Fostering unassisted off-hour deliveries: The role of incentives

Transportation Research Part A xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Transportation Research Part A journal homepage: www.elsevier.com/locate/tra

Fostering unassisted off-hour deliveries: The role of incentives José Holguín-Veras a,⇑, Xiaokun (Cara) Wang b, Iván Sánchez-Díaz c, Shama Campbell b, Stacey D. Hodge d, Miguel Jaller e, Jeffrey Wojtowicz f a Center of Excellence for Sustainable Urban Freight Systems (COE-SUFS), Department of Civil and Environmental Engineering, Rensselaer Polytechnic Institute, 110 8th St., Troy, NY 12180, USA b Department of Civil and Environmental Engineering, Rensselaer Polytechnic Institute, 110 Eighth St., Troy, NY 12180, USA c Department of Technology Management and Economics, Chalmers University of Technology, Vera Sandbergs Allé 8, Göteborg 41296, Sweden d Office of Freight Mobility, Division of Traffic and Planning, New York City Department of Transportation, 55 Water Street, 9th Floor, New York, NY 10041, USA e Department of Civil and Environmental Engineering, University of California, One Shields Avenue, Davis, CA 95616, USA f Center for Infrastructure, Transportation, and the Environment, Rensselaer Polytechnic Institute, 110 Eighth St., Troy, NY 12180, USA

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: Off-hour deliveries Incentives Off-peak deliveries Freight demand management Freight behavior Unassisted deliveries

a b s t r a c t This paper describes the chief findings of research conducted to assess the willingness of receivers of supplies to accept unassisted off-hour deliveries (U-OHD), which are those conducted outside regular business hours and without the assistance of the receiving establishment staff. U-OHD have potential to increase economic competitiveness, reduce congestion, improve environmental conditions, enhance livability, and increase quality of life in urban areas. This study considers the role that public policy initiatives could play in fostering receivers’ acceptance of U-OHD by analyzing survey data collected from potential U-OHD adopters. The paper describes the survey conducted, performs descriptive analyses of the data, analyzes the respondents’ stated willingness to participate in unassisted off-hour deliveries, estimates discrete choice models to gain insight into receivers’ decision-making processes, and analyzes the effectiveness of alternative policy scenarios. It is found that a number of policy levers can foster U-OHD: (1) public sector provision of a one-time incentive, a public recognition program, and business support services; (2) carriers providing shipping discounts to receivers of U-OHD; and (3) the creation of a Trusted Vendor Certification Program. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Increasing urban congestion puts great pressure on decision makers and transportation agencies that must also be responsive to citizens’ desires for vibrant urban economies, enhanced livability, and high quality of life. This quest has led to the identification, design, and testing of a number of initiatives designed to foster greater efficiency of freight systems. One type of initiatives are the receiver-centered Freight Demand Management (FDM) measures which attempt to change the nature of the demand for cargo. These policies take advantage of the fact that the receivers—by virtue of being the primary customers in the economic transaction, or the ones that create the demand—have a great deal of power over vendors and carriers, who must respect their wishes if they want to stay in business. Examples of FDM include encouraging receivers to reduce the number of deliveries they get, or to retime some or all deliveries (Holguín-Veras and Sánchez-Díaz, 2016). ⇑ Corresponding author. E-mail addresses: [email protected] (J. Holguín-Veras), [email protected] (Xiaokun (Cara) Wang), [email protected] (I. Sánchez-Díaz), campbs4@rpi. edu (S. Campbell), [email protected] (S.D. Hodge), [email protected] (M. Jaller), [email protected] (J. Wojtowicz). http://dx.doi.org/10.1016/j.tra.2017.04.005 0965-8564/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

2

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

Regrettably, the potential role of FDM has been largely overlooked. This, in turn, presents a unique opportunity to reduce congestion and improve economic competitiveness, by developing and implementing FDM programs (Holguín-Veras and Sánchez-Díaz, 2016). Achieving this goal requires gaining insight into the behavioral underpinnings of the urban freight system, and how best to influence it to minimize its negative externalities, without detracting from its contributions to the economy and quality of life. In recent years, there has been an increase of interest in Off-Hour Deliveries (OHD), a particular form of FDM where the main intent is to induce receivers to accept deliveries outside regular business hours. To some extent, the interest in OHD is the result of its successful pilot testing and implementation in New York City (NYC), which demonstrated that the use of incentives and collaborative private-sector engagement could induce large numbers of receivers to accept OHD. Without doubt, OHD is an attractive option, that: can be implemented in partnership with the private sector; does not require large expenditures in capital programs; reduces conflicts with other road users (e.g., pedestrians, bicyclists, and buses); facilitates the implementation of bike lanes and Bus Rapid Transit by reducing the need for loading spaces at the curb; improves economic productivity; reduces pollution and congestion; and enhances livability and quality of life. However, as is typical of a new concept, multiple aspects deserve further research, including the behavioral response of receivers to OHD adoption under the influence of public-sector policies. This paper seeks to contribute to the nascent field of FDM through the behavioral investigation of factors that could foster receivers’ acceptance of unassisted OHD, (U-OHD). Unassisted OHD is a form of OHD whereby the receiving establishment has no staff present at the time of delivery. To this effect, the authors collected stated preference data to estimate behavioral models and ascertain the role that public-sector policies could play in fostering acceptance of U-OHD. The paper has four sections in addition to this introduction. Section 2 summarizes the background of this research. Section 3 briefly describes the survey conducted, and provides a descriptive analysis of the data. Section 4 presents the behavioral models estimated. Section 5 discusses policy implications and provides concluding remarks.

