the positive role of sharing platforms on the environment, it is quite challenging to evaluate the actual sustainability orientation of sharing economy platforms. Sharing companies are highly heterogeneous in this regard and can create a wide range of unintended consequences for the environment. The authors describe a number of these unintended consequences and enumerate several back-end and front-end design opportunities for incentivizing beneficial environmental outcomes. However, research about environmental costs and benefits of peer-to-peer sharing platforms has been limited, and more studies are needed to further guide and prioritize such design opportunities.

Resilience of complex sociotechnical systems has been a topic that has attracted increasing interest from several academic communities. Resilience in these system types is the result of a combination of top-down and bottom-up responses at different levels and by various actors, including the synergistic role of policy design on the side of the regulators, behavioral change on the side of human agents, and repurposing of existing capacity and technological adaptation on the side of businesses. Studying resilience in sharing economy platforms can not only prepare us for future disruptions but can also teach us important lessons about multilevel, synergistic responses at the system level that are useful for the broader context of complex sociotechnical systems. We hope the chapters in this book will inspire new thinking about a range of crucial questions regarding system resilience, especially in the wake of the COVID-19 experience. For example, how can we identify and characterize existing capacities in sharing platforms that can be quickly and efficiently repurposed and reallocated in the face of major disruptions? How might we create more synergy between top-down responses (including policy and business decisions) and the behavioral changes that often act as bottom-up adaptation mechanisms in the face of a disruption? How can we create scenario-study models that incorporate these levels of system responses and that can be used both to identify the trade-offs of resilience decisions and to communicate them to the key stakeholders? And finally, how can we better integrate the adaptability aspects of platform-based systems with the objectives of public decisions – such as through public–private partnerships – to better steer the direction of a system’s response towards the public good?

13.3 Future Business Directions

The focus of this volume has been both to provide a systemic perspective on sharing economy platforms and to discuss design and governance issues at the intersection of engineering, regulation, and operations. This is in contrast and complementary to recent books that look at the sharing economy from the perspective of the firm making key business decisions. The recent focus on the firm’s perspective is not surprising, since the trends that gave rise to the modern sharing economy (which are examined in the Introduction to this book) have been associated with substantial new value creation over the last decade. Hundreds of market-based platforms continue to take advantage of growing algorithmic and data capabilities coupled with rapidly advancing technologies to allocate access to goods and expertise in such a way as to keep transaction costs to a minimum while in many cases utilizing available capacities of physical assets as fully as possible. Collectively, these features make sharing economy platforms unique from a business perspective. Although this volume’s focus has been elsewhere, a number of chapters in this volume have raised important business implications that may serve as catalysts for future research in this field.

First, this book makes a case for the possibility of bringing greater shareability to a range of platform services. For example, Chapter 9 observes that on-demand mobility services currently provide very few shared rides, and the authors present models and recommendations designed to improve sharing in these services. Sharing economy models are discussed in Chapter 12 as a way to solve the challenging problem of last-mile deliveries in e-commerce. Similarly, Chapter 11 discusses the possibility of using sharing economy models in energy systems.

Second, several chapters emphasize that establishing and maintaining trust is essential for the operation of sharing economy businesses. Tadelis, in Chapter 5, argues this could be accomplished by designing appropriate feedback mechanisms. Additionally, privacy is becoming a top concern for users of platform-based businesses, and as discussed in Chapter 4, it is important to understand the privacy calculus of platform users in order to identify potential trade-offs associated with privacy protection measures and other business metrics.

Third, several chapters in this volume highlight ongoing concerns about business competition in the sharing economy. For example, looking back at how this sector grew over the past decade, there seems to be a wide gap between the multibillion-dollar valuations of platform companies like Uber and their lackluster profitability. This gap is often attributed to the power of strong network effects, as discussed in Chapter 2, which create entry barriers and effectively lock in users once the platform company succeeds in attracting many of them during the initial growth spurt. The lock-in problem has been typically discussed in the context of competition among online social networks, where it arises due to embedded positive direct network effect (for example, the more of your friends use a social network site, the higher your value of using it). Leading online social networks, such as Facebook and LinkedIn, have pursued this strategy to establish a sustainable competitive advantage. Thus, it is important to gain a better understanding of how various components of switching costs may reduce competition among sharing economy platforms. Reduced competition may result from several factors, including the cross-side network effects, possible multihoming (users joining multiple platforms at the same time), and myopic decisions on the part of users. A great deal more analytical and empirical research about these nuanced factors is needed to better understand concerns about sharing platform competition.