2. Background OHD programs have been shown to help urban economies reach their most efficient outcomes, providing greater supply chain efficiencies, as well as environmental and quality of life benefits. There is indeed ample evidence—see Yannis et al. (2006) and Holguín-Veras et al. (2011)—to indicate that the economic welfare of urban areas would improve if the market share of OHD increases. Estimates suggest that for New York City the optimal amount of staffed OHD (S-OHD) is in the range of 14–21% of the total number of deliveries, depending on the composite value of time and the traffic composition (HolguínVeras et al., 2011). In the case of U-OHD, optimal participation is probably around 40% (Holguín-Veras et al., 2012). These numbers stand in contrast with the current market share of OHD of about 4–5% in New York City area (Holguín-Veras et al., 2007) and the Spanish cities of Santander and Barcelona (Domínguez et al., 2012). Thus, public-sector intervention is needed to increase OHD, especially U-HOD, as a market failure prevents the system from reaching the most efficient outcome (Holguín-Veras, 2011). In acknowledging the potential merit of greater OHD market share, it is important to identify the conditions that are preventing its wider acceptance. A combination of factors is to blame. In most urban freight markets, receivers have a great deal of power in their dealings with carriers and vendors. As established in Holguín-Veras (2008) and Holguín-Veras (2011), receiver opposition—together with the power that receivers have in determining delivery times, and the inability of carriers to compensate the receivers—act together to create the market failure that prevents an increase in OHD market share. A number of approaches can be considered to remove the market failure: regulations banning deliveries in the congested hours of the day, freight road pricing (e.g., cordon time of day tolls, time-distance pricing), charging receivers for deliveries received in the congested hours, and providing incentives to receivers to induce them to accept OHD (Holguín-Veras, 2008; Holguín-Veras et al., 2011, in press). The first two are carrier-centered, while the last two are receiver-centered approaches. Regulations could be used to promote OHD, as practiced in a number of Chinese cities (Changsha Bureau of Public Security, 2013; Shenzhen Bureau of Public Security, 2013; Beijing Traffic Management Bureau, 2014). In these cases, public-sector agencies issued regulations requiring selected industry sectors to perform OHD. Experience has shown that these programs are frequently counterproductive. The receivers oppose the mandate, because it forces them to accept OHD and pay for either the additional staff costs, or accept additional risk without any form of compensation. Although carriers stand to benefit from the OHD operations, they tend to dislike the mandate because it puts them in an untenable situation. In Beijing, for instance, carriers complain that the receivers exert pressure on them to make deliveries in the regular hours, in violation of the mandate, causing them to get numerous fines (Beijing Traffic Management Bureau, 2010). In fact, managers of parcel delivery companies admit that they use passenger vehicles to make deliveries during the day hours. This decision leads to higher congestion levels because multiple passenger vehicles are needed to do the deliveries that a single truck could transport. This situation illustrates a key limitation of regulatory approaches: they are blunt policy instruments. Requiring that broad segments of the industry perform OHD will impact numerous receivers for which OHD is not the best outcome. Forcing receivers to comply reduces the net economic benefits of the program, as receivers must either absorb the extra costs, or force carriers to make deliveries in smaller vehicles (and even passenger vehicles) or simply absorb extra fines. Although there are social benefits in terms of congestion and pollution reductions, this suboptimal regulation produces unintended negative effects, a government failure, that make things worse. Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

3

Another option is freight road pricing. After all, internalizing the externalities caused by an economic activity may induce behavior change, moving the system towards optimum (Pigou, 1920). The assumption is that charging higher tolls for travel in the congested hours would lead carriers to switch their deliveries to the off-hours, or to reduce the frequency of travel. Unfortunately, this point of view neglects two supremely important facts: (1) carriers travel because they are making deliveries to a receiving establishment; and (2) the receivers are the ones that primarily determine delivery times and frequency of deliveries. In most cases, carriers—the weakest element in the supply chains—cannot force receivers to change behavior. The empirical evidence in this regard—from time-of-day pricing implementations in NYC and London—is compelling; in neither case did higher tolls reduce freight traffic in a meaningful way. See Holguín-Veras et al. (2006, in press). There are theoretical reasons that explain the challenge faced by carrier-centered approaches. More specifically, carriers in competitive markets cannot pass cordon-time-of-day tolls on to their customers because the tolls are a fixed cost that does not enter into the computation of the rates. The empirical data confirm this assertion. After experiencing an increase in daytime tolls of 50%, eight out nine carriers could not pass any toll increases to their customers. Only one out of nine carriers could pass a rather mild increase to their customers Holguín-Veras et al. (2006). Consequently, the transmission of price signals is damped. Carriers could pass on to customers time-distance tolls, though the tolls would need to be very high, and doing so would still not necessarily be feasible politically. Readers interested in the theoretical reasons that explain these admittedly counter-intuitive results, together with the empirical evidence that supports them, are referred to Holguín-Veras et al. (2006, in press) and Holguín-Veras (2008, 2011). Another alternative is to charge receivers for the externalities produced by their deliveries. The internalization of these costs is likely to lead the receivers to reduce the number of deliveries, retime deliveries to the off-peak hours, or accept OHD. By focusing directly on receivers, this alternative may circumvent the problem of damped price signals that limit the effectiveness of pricing. Holguín-Veras and Sánchez-Díaz (2016) used an Economic Order Quantity (EOQ) model (Harris, 1915) to assess the effects of a social optimal receiver charge, equal to the value of the externalities produced by the delivery. They found that the receiver charge could reduce the number of deliveries received by about 10%. However, charging receivers will likely generate considerable opposition from the private sector, opposition that could deter implementation. The contrasting performance between carrier-centered and receiver-centered approaches—explained in detail in Holguín-Veras et al. (in press)—is the result of the asymmetric power relations present in most urban supply chains, whereby receivers influence the behavior of carriers, and not the other way around. The fourth possible approach is to use public policy to incentivize receivers to accept OHD. In response to these policies, potential participants assess the pros and cons and decide whether or not to participate, on the basis of their specific conditions. If OHD is beneficial, the receiver will join the program; otherwise, it will not. The element of choice, based on a business’s particular economic considerations, ensures that voluntary OHD programs always increase economic welfare. Thus, as long as the implementation cost is smaller than the benefits from switching deliveries to the off-hours, increasing OHD will bring about net economic benefits. One of the virtues of a voluntary OHD program is that it eliminates the risk of negative unintended effects associated with what would be viewed as a government failure. Moreover, voluntary OHD programs are easier to sell to the private sector, which facilitates implementation. As should be clear, OHD is a nuanced and complex topic. At the core of the complexity are the impacts on the stakeholders involved, including local communities. The carriers stand to benefit if a sufficient number of receivers participate; the receivers are bound to enjoy increased reliability and lower shipping costs though they may experience an increase in staffing costs if they use staff to accept S-OHD, or a higher perception of risk if U-OHD are adopted. Local communities could benefit from reductions in congestion and pollution, and enhanced quality of life during daytime hours, though they may be impacted by night noise and pollution. The latter concern has led European cities to ban OHD, which has created other problems. The 4–5% of deliveries that naturally would take place in the off-hours are being forced into the day hours, increasing pollution and congestion during the day. For a discussion of European practices, see Browne et al. (2006). The authors’ research indicates that incentives are key to changing receivers’ behavior. Since the incentives remove receiver opposition, and the carriers are generally in favor of U-OHD, entire supply chains could switch to the off-hours. The economic impacts would be considerable. Implementing U-OHD in Manhattan could generate benefits of about $150–$200 million/year (Holguín-Veras et al., 2011), which include substantial reductions in criteria pollutants (HolguínVeras et al., 2013b). It is interesting to note that the bulk of these savings are accrued in the congested regional networks that link warehouses and distribution centers to NYC. Changing the time of travel of delivery tours would therefore improve congestion and environmental conditions in the entire metropolitan region. As part of the authors’ research, a successful pilot test of OHD was conducted in NYC (Brom et al., 2011; Holguín-Veras et al., 2011) using two different modalities of OHD. Half the establishments used S-OHD (with receiving establishment staff present at the time of delivery), while the other half used unassisted OHD, or U-OHD (without staff from the receiving establishment present). These two modalities exemplify the tradeoff between risk and reward. Participants in S-OHD minimize the risk of a negative outcome, and use the financial incentive to pay for staff to be present at the time of delivery. In SOHD, the incentive only generates a small profit to the receiver because the bulk of it is used to pay the staff. In contrast, in U-OHD, receivers take the risk and allow their vendors to enter their establishments unsupervised. If nothing goes wrong, the incentive becomes profit. The pilot test confirmed the chief findings of the research conducted before the pilot (Holguín-Veras et al., 2007, 2008; Holguín-Veras, 2008, 2011), but equally important, the pilot produced important findings on its own: U-OHD could play a key role in an FDM strategy; OHD could be conducted with minimal or no disturbance to local communities (no noise Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

4

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

complaints were received during the pilot); OHD are more reliable than regular-hour deliveries; and, U-OHD could be safely performed without putting drivers or establishments at risk. The pilot also revealed that, if delivery errors are made in U-OHD they could be remedied early in the day with minimal interruptions. In contrast, errors in regular-hour-deliveries found later in the morning cannot be remedied until the afternoon, which disrupts business operations and requires a larger safety stock. However, the most important finding of the pilot test was related to the permanence of the behavior changes. At the end of the pilot, the receivers that used S-OHD reverted to regular-hour-deliveries, because without incentives, they could not afford OHD. In contrast, the vast majority of the receivers doing U-OHD opted to stay in the program. When asked why they did so, the managers indicated that the convenience and reliability were such that they found it worthwhile to continue with U-OHD. Instead of uncertain delivery times because of congestion—which forced them to maintain a safety stock of supplies—U-OHD guaranteed that the supplies would be in place when the business opens. Once receivers tried U-OHD, and ‘‘nothing bad” happened, they were able to benefit from: superior reliability, fresher products, reduced inventories, and more efficient use of their staff. The success of the pilot, widely reported in the press (Cassidy, 2009, 2010; Wall Street Journal, 2010), led the United States Department of Transportation and the New York City Department of Transportation (NYCDOT) to launch an implementation phase (expected to take three to five years), and to the inclusion of OHD in NYC’s long-term plans (City of New York, 2011, 2015). Moreover, the Federal Highway Administration and the Environmental Protection Agency created a program to foster OHD in other cities (Federal Highway Administration, 2012), awarding the first grants to Orlando, FL and Washington, DC. Other cities, including Chicago, Boston, Atlanta, London, and Toronto also expressed interest in OHD programs. Sao Paulo (Brazil), Bogota (Colombia), Copenhagen (Denmark) and Brussels (Belgium) have conducted successful pilot tests with results that are in general agreement with the NYC experience. In the case of the European trials, relaxation of the night delivery bans are being considered. For typical results see (Centro de Inovação em Sistemas Logísticos, 2015). The research reported in this paper was a central component of the project’s implementation design (Holguín-Veras et al., 2013b). The behavioral models discussed here were incorporated into a Behavioral Micro-Simulation (Silas and Holguín-Veras, 2009; Holguín-Veras and Aros-Vera, 2014), which models the response of carriers and receivers to public policy. The key findings from these analyses, and input from NYCDOT’s Industry Advisory Group, helped determine the incentives to be provided, and the industry sectors that should be targeted. During the course of the 2012–2014 phase of the research project, more than 400 establishments shifted deliveries to the off-hours (Holguín-Veras et al., 2013b, 130). It is worthy of mention that about 50% of the receivers switched to U-OHD without receiving a financial incentive. The authors’ conjecture is that these participants decided to accept U-OHD on their own, after seeing other businesses have a positive experience with U-OHD. This is consistent with key findings from behavioral economics about network effects that induce individuals to act like the network of peers they are part of (Brown and Reingen, 1987). Further research should identify the reasons for their decision.

3. Descriptive analyses To gain insight into the effectiveness of public-sector policies to foster U-OHD, the team collected stated preference data and estimated behavioral models. The data were collected using a random sample of 263 receivers in Manhattan. As a whole, the sample resembles the statistical distribution of establishments in Manhattan. For analysis purposes, 2-digit North American Industry Classification System (NAICS) codes were used to group the data. The largest group (40.68%) is in the retail trade (NAICS 44-45); ‘‘Other Services” (NAICS 81) such as laundry and dry cleaning businesses, spas, fitness centers, pet grooming, and pharmacies comprise 20.91% of the sample; 16.73% are in manufacturing (NAICS 31-33); another 10.65% do wholesale trade (NAICS 42); arts, entertainment, recreation (NAICS 71) represent 5.7%; accommodations and food services (NAICS 72) 3.42%; and 1.90% are in transportation and warehousing (NAICS 48-49), finance and insurance (NAICS 72), and health care and social assistance (NAICS 62). The results related to establishment characteristics are based on the number of establishments in the sample (263), while the stated preference results are based on the total number of responses (depending on the variable used, the number ranges from 1540 to 1580 responses). This section summarizes the most salient features of the data. The survey collected data on deliveries and shipments received, current operations and flexibility to change, and willingness to participate in U-OHD in response to a set of incentives. The delivery section included questions about number of deliveries received, delivery costs, shipment size, type of goods received, primary transporter of the goods, form of access to building, and number of vendors. The section on operations and flexibility inquired about hours of operations, number of employees, who needs to provide consent for the establishment to accept OHD, interest in consolidating shipments, interest in OHD, and if they currently have a vendor they would trust to do OHD. In the stated preference section, respondents were presented with six randomized scenarios involving various combinations of incentives, and asked to indicate their willingness to accept U-OHD. The choices given to respondents were: ‘‘not at all willing,” ‘‘not too willing,” ‘‘neutral,” ‘‘somewhat willing,” or ‘‘very willing.” The incentives considered were a one-time-incentive ranging from $1000 to $9000, a carrier discount ranging from 0% to 50%, public recognition (or none), and business support services (or none). The one-time-incentive and the shipping discount provided by the carriers were considered because previous research found them to be influential in convincing receivers to accept OHD (Holguín-Veras et al., 2007). The use of public recognition and business support services requires some discussion. Public recognition refers to the use of awards, a branding logo, and other mechanisms to certify that the receiver is participating in sustainability efforts that are beneficial to the community. Such programs would help businesses raise their Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