Fourth, on the operational side, the on-demand service delivery promise of sharing economy platforms, together with a less predictable crowd-sourced contractor resource structure for performing the services, can lead firms to maintain excessive capacities. Perhaps surprisingly, such high (and costly) capacities often correspond to low utilization of resources, contrary to common claims made by sharing economy platforms. An example of this can be seen in ride-sharing platforms: They encourage many drivers to join and be active on the platform app but impose on those drivers to spend a significant amount of time waiting for fares. Another example can be seen in the last-mile delivery context, as discussed in Chapter 12. More research is required to understand these unexpected phenomena on the operational side of platforms.

More broadly, over a decade has passed since most sharing economy leaders started their businesses, and major industry changes have followed the COVID-19 pandemic, thereby raising the possibility that it is time to reassess some of the business assumptions that have been widely and, for the most part, silently accepted in the sharing economy industry. For instance, it has been suggested that one of the factors driving sharing economy growth is the shift from ownership to use among millennials. However, we have seen much evidence of the opposite trend during the COVID-19 pandemic. In response to rising demand, the average price of used cars increased by more than 40 percent in less than two years from the start of the pandemic.¹ Meanwhile, millennials significantly contributed to the real estate market boom in 2020–2021.² It remains to be seen whether these recent trends are temporary and could be fully explained by supply chain disruptions or whether these are harbingers of long-term structural changes that present a serious challenge for sharing economy platforms.

Finally, we emphasized in the Introduction to this book that most business decisions cannot be divorced from platform governance decisions for sharing economy companies. While this is true for all businesses, governance decisions are crucial for the business success of sharing economy companies for many reasons. These reasons include regulatory compliance, safety imperatives, and resilience and environmental concerns. Therefore, we expect more research to be conducted on integrated modeling and analysis of business and governance decisions in different platform types, allowing public policy stakeholders to better assess regulatory environments and possible trade-offs.

13.4 The Future of Work

The workplace has undergone dramatic changes in recent decades as a result of numerous disruptive forces including globalization and automation. As multiple chapters in this volume discuss, sharing economy platforms are the most recent and rapidly accelerating disruptive force on the structure of work, with both intended and unintended consequences. For example, while digital platforms have made remote work possible during the COVID-19 pandemic, gig-workers employed by many sharing economy platforms have precarious working conditions, as described by Schor and Vallas in Chapter 6. These circumstances have fueled a fierce debate on the employment status of platform economy workers and on the future of work in general. Although some sharing economy platforms have existed for over a decade and research on platform-based work has grown rapidly^{3, 4, 5} it remains unclear how platform jobs affect the quality of employment, whether workers are exposed to risk with potentially adverse effects, and how platform workers view their position as independent contractors. The future of work is thus a key dimension that must be prioritized in any serious effort to reengineer the sharing economy.

Almost all sharing economy platforms have two core characteristics. First, they use Internet-based digital technology and algorithms to mediate transactions between buyers and sellers of goods and services. Second, they define themselves not as employers, but simply as providers of information systems that “match” independent contractors with potential customers or clients – an important economic and legal shift that redefines the nature of employment and that externalizes many financial and legal risks. In turn, many of these risks have been imposed on the workers themselves. As Chapter 12 on last-mile delivery and Chapters 8 and 9 on mobility discuss, crowdsourced independent-contractors help platforms to achieve greater operational flexibility to provide on-demand services, agilely matching supply to demand with minimal risk to the firm. While many on the platform side argue that this flexibility provides the necessary competitive advantage to firms and flexible working hours to workers, it also introduces a significant level of uncertainty for all actors involved in the operational environment.