5

profile among environmentally-sensitive customers, and enables these customers to identify and patronize businesses that participate in environmentally-friendly practices. This conjecture has led to the creation of recognition and certification programs, such as SmartWay (U.S. Environmental Protection Agency, 2013), the John Connell Awards (Noise Abatement Society, 2013) and driver management and vehicle maintenance certification programs (Transport for London, 2013). However, not much is known about how effective public recognition actually would be in fostering U-OHD. Providing business support services to induce OHD is an idea originally suggested by a member of the NYCDOT’s Industry Advisory Group. In his opinion, the bulk of commercial establishments in urban areas are small businesses that struggle, as their owners do not always have a good command of the wide range of regulations, administrative procedures, legal aspects, and financial aspects associated with doing business in a complex urban environment. Should public recognition and business support services be found effective in inducing U-OHD, public-sector support for these initiatives may be justified. These programs have other notable benefits, as they: convey to the private sector a message of support from relevant public agencies, could indeed foster economic activity and efficiency, and create an environment where public and private sectors could cooperate. These considerations speak in favor of implementing public recognition and business support services programs. 3.1. General results The employment data provided by the respondents were analyzed. To estimate total manpower, part-time employees were converted to equivalent full-time employees using a conversion factor of 0.45. Total employment (measured in fulltime equivalents, or FTE) ranges from one to 422.50. The number of deliveries received in a week ranges from one to 500. The average number of deliveries/week received is 18.96, with a standard deviation of 40.14. The commodities received were classified into two groups: perishable and nonperishable. Receivers of perishable goods (i.e., agriculture, forestry, fishing, or food) are 13.31% of the total, while 86.69% only receive nonperishable items. In terms of costs per delivery, the largest is $1500 and the lowest is zero (i.e., receivers do not pay separate delivery costs), with an average delivery cost $139.60. An estimated 6.11% transport their own goods, 59.96% use vendors, and 34.73% use both the company’s transportation and vendors. The number of vendors ranges between one and 1000, with an average number of vendors 27.71. For operation hours, a significant portion of businesses (32.32%) operate, at some point, during the off-hours, either in the morning or at night. The survey asked respondents if they have a vendor they trust to perform U-OHD unsupervised. The data show that 32.02% do, while 67.98% do not. The significance of this result is difficult to overstate, as it indicates that a sizable proportion of commercial establishments do not have any major impediment to accepting U-OHD. In terms of consent to do U-OHD work, 48.58% indicated that they do not need approval from anyone else; the survey also asked about willingness to consolidate shipments, and the number of truck trips generated by the establishment. The data indicate that 20.63% of the companies are interested in consolidating shipments, and an estimate of 19.69% of the sample have considered doing or are already doing OHD. As to what any city agency could do to foster participation in OHD, the responses indicate that the majority (82.46%) believe that there is nothing that city agencies can do. However, a solid 17.54% suggested that city agencies could: address parking regulations for freight vehicles, provide incentives, develop relationships with landlord/building management/companies, decrease tolls, and ‘‘others”. The parking issues mentioned include lack of parking, strict regulations and restrictions, and fines. 3.2. Willingness to accept U-OHD This sub-section provides an overview of the willingness to participate in U-OHD in relation to business characteristics and policy variables, on the basis of the aggregated responses for the various scenarios. The resulting values provide an indication of the average responses for the wide range of values of the policy variables. The categories used in the analyses are shown on the horizontal axis (the percent shown are the frequency of cases for that particular group). The vertical axis shows the results in decreasing order of willingness. To facilitate interpretation, lines have been drawn to separate the results between ‘‘willing” (‘‘somewhat willing” and ‘‘very willing”), neutral, and ‘‘unwilling” (‘‘not too willing” and ‘‘not at all willing”), which are the terms used throughout the paper. The breakdown of responses concerning willingness to do U-OHD by industry group is presented in Fig. 1. As shown, establishments in ‘‘Other Services” (e.g., laundry/dry cleaning businesses, spas, fitness centers, pet grooming, and pharmacies) have the highest willingness (36.21%) to accept U-OHD, with accommodations and food services and retail trade closely behind with 35.19% and 34.18%, respectively. Manufacturing (28.04%), wholesale trade (27.54%), and arts, entertainment, and recreation (25.55%) industries also showed significant willingness levels. It is worth highlighting the sizable proportion of responses in the ‘‘neutral” category, which ranges between 10% and 40%, as these respondents may be persuaded to accept U-OHD if things go well with the program. This would imply that the potential market for U-OHD, the summation of the ‘‘neutral” and the ‘‘willing”, would be about 50% of the establishments in the NAICS 81, 72, 44-45, 31-33, 42, 71, and 48-49. With respect to number of employees (FTEs), the willingness to participate increases for those establishments with 20 employees or fewer. For those with more than 20 employees there is a significant decrease in willingness. It is worth noting that those that fall in the 21–50 range have a ‘‘very willing” (10.61%) level that is comparable to those with a smaller number of employees, so these should not be excluded when targeting businesses to do U-OHD. However, the willingness to participate for businesses with over 50 employees is significantly smaller than other groups. On the other hand, establishments Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

Willingness to receive UOHD

6

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Not At All Willing Not Too Willing Neutral Somewhat Willing Very Willing 81 72 44-45 31-33 42 71 48-49 62 52 20.92% 3.42% 40.68%16.73%10.65% 5.70% 0.76% 0.76% 0.38% 2-digit NAICS

Fig. 1. Willingness versus 2-digits NAICS. Note: NAICS 81: Other services; NAICS 72: Accommodations/food services; NAICS 44-45: Retail trade; NAICS 3133: Manufacturing; NAICS 42: Wholesale trade; NAICS 71: Arts/entertainment/recreation; NAICS 48-49: Transportation/warehousing; NAICS 62: Healthcare/social assistance; and, NAICS 52: Finance/insurance. Percentages under the X-axis represent the share of responses in the sample.