The challenges of the platform economy impose unforeseen costs on platform firms themselves, which often struggle to scale up their business models in sustainable fashion. As the chapters on labor, urban mobility and last-mile delivery discuss, one major challenge that businesses face flows directly from their use of the independent contractor model. Firms cannot simply impose work schedules on workers, since freedom over working hours constitutes an important selling point for the recruitment of workers. As a direct consequence, firms encounter heightened levels of uncertainty about staffing levels, which are often vital to their business success. Moreover, since platform workers must assume responsibility for many operational costs and risks, they exhibit extremely high levels of turnover, which imposes substantial costs on platform firms in the forms of bonuses, marketing campaigns, and promises of minimum levels of earnings. Dynamics such as these mean that gig-workers can be less reliable (for example, by not showing up on time) and less experienced at the task at hand (for example, by not knowing the details of doing a delivery at a customer location), thereby forcing firms to increase their supply buffers in order to ensure a given service level in their operations. These problems can jeopardize firm viability. They reveal that firms have yet to develop sustainable models for the governance and control of the workforce on whose labor they rely.

Workers participating in the sharing platform economy also face distinctive challenges that differ from those of “traditional” paid employees (as described in Chapter 6). Many workers are attracted to platform work by the possibility of more autonomy over work schedules and greater freedom from supervision. However, the terms of their employment may be obscure. For example, transportation workers must “accept” jobs without knowledge of the destinations. From the perspective of the gig-worker, on top of not having the benefits granted to an employee, this type of work arrangement generates significant anxiety from not knowing the actual income that an intended number of work hours will generate. In addition, gig-workers must satisfy the conditions of reputational management systems in order to avoid “de-activation,” even though such conditions often are unknown to them (as described in Chapter 5).

Another clear danger for workers is that the expansion of platform-based work may open up significant gaps in the social safety net, since platforms seldom provide access to health or retirement insurance and platform workers are ineligible for protections under labor standards and minimum wage laws. Collective action is the traditional approach to balancing these information, economic, and social asymmetries, either through formal labor unions or through informal information sharing. Because platform workers typically are contractors, and not employees, however, they are limited in their ability to unionize. In addition, in traditional workplaces, informal worker collectives result from conversations “around the water cooler.” However, in the sharing economy, which lacks a physical workplace, these conversations come at a greater cost and often are relegated to online forums. Thus, one of the main questions for comprehensive reengineering of the sharing economy is the question of how to determine the optimal conditions of work and the regulatory actions and protection that need to be taken to ensure those conditions, as discussed in Chapter 7 on regulation and Chapter 6 on labor and work.

Urban and state governments, too, face unforeseen challenges from the platform revolution. Since platforms represent new forms of business for which decades-old regulations were not designed,⁶ platforms can often operate in an unregulated space, free of the dictates that constrain their more traditional competitors. Typically, city governments lack the most basic information about platform firm operations, even though the latter have major consequences for the transportation, housing, and employment systems on which the public relies.

More generally, the sharing economy has generated important gaps in the flow of information that is vital to the interests of workers, governments, and firms. For example, large-scale proprietary information generated by ride-hailing platforms such as Lyft is valuable. As a result, firms rarely share such information with regulators, who could use it to better understand the effect of the firm on the public. Nor do firms share such information with workers, who could use it to make career and daily employment decisions. Ironically, firms themselves suffer from information gaps, since they typically lack access to information about the long-term well-being of the workers who provide the lion’s share of their service. As some of our contributors have discussed, new research methodologies can help produce, disseminate, analyze, and share information previously unavailable about the sharing economy, which in turn should help improve market efficiencies, reduce labor market uncertainty, and support proactive regulatory structures, thereby strengthening the entire sharing economy ecosystem.

The multiple observations in this volume about the nature of work in the sharing economy teach us a crucial lesson: Comprehensive optimization of work conditions by platform owners, workers, and regulators should be one of the core concerns of reengineering the sharing economy. Currently, platform owners optimize for efficiency, growth, and profit through the design of their matching and pricing algorithms. Government regulators optimize for the public good through regulations. Workers, as suggested by Hall and Krueger (2018)⁷ and Schor et al. (2020),⁸ currently optimize for both income and flexibility. However, it is not clear if each stakeholder optimizing myopically without a systems perspective of the entire ecosystem can possibly achieve the desired outcomes (see Chapter 2).

Looking ahead to the prospect of reengineering the sharing economy, there remain important open questions for all stakeholders related to the future of work. From the perspective of the firm, many sharing economy platform companies struggle to be profitable even after operating for years with significant market shares, raising questions about the sustainability of the business model as it concerns firms’ relationship with their workforce. This relationship impacts several aspects of a firm’s profitability including how it recruits, maintains, and pays its workforce and how the firm’s operational efficiency is affected by issues such as workers’ hours, dependability, and professionalism. In turn, workers are low-paid and lack meaningful control over working conditions and data (see Chapters 5 and 6). Finally, regulatory authorities have thus far had little success at regulating the underlying business activity and service delivery that sharing platforms make possible, in turn limiting their ability to constrain negative effects on the public good, including the future of work (see Chapter 7). While this limited regulatory success is partially due to the strong lobbying efforts of sharing economy firms, such as in the case of the ballot defeat of Proposition 22 in California (discussed in Chapter 6), in other cases it is due to the challenge of anticipating the externalities that will be caused by the regulation itself. In this complex ecosystem, regulatory and other actions targeting part of the system may broadly impact consumer behavior or workforce dynamics in unintended ways, potentially causing more harm than good, as discussed in Chapter 4. This observation emphasizes the point by Duman, Ergun, and Behroozi (Chapter 12) as well as Heydari (Chapter 2) that a comprehensive analysis of the nature of work in sharing ecosystems is crucial.

13.5 Equity and Access

Although this volume is by no means the first to emphasize the significance of equity considerations in the sharing economy, it does resoundingly affirm equity’s centrality. Indeed, equity is a core theme in many of the analyses contributed by our authors, though regularly only implicitly so. These analyses provide rich detail about the range of equity-related harms and benefits that have occurred in sharing economy markets. They also provide significant information and inspiration for creating a more equitable sharing economy. Before reviewing the lessons learned from this volume about equity, it is important to consider how equity is defined – and how it manifests – in the sharing economy.

Plainly, one crucial vein of concern and analysis that invokes equity considerations relates to race and racial relations. Because the focus of this book is the American sharing economy, analysis of racial equity in the sharing economy could not be a more pressing matter. As Dyal-Chand observes in Chapter 7, the COVID-19 pandemic is not the only pandemic that has plagued the United States for many months now. Racial violence has also reemerged as a crisis that demands cross-disciplinary analysis and response. Not surprisingly then, concerns about racial equity surface throughout this volume. Schor and Vallas describe the emergence of a “third, implicitly racialized employment status,” between independent contractor and employee – a status that is both unequal and “substandard” in the level of protections and value that it affords workers who have it. Dyal-Chand discusses the dawning recognition among those who study the sharing economy that at least some proprietors of sharing platforms seem to be developing their businesses in a direction that capitalizes on the racist results produced by their algorithms. More implicitly, both Chapters 10 and 11 raise troubling questions about the racialized effects of sharing innovations intended for (often commendable) purposes such as providing greater access to goods and services within neighborhoods and the democratization of energy production and control.

Issues of equity and equality also arise with respect to gender, disability, and other identity categories. Research on whether sharing economy platforms discriminate on the basis of gender has only scratched the surface, and this is reflected in the contributions to this volume. Yet it is also apparent that many of the questions raised by the research on disparate racial impact also necessitate a robust research agenda concerning other disparate effects. These effects will no doubt be different from the effects of racism within the sharing economy, but the research on race in the sharing economy can provide helpful clues to guide additional research.

This volume also overwhelmingly makes the case that equity within the sharing economy is defined by level of income and wealth. For example, Heydari’s proposal of a sociotechnical examination of the many positive and negative externalities produced by sharing platforms provides an analytical perspective that reveals the hidden burdens and benefits that depend partly on the wealth of sharing economy participants. Focusing their analytical lens on the increasingly ubiquitous mobility industry, Koutsopoulos, Ma, and Zahedi provide nuanced information about the differential impact of ride-sharing innovations in the first generation of mobility platforms. While their attention is on reducing congestion, increasing sustainability, and improving the profitability of mobility companies, the detailed innovations they propose also provide a template for achieving more equitable access to mobility platforms by consumers with lower incomes and less access to traditional goods and services such as privately owned cars and taxis.