Willingness to receive UOHD

with over 50 employees had the largest percent (40.48%) of neutral level, willingness which may indicate that there is room to convince them to accept U-OHD. The data were analyzed in relation to the type of commodities received by the businesses, which were classified into perishables and nonperishables. Those that receive perishable goods showed a higher willingness to participate (42.19%), than the receivers of nonperishable commodities (30.27%). This is consistent with the experience of the 2010 pilot, which indicated that receivers of perishables—primarily food stores and restaurants—prefer to receive their supplies early in the morning before the business opens, which results in more reliable delivery times as the trucks travel before peak morning traffic. This result is also in line with Nuzzolo and Comi (2014). Fig. 2 shows that the willingness to accept U-OHD decreases with the number of deliveries received, as receiving a large number of deliveries makes it more complicated to switch operations. The data about working hours were also analyzed. Those businesses that are open during off-hours are more willing to accept U-OHD than those that are not, with 43.06% compared to 37.30%. Companies with long operating hours, such as those that overlap with morning and evening/night off-hour periods and those that have 24-h operations, are very willing to do UOHD. More specifically, businesses that have both morning and evening/night off-hours operations, including 24-h operations, have the highest willingness (43.94%) compared to those that have off-hour operations in only one period (morning or evening/night), or none. Fig. 3 shows willingness in relation to the one-time financial incentive. As shown, there is a steady increase from $1000 (willingness of 20.11%) to $4000 (willingness of 36.46%), where the willingness evens out. For the first range of one-timeincentive, receivers that can switch without major inconvenience would respond positively to the incentive. However, for firms that have to incur high operational costs to participate in the program, a one-time incentive—even when it is higher than $4000—will not offset the costs to participate. As expected, the incentive of $9000 exhibits the largest willingness (38.10%). Another form of incentive is the discount in shipping costs that carriers/vendors may provide if the receivers accept U-OHD. The scenarios considered discount rates ranging from 0% to 50%. The results, shown in Fig. 4 indicate that the largest

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Not At All Willing Not Too Willing

Neutral Somewhat Willing Very Willing

1-5 6-10 11-15 16-20 21-25 26-30 31-50 51+ 39.23% 22.69% 10.77% 6.54% 6.54% 2.31% 6.15% 5.77% Number of deliveries/week

Fig. 2. Willingness versus average number of deliveries received per week. Note: Percentages under the X-axis represent the share of responses in the sample.

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

7

Willimgness to receive UOHD

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Not At All Willing Not Too Willing Neutral Somewhat Willing Very Willing $1000 $2000 $3000 $4000 $5000 $6000 $7000 $8000 $9000 12.27% 11.17% 11.43% 11.75% 10.97% 10.39% 11.23% 9.87% 10.92% One-time incentive ($US)

Willingness to receive UOHD

Fig. 3. Willingness versus one-time-incentive to influence switch to U-OHD. Note: Percentages under the X-axis represent the share of responses in the sample.

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Not At All Willing Not Too Willing Neutral Somewhat Willing

Very Willing 0 14.94%

10 20 30 40 19.94% 17.08% 16.62% 16.56% Carrier discount (in %)

50 14.86%

Fig. 4. Willingness versus carrier discount. Note: Percentages under the X-axis represent the share of responses in the sample.

increase in willingness takes place from zero to the 10% discount (9.53% increase). From that level, the willingness gradually increases up to the 50% discount. The analysis of the willingness to participate in relation to the availability of a trusted vendor revealed that businesses with a trusted vendor are more willing (39.17%) to accept U-OHD than those that do not have one (27.30%). A sizable 58.74% of those without a trusted vendor were not willing to accept U-OHD, with 50.00% ‘‘not at all willing.” Obviously, trust is one of the most influential considerations, because if receivers allow vendors access to their property they become vulnerable to loss or damage of property. The importance of trust is both a challenge and an opportunity. It is a challenge because lack of trust could hamper adoption of U-OHD; it is an opportunity because trust could be gained, as the example of Internet sales clearly demonstrates. At the start of the Internet sales revolution, many thought that Internet sales would not take hold because consumers would not buy things they could not physically inspect. Customer reviews and insurance programs to compensate consumers for losses gradually changed consumer attitudes. The authors believe that receivers could be induced to develop trust in carriers, and allow them to perform U-OHD. Public recognition increased willingness by 3.22%. The inclusion of business support services as an incentive increased willingness to 34.09% from 29.56%. It should be noted that the trends described above are largely qualitative and may not have statistically significant effects. Moreover, some of the business characteristics discussed above could be correlated. It is thus inappropriate to derive policy insights directly from these observations. To disentangle the effects of the various influencing factors and identify their statistical significance, the authors estimated discrete choice models, as discussed in the next section. 4. Modeling results Building on the insight gained from the descriptive analyses, the authors used the stated preference data to estimate behavioral models. These models can provide further insights into receivers’ willingness to accept (WTA), help assess effectiveness of different incentives, and enable the simulation of receivers’ reactions to a given stimulus. This kind of model has been used successfully in the past to assess the response of freight stakeholders towards freight policy. For instance, Holguín-Veras et al. (2007, 2008) analyzed the response of receivers and carriers to a set of policies aimed at fostering Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

8

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

OHD; Marcucci and Danielis (2008) studied carriers’ use of consolidation centers; Domínguez et al. (2012) investigated the role of receivers’ attributes and policy to influence their inclination to use consolidation centers and adopt OHD; HolguínVeras and Sánchez-Díaz (2016) assessed the potential of receiver-led consolidation programs; and Marcucci et al. (2015) studied the preferences of carriers for alternative parking and pricing scenarios. The reasons to use these models are numerous, as they: allow for a formal assessment of the role played by company attributes, characteristics of the alternatives considered, and policy variables; could be used to obtain estimates of market responses for combinations of policy variables; enable the study of the econometric interactions between independent variables; and provide a solid foundation for estimation of elasticities. The insight gained from the models is extremely useful for policy making. The dataset assembled consists of panel data gathered by the stated preference questions described in Section 2. The dependent variable is an ordinal scale that measures willingness to accept U-OHD, while the independent variables are the attributes of the receivers and policy variables. Given the nature of the dependent variable, it is necessary to use ordered choice models to estimate the outcome probabilities. The willingness level is represented by a numerical scale ranging from one to five, with one being the least willing, and five being the most willing. In ordered choice models, the dependent variable (an ordinal scale) is modeled through the joint estimation of a score function—which is a linear combination of independent variables—and a set of parameters, l, (thresholds) that split the domain of the score function in a way that the model replicates the original choice probabilities. The log-likelihood function is used to determine the efficiency of the ordered discrete model; this function is maximized to arrive at the best model. The models belong to the family of Random Utility Models (RUM) (Domencich and McFadden, 1975). In the case of ordered discrete choice modeling, the choice of individual n is explained by a utility (Un) that depends on the attributes of the individuals and the alternatives. The utility is linear, as follows:

U n ¼ Xn b þ en

ð1Þ

where Xn A vector of variables or attributes proper to the observation n Β A vector of estimable parameters en A random disturbance Given the panel structure of the data (a single company provides multiple responses), there is a high risk of correlation between the unobserved attributes. Numerous techniques have been proposed to address this issue (Hess and Rose, 2009; Yáñez et al., 2011), including the use of panel data models, with either fixed effects of random effects, random coefficient models, or a combination of both. The authors decided to use an Ordinal Logit model with random effects because it is a practical, yet conceptually solid, way to account for the correlation caused by repeated choices. Thus, Equation (1) becomes:

U nt ¼ Xn b þ ent þ un

ð2Þ

where t indexes the scenarios for which respondent n provided data. The unobserved heterogeneity un is invariant across all t, capturing individual specific effect (such as taste) that remains the same across observation n’s repeated choices. The model assumes that un is normally distributed with zero mean and standard deviation. The parameters, l(j), 0 < l(1) < l(2) < l(3), are estimated during the modeling process. These parameters, referred to as thresholds, define numerical ranges that indicate the level of willingness, as shown below:

Y nt ¼ 1; Y nt ¼ 2; Y nt ¼ 3; Y nt ¼ 4; Y nt ¼ 5;

if if if if if

U nt 6 0 0 < U nt 6 lð1Þ lð1Þ < U nt 6 lð2Þ lð2Þ < U nt 6 lð3Þ U nt > lð3Þ

ð3Þ

Eq. (3) suggest that higher utility leads to higher willingness level. To determine the probability for each willingness level, the random disturbance (e) is assumed to follow a Gumbel distribution. Therefore, the probability that Unt corresponds to a willingness level can be described by the following equations:

Pnt ð1Þ ¼ Pðynt ¼ 1Þ ¼ Pnt ðjÞ ¼ Pðynt ¼ jÞ ¼

1 1 þ expððX nt b  0Þ=ðj1 þ r2 ÞÞ

1 1  1 þ expððX nt b  lðj  1ÞÞ=ðj1 þ r2 ÞÞ 1 þ expððX nt b  lðj  2ÞÞ=ðj1 þ r2 ÞÞ

Pnt ð5Þ ¼ Pðynt ¼ 5Þ ¼ 1 

1 1 þ expððX nt b  lð3ÞÞ=ð1 þ r2 ÞÞ

ð4Þ ð5Þ ð6Þ

Once the final model was obtained, the elasticities of the continuous variables were calculated using Eq. (7): nt ðjÞ gPxntk ¼

@Pnt ðjÞ Pnt ðjÞ

 @xntk @Pnt ðjÞ xntk ¼  @xntk Pnt ðjÞ xntk

8j 2 W; n 2 N; k 2 K; t 2 T

ð7Þ

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

9

where Pnt(j) is the probability that an individual n presents a willingness level j in scenario t xntk is the variable k for the individual n in scenario t nt ðjÞ gxPntk is the individual direct elasticity of choosing j with respect to variable k in scenario t Elasticities of binary variables are calculated as the arc elasticities, i.e., the percentage of probability changes resulting from the change of the binary variable value, as shown in Equation (8):

Pðynt ¼ jjxntk ¼ 1Þ  Pðynt ¼ jjxntk ¼ 0Þ 8j 2 W; n 2 N; k 2 K; t 2 T Pðynt ¼ jjxntk ¼ 1Þ þ Pðynt ¼ jjxntk ¼ 0Þ 2

gxPnknt ðjÞ ¼ 1

ð8Þ

There are numerous reasons to estimate and analyze the elasticities of choice. First, the dimensionless nature of elasticities enable to use them to identify which variables play the most important role in the probability of choices. Consequently, the analyst could identify which user segments are most inclined to respond in the manner desired by policy makers. Second, their economic nature foster an economic interpretation of the results using the arsenal of economic tools and concepts. Third, the individual (disaggregate) elasticities in Eqs. (7) and (8) could be used to estimate the market elasticity that represent the behavior of conglomerates of users. The latter is particularly important because in most cases, policy makers are interested in the market response to policy, as opposed the response of individual users. In this paper, the elasticities are calculated for each individual at the current variable values. The individual elasticities were used to compute the market elasticity to assess the overall effects of incentives (or attributes) on the willingness to accept U-OHD. For more discussions on how to estimate market elasticities from choice models, see Ben-Akiva and Lerman (2010) and Varian (1992). The next section discusses the results of the model estimation process. 4.1. Final models As part of the modeling process, the authors tested a wide range of models to identify the systematic relation between attributes of the firms, policy variables, and willingness to accept U-OHD. The establishment attributes considered included: industry segment, commodity type, employment, sales, number of vendors, area, transportation costs, among others. However, some of these were discarded because they were not statistically significant, or because their parameters were not conceptually valid. Moreover, to ensure the robustness of industry sector-specific parameters, the authors decided to only include industry segments with more than six observations. This threshold was determined by a sensitivity analysis as the minimum number of observations that leads to robust results. The availability of a trusted vendor deserves special discussion. Although it is an attribute of the firm, not an incentive, the authors’ conjecture is that it may be influenced by policy. Consider, for instance, a ‘‘trusted vendor certification” program that provides receivers with assurances that the certified vendors could be trusted. A credible certification program that includes customer reviews, akin to those used in Internet retail sites, could help overcome receivers’ concerns and expedite adoption of U-OHD. Such a program is attractive because it could be administered by private-sector groups that could offer the certification as a benefit to their members, thus relieving the public sector from that responsibility. After a comprehensive modeling effort, the final model, in which all variables are statistically significant and conceptually valid, is shown in Table 1. The authors performed a test of parallel lines (Brant, 1990), that indicated that the current model form, which assumes all parameters remain consistent across all willingness levels, is too restrictive. Future modeling efforts should look into alternative model forms that allow parameters to vary across different willingness levels. The results shows that willingness to accept U-OHD is a function of the number of deliveries received, policy variables, and industry sectors. The results show that a one-time incentive, business support, public recognition, carrier discount, and availability of a trusted vendor help foster adoption of U-OHD, and that the number of deliveries has a negative relation to willingness to accept U-OHD. This result lines up with expectations, as it is more difficult to switch large number of deliveries to the off-hours than to switch only a few. The model shows that firms have different attitudes towards U-OHD depending on the industry segments they belong to. Performing arts businesses seem to be particularly willing to accept U-OHD. The model also reveals that nondurable wholesalers, apparel manufacturing, and retail establishments (e.g., clothing, food and beverages stores) are very willing to participate in U-OHD when incentives are provided. Clothing stores have interesting reactions: they are generally unwilling to accept U-OHD, but their attitudes change dramatically if they have a trusted vendor. These results are consistent with the previous research conducted for staffed OHD for both NYC (Holguín-Veras et al., 2007) and Spanish cities (Domínguez et al., 2012). It should be said that the research conducted by Holguín-Veras et al. (2007) and Domínguez et al. (2012) are not fully comparable because they did not consider the role of the one-time incentive (they considered an on-going incentive), business support, public recognition, carrier discount, and trusted vendor. Table 2 shows the market elasticities for the key policy variables. The results reveal that the most potent variables are, in descending order of importance: the availability of a trusted vendor, the one-time incentive, carrier discount, business support, and public recognition. In terms of response to incentives, the elasticities for ‘‘somewhat willing” and ‘‘very willing” are positive, meaning that an increase in the incentive would increase the willingness to accept U-OHD. It should be mentioned that there are instances where there are both a generic (that apply to all industry sectors) and industry-specific parameter that impact the same variable. In these cases, the total effect is the summation of both effects, generic plus industry-specific. Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