The contributions in this volume additionally make clear that equity concerns exist on both the demand and supply sides of sharing platforms. On the demand side, the analyses by Koutsopoulos, Ma, and Zahedi as well as O’Brien, Heydari, and Ke provide deep empirical insights into how access to first-generation sharing platforms can vary for consumers by neighborhood, income, and other demographics. Dyal-Chand describes the proliferating literature on the genderized and racialized consumer harms wrought by first-generation sharing platforms. Kane, Allen, Si, and Stephens raise similar concerns in the new frontier of energy sharing. As the discussion by Lambillotte and Bart suggests, such fundamental concerns as privacy may intersect in significant ways with the axis of equity.

On the supply side, Schor and Vallas raise deeply troubling questions about the future of equitable work, especially for the sharing economy workers who rely on platform jobs as their primary source of income. Such concerns are amplified when considered in contexts such as last mile delivery (see Chapter 12) and the development of clean energy systems (see Chapter 11). These chapters provide the detailed examples for the conclusion reached by Heydari and Dyal-Chand in their chapters that sharing platforms have been able to develop in a regulatory environment that does not constrain platform proprietors in their treatment of those who provide goods and services through those platforms.

The analyses in this book thus present a vexing puzzle: On the one hand, sharing economy platforms maximize opportunities for maintaining anonymity and for sharing the value of expensive goods and services. Through technology, such platforms reduce the costs of market entry and exit by making it easy and cheap to provide – and also to access – goods and services. They significantly increase access to information at a very low cost. In short, the sharing economy should be a means of equalizing access to an enormous range of markets. Yet, on the other hand, many of these very platforms have innovated in ways that allow proprietors and suppliers to differentiate – and outright discriminate – on the basis of race, gender, disability, income, and other characteristics. In so doing, these platforms have limited access to sharing economy participation on the basis of criteria that should have been rendered invisible and irrelevant by platform technology. They have regularly contributed to inequity rather than increasing equity.

While this volume has contributed to the conversation about equity by providing important empirical and interdisciplinary evidence of this puzzling phenomenon, it has also contributed to a basic diagnosis. As the scholars in this volume have described from a range of disciplinary perspectives, the proprietors of sharing economy platforms innovate in directions that optimize for their priorities. The first generation of sharing economy platforms have overwhelmingly optimized for fast growth, and more broadly, profit. In the course of doing so, they have produced a range of positive and negative externalities, some of them startling. These externalities teach us important lessons about the multiple impacts of the sharing economy – and its potential for achieving equity among other things. Yet, realizing this potential requires more deliberate and concerted action. In short, the current state of the sharing economy demands a rebalancing in the direction of greater equity. Whether by choice, by mandate, or by some combination of the two, such a rebalancing can only occur if platform proprietors optimize for equity in addition to growth and profit.

Moreover, the diagnosis that emerges from this volume makes clear that the problem of inequity in the sharing economy is deeply systemic in nature. Currently, market design, industry practices, and law all provide ample space for sharing platform proprietors to make their own choices about goals, priorities, and innovations, including those that increase inequity. For example, the design of sharing platforms provides ample opportunities to innovate new forms of business transactions that capitalize on reputation and trust. As Tadelis describes, such innovations are exciting and disruptive, allowing a broad range of participants in the sharing economy to rely on new forms of information and new business methods. Transparent rating systems allow suppliers of services on sharing platforms to develop good will rapidly and efficaciously, as compared to traditional businesses. Yet, as other contributors point out, these very forms of market design also can reduce equity by eliminating anonymity and thereby reinstating the ability to discriminate on the basis of race, geography, wealth, and other criteria. Industry practices can exacerbate such effects. By leveraging just such design mechanisms, sharing platforms can use surge pricing and other methods to take advantage of unequal access by consumers. As Schor and Vallas discuss, they can also increase the precarity of low-wage workers who depend on sharing platforms for meaningful income.

Currently also, as the chapters by Heydari, Dyal-Chand, and others discuss, law creates ample space for innovation in market design and industry behavior without systemic analysis of the connection between such behavior and equity considerations. Powerful intellectual property rights, contracts of adhesion, weak labor and employment laws, piecemeal and reactive regulations, and lack of political will or even direction in protecting widespread access to sharing platforms at times combine to nurture and even valorize disruption at the expense of necessary protections.

Crucially, the contributions in this volume have supplemented these diagnostic insights by enhancing our understanding of a range of possible solutions to the problem of rising inequity in the sharing economy. One of the most important messages from the volume as a whole is that, because of the multiple sources for inequitable development and operation within the sharing economy, the solutions must also be cross-disciplinary. To examine the potential of cross-disciplinary solutions to address inequity in the sharing economy, consider one set of solutions that has come to the fore in this volume, and indeed that invokes the title of this volume. Specifically, consider the potential that some of the necessary regulations of the sharing economy may be best imposed by means of the engineering of the platforms. In other words, the concept of “regulation by design,” which has been the subject of much scholarship in the privacy domain, may also be a valuable form of regulation for the purpose of prioritizing equity.⁹

As the analyses in this volume suggest, the incorporation of regulation into the design of sharing platforms would require at least two indispensable ingredients. First, a certain level of what has been described as “self-regulation” would be required. Self-regulation might originate in the design choices made by businesses within a sector that choose to maximize value for different stakeholders, beyond profit, growth, or efficiency. For example, recognizing its role as a necessary component of the transportation infrastructure (especially during crises or other periods when public transportation is disrupted), a ride-sharing company could choose a more socially responsible pricing structure that would reduce or at least stabilize prices during times of crises, instead compensating for this lost opportunity by charging higher prices in business districts or other geographic regions where riders could expect to be subsidized by their employers or would have incomes high enough to support paying higher prices. Just as some platforms have already marketed their products and services on the basis of their greater contributions to environmental sustainability (see Chapter 3), or consumer safety,¹⁰ such a business could distinguish itself in the market on such grounds.

In addition, government would have an important role in developing and maintaining this kind of self-regulation. In the ride-sharing pricing example just provided, it is possible that the hypothetical company could achieve market success by means of this combination of social responsibility and pricing differentiation, and it is even possible that it could begin a “race to the top,” rousing other companies to explore the benefits (and costs) of fulfilling their function as a necessary part of an urban transportation infrastructure. However, given the many apparent incentives toward monopolization, it is likely that such moves would need to be encouraged by governmental involvement. Fortunately, there already exist numerous regulatory models from which regulators could draw. One interesting model is the development of “certificates of trust,” which could originate either within an industry or with a governmental agency.¹¹ Another example could be for regulators to lead the development of an industry-wide code of conduct for platform design. Another would be to provide design guidelines such as those issued by regulatory agencies to ensure compliance with the Americans with Disability Act and similar federal and state laws.¹² Certainly, also, it may be appropriate for regulators to at times require the incorporation of certain design standards that would optimize for one or more of these principles.

As Chapter 2 makes clear, this kind of coordination between platform design and regulatory design would require a sociotechnical approach that could account for a broad range of positive and negative externalities, design characteristics, and individual and group behaviors on both the supply and demand sides of sharing platforms. The analyses in the chapters on sharing in the neighborhood, sharing and sustainability, sharing and last-mile delivery, and sharing energy all provide vivid examples of the need for system-wide analysis in engineering self-regulatory approaches.

This, however, would be just the beginning. While such an approach to reengineering a more equitable sharing economy holds much promise, crucial questions will arise and will require interdisciplinary research and analysis. Three sets of questions seem particularly salient on the equity front. None can be answered on the basis of the research presented in this volume, though much of this research certainly lays the foundation for an ambitious forward-looking research agenda.

First, how could platforms be inspired to choose to self-regulate in the direction of greater equity? Given the extraordinary impetus to optimize for immediate fast growth – and the reality that many platforms have yet to achieve any meaningful profit while taking in vast amounts of venture and other private capital – how could platforms be motivated to race to the top in designing equitable industry practices? This question seems particularly salient in light of the extraordinary level of social and political polarization in US society (and many other societies) today. Indeed, it is reasonable at least to wonder whether the type of “regulatory entrepreneurship” described by Pollman et al. is possible partly because of such polarization.¹³ If so, then the challenge of solving the contemporary polarization in wealth depends partly on solving these other forms of polarization as well – a daunting task.