10

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx Table 1 Ordered logit model with random effects. Model

Model

Independent variables

Coefficient

t-stat

Constant Number of deliveries (per week) Trusted Vendor (TV) Incentives Business Support (BS) Public Recognition (PR) One time incentive in $1000 (OTI) Carrier discount in percent (CD) NAICS Performing arts, binary variable Clothing stores, binary variable Interaction terms: One Time Incentive (OTI) and NAICS OTI for nondurable wholesalers OTI for clothing stores OTI for apparel manufacture stores OTI for food and beverage stores Interaction terms: Carrier Discount (CD) and NAICS CD for personal laundry Interaction terms: Trusted Vendor (TV) and NAICS TV for clothing stores TV for miscellaneous stores retailers TV for food and beverage stores TV for performing arts Parameters l(1) l(2) l(3) Sigma n Log likelihood

1.727 0.040 2.270

(4.66) (2.93) (5.17)

0.522 0.381 0.197 0.029

(3.31) (2.46) (7.29) (6.18)

2.057 2.045

(2.41) (2.28)

0.272 0.253 0.216 0.150

(3.94) (3.34) (1.59) (1.81)

0.010

(1.07)

4.756 4.252 4.118 2.706

(3.00) (3.99) (2.87) (1.55)

1.980 4.670 7.801 5.841 1522 1378.66

(21.37) (33.85) (40.97) (17.10)

4.2. Subjective monetary value of non-monetary incentives As shown in Table 1, the model includes both monetary and non-monetary incentives. This provides an opportunity to estimate the subjective monetary value of the non-monetary incentives. The monetary value of each incentive is the marginal rate of substitution of the incentive with respect to the one-time-incentive:

MV I;j ¼

   @U @U @I @OTI

ð9Þ

where MVI,j: monetary value of incentive I for segment j I: incentive J: industry segment

U: utility function specified in the model OTI: one time incentive offered (US$)

Table 2 Summary of market elasticity values. Variable

Not at all willing

Not too willing

Neutral

Somewhat willing

Very willing

Trusted vendor (TV) TV for miscellaneous retailers TV for clothing stores TV for performing arts TV for food and beverage stores One-time incentive (OTI) OTI for nondurable goods wholesalers OTI for clothing stores OTI for food and beverage stores OTI for apparel manufacture stores Business support Public recognition Carrier discount (CD) CD for personal and laundry

0.2451 0.4225 0.4812 0.2855 0.4355 0.1048 0.0121 0.0127 0.0048 0.0033 0.0568 0.0415 0.0767 0.0053

0.0094 0.1048 0.1359 0.0449 0.1097 0.0024 0.0003 0.0003 0.0001 0.0001 0.0012 0.0010 0.0017 0.0001

0.0645 0.0291 0.0155 0.0454 0.0281 0.0330 0.0038 0.0040 0.0015 0.0011 0.0178 0.0131 0.0241 0.0017

0.1390 0.1896 0.2023 0.1457 0.1939 0.0622 0.0072 0.0075 0.0028 0.0020 0.0337 0.0247 0.0455 0.0031

0.2331 0.4728 0.5542 0.2970 0.4891 0.0950 0.0109 0.0115 0.0043 0.0030 0.0515 0.0376 0.0695 0.0048

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

11

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

The resulting estimates, shown in Table 3, indicate that the various policy incentives have a sizable monetary value, though the valuation varies with industry sector. The results reveal that offering business support services has the same effect on utility as a $987 to $2643 one-time incentive, depending on the industry segment. Providing public recognition is equivalent to a one-time-incentive ranging from $851 to $1935. This range is lower than the one for business support, though it is an attractive alternative for the public sector because there is no need to incur large expenses to make it possible. Having a trusted vendor was valued the highest by the responders, with a subjective monetary value ranging from $4863 to $33,139. These results highlight the important role that a trusted vendor certification program could play to foster U-OHD. Tables 4 and 5 show the U-OHD market shares for different combinations of incentives for all industry segments, and retail. The top of the table shows the percentage of firms that would be willing to accept U-OHD when no business support is offered, for each one-time incentive, for each carrier discount, and when public recognition is provided (or not). To study the potential impacts of a trusted vendor program two scenarios were developed: on the left of the table, the market shares are shown for the original sample (30.31% of establishments with a trusted vendor, and 28.16% for retail establishments), market shares on the right correspond to a simulated scenario where 50% of the firms have a trusted vendor. Since it is not possible to define which firms are likely to participate in the trusted vendor program, the latter scenario is created using a Monte Carlo simulation to randomly select the firms. The market shares presented in the tables are the average of ten simulations. The corresponding standard deviations are shown in Appendix A. As shown in Table 4, the market share of U-OHD for all industry segments ranges from 3.0% to 24.0% for the current proportion of trusted vendors (30.31%); and from 5.1% to 33.8% if the proportion of trusted vendors increases to 50%. However, the larger impact in market shares is produced by the one-time incentive. Table 5 confirms previous findings indicating that retail establishments are inclined to accept U-OHD. In retail, the U-OHD market share ranges from 5.2% to 25.4% with the current proportion of trusted vendors (28.16%); and from 8.6% to 39.9% if that proportion increases to 50%. Targeting the retail sector, and implementing strategies to raise the proportion of trusted vendors would increase U-OHD. As expected from the model structure, the response to policy variables is non-linear. Since different combinations of incentives could achieve the same level of market share, it is important to determine which combination is the most appropriate for the specific conditions in the city. Moreover, since some of the policies require the involvement of other stakeholders, it is crucial to establish multi-stakeholder collaborative approaches to achieve the desired behavior change most efficiently. 5. Implications and concluding remarks The research reported in this paper assessed the effectiveness of various public policies to foster U-OHD in NYC. To this effect, stated preference data were used to estimate random effect Ordered Logit models. The final model: identifies how firms’ attributes affect the willingness to participate in U-OHD; determines which industry sectors are most inclined to accept U-OHD; assesses the effectiveness of alternative policies; and estimates potential participation for the various policy scenarios. The policy variables considered were: one-time incentive, carrier discounts, public recognition, business support, and the availability of a trusted vendor. The results indicate the potential of these policies. The data show that businesses with a trusted vendor are about 1.4 times more inclined to participate in U-OHD (39.17% vs. 27.30%). A sizable number of establishments could do U-OHD with minimal inconvenience. There is also the exciting possibility of developing, in collaboration with business groups, a ‘‘Trusted Vendor Certification Program” that would certify vendors that have completed training on how to conduct U-OHD safely, and in compliance with all appropriate regulations and best practices. This, together with the finding that 48.58% of establishments do not need the consent of anyone else to implement U-OHD, suggest that a trusted vendor program could play a key role in fostering U-OHD. There is a real possibility of inducing large amounts of receivers to participate in U-OHD, particularly the 15.44% that have considered accepting OHD, most notably, without any incentive. The analyses of the willingness to participate in connection to business characteristics produced important findings. The first is that willingness to participate in U-OHD increases with business size up to a point (16–20 employees), and then decreases. This makes sense, as if the establishment is too small, there may not be a need for U-OHD; while if the establishment is too large, the coordination effort and risk may be a challenge. However, this does not diminish the potential of U-OHD, because small establishments produce the bulk of the freight traffic in urban areas. A second important finding is that industry sectors exhibit different levels of willingness to accept U-OHD. In descending order of willingness: other services (NAICS 81), accommodations and food services (NAICS 72); retail trade (NAICS 44-45), manufacturing (NAICS 31-33),