Second, while the benefits of governmental regulation to promote reengineering of a more equitable sharing economy may be apparent, it is also important to consider the costs of such regulatory interventions. One such cost could be social and political backlash, thereby leading to even greater polarization. Such a counterproductive result would be to no one’s advantage. A second cost could be the potentially high level of investment required to achieve a regulatory approach that is responsive, thoughtful, and sophisticated enough to nurture industry-led design that could successfully achieve equity over the long-term. Unfortunately, our regulatory history has produced too many examples of analogous regulatory failures, despite the good intentions behind them. A third obvious cost is that regulation could delay and stunt positive industry innovations as well as profits, thereby harming the very individuals and groups, such as low-wage workers and consumers of color, that the regulations would be intended to help. These are serious concerns, and they demand careful attention going forward.

Finally, it will be important for future research and analyses of equity – and the possibility of engineering for equity – to consider the implications for sharing economy governance more broadly. All of the complicating factors just described and many others, including political and other forms of polarization and the globalized nature of the sharing economy, also complicate the prospects of stable governance. The multiple and varied examples of sharing economy platforms provide fertile ground for further research about governance. For example, as the chapters on sharing delivery systems, energy, and mobility discuss, genuinely peer-to-peer platforms have addressed a range of access and even equity issues. Heydari and Dyal-Chand both raise questions about whether these and other examples could point us toward deeper examinations of the emerging democratic principles in some sharing economy contexts.¹⁴ Here again, the contributors to this volume have raised significant questions that deserve further research attention.

13.6 Reengineering Sharing: What Lies Ahead?

In addition to these core dimensions that define the challenge of reengineering a more just sharing economy, the contributions in this volume have raised a set of questions that might best be described as more philosophical, epistemic, or even existential in nature. These questions present some of the most difficult and complex challenges of all. Yet we believe that it is only wise for those involved actively in reengineering the sharing economy to reflect on these questions, and in so doing, to make their best efforts to proactively address them. While we make no pretense to answer these questions, we conclude this chapter by raising two salient areas of necessary exploration.

13.6.1 How Have Platforms Contributed to Globalization?

The relationship between platforms and globalization is deep and diffuse. Indeed, at this stage, we can only raise more specific questions about this relationship in an effort to define its contours. On the labor side, for example, how has platform work affected patterns of migration and immigration? In what ways has platform access replicated patterns of discrimination, colonialist behavior, and nationalisms, and in what ways has it disrupted those patterns?

Some of these questions are relevant on the consumer side as well. Additional questions also arise: How have consumers benefited from accessing platforms across borders? On the other hand, in what ways have they assumed greater risk?

Finally, a number of crucial questions arise for businesses and those who govern them. For example, how have platform-based businesses responded to taxation, and more generally to other laws that depend in meaningful measure on physical location within a territory? These are just some of the many, many questions inspired by the connection between platforms and globalization.

13.6.2 Who has the Right to Govern the Sharing Economy?

Finally, and relatedly, one of the most vexing set of questions moving forward will no doubt concern platform governance. The diffuse, indeed globalized, nature of platforms deeply impacts the question of governance for the obvious reason that it raises foundational questions about who has the right to govern the platform economy, or any given piece of it. While this question at times feels rhetorical, especially in light of claims that the Internet is too diffuse a phenomenon to be governable, it remains imperative to search for a more substantive answer in response to such claims.

On this question, the editors of this volume have a clear normative position: As we, and many of our contributors, have expressed, we believe it is imperative to develop processes, structures, and norms that move the sharing economy in the direction of genuinely democratic governance. While such a statement is rhetorically powerful, it is also rhetorically straightforward. It will of course be much more difficult to operationalize this statement. Doing so will require serious attention to the current power imbalance between platform owners on the one hand and consumers, workers, and even regulators on the other. It will require reconsideration of intellectual property rights and other legal and market structures that perpetuate this power imbalance. Moreover, just as is the case with any political democracy, it will require vigilance and nurturing over the long term.

Yet we believe that it will be imperative to engage in just such an effort as we seek to optimize for a more just sharing economy. We hope that the analyses in this volume have provided both information and inspiration for future work in this direction.