Table 3 Subjective monetary values of incentives and key characteristics for each industry segment. NAICS Description

315 Apparel manufacturing

424 Non durable wholesaler

445 Food and beverages

448 Clothing stores

453 Miscellaneous store retail

711 Performing arts

812 Personal & laundry services

Other Segments

Range

Public Recognition Business Support Trusted Vendor

$912 $1248 $5447

$814 $987 $4863

$1096 $1500 $6547

$851 $1164 $15,736

$1931 $2643 $33,139

$1931 $2643 $25,311

$1931 $2643 $11,537

$1931 $2643 $11,537

$851–$1931 $987–$2643 $4863 –$33,139

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

12

Actual Trusted Vendors: 30.3% Carrier Discount

0%

One-Time Incentive

No PR

Simulated Trusted Vendors: 50% 10%

20%

30%

40%

50%

0%

10%

20%

30%

40%

50%

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No Business Support $0 3.0 $1000 3.5 $2000 4.1 $3000 4.7 $4000 5.4 $5000 6.1 $6000 6.9 $7000 7.7 $8000 8.6 $9000 9.5

3.7 4.3 4.9 5.6 6.4 7.1 7.9 8.8 9.8 10.9

3.5 4.1 4.7 5.4 6.1 6.8 7.6 8.5 9.4 10.5

4.3 4.9 5.6 6.3 7.1 7.9 8.8 9.7 10.8 12.1

4.1 4.7 5.4 6.1 6.8 7.6 8.4 9.4 10.4 11.6

4.9 5.6 6.3 7.1 7.9 8.7 9.7 10.8 12.0 13.5

4.7 5.3 6.0 6.8 7.6 8.4 9.3 10.4 11.5 12.9

5.6 6.3 7.0 7.9 8.7 9.7 10.7 12.0 13.4 15.1

5.3 6.0 6.8 7.5 8.4 9.3 10.3 11.5 12.8 14.4

6.3 7.0 7.8 8.7 9.7 10.7 11.9 13.3 15.0 16.9

6.0 6.7 7.5 8.4 9.3 10.3 11.4 12.8 14.3 16.1

7.0 7.8 8.7 9.7 10.7 11.9 13.3 14.9 16.8 19.0

5.1 6.0 7.1 8.3 9.5 10.8 12.2 13.5 15.0 16.4

6.3 7.4 8.6 9.8 11.1 12.5 13.9 15.3 16.8 18.5

6.0 7.0 8.2 9.4 10.7 12.0 13.4 14.8 16.3 17.9

7.3 8.5 9.7 11.0 12.4 13.8 15.2 16.7 18.3 20.0

7.0 8.1 9.3 10.6 11.9 13.3 14.7 16.2 17.7 19.4

8.4 9.6 10.9 12.3 13.6 15.1 16.6 18.2 19.9 21.8

8.0 9.2 10.5 11.8 13.2 14.6 16.0 17.6 19.3 21.1

9.6 10.8 12.2 13.5 15.0 16.5 18.1 19.9 21.7 23.7

9.1 10.4 11.7 13.1 14.5 15.9 17.5 19.2 21.0 22.9

10.7 12.1 13.4 14.9 16.4 18.0 19.8 21.7 23.7 25.9

10.3 11.6 13.0 14.4 15.9 17.4 19.1 20.9 22.9 25.0

12.0 13.4 14.8 16.3 18.0 19.7 21.6 23.7 25.9 28.3

Business Support $0 $1000 $2000 $3000 $4000 $5000 $6000 $7000 $8000 $9000

4.8 5.5 6.2 7.0 7.8 8.7 9.6 10.7 11.9 13.4

4.6 5.3 6.0 6.7 7.5 8.3 9.2 10.3 11.4 12.8

5.5 6.2 7.0 7.8 8.6 9.6 10.7 11.9 13.3 14.9

5.3 5.9 6.7 7.5 8.3 9.2 10.2 11.4 12.7 14.3

6.2 7.0 7.8 8.6 9.6 10.6 11.8 13.2 14.9 16.8

5.9 6.7 7.4 8.3 9.2 10.2 11.3 12.7 14.2 16.0

6.9 7.8 8.6 9.6 10.6 11.8 13.2 14.8 16.7 18.9

6.7 7.4 8.3 9.2 10.2 11.3 12.6 14.1 15.9 18.0

7.8 8.6 9.6 10.7 11.9 13.2 14.8 16.7 18.8 21.3

7.4 8.3 9.2 10.2 11.3 12.6 14.1 15.9 17.9 20.3

8.7 9.6 10.7 11.9 13.3 14.8 16.7 18.8 21.2 24.0

6.8 7.9 9.1 10.4 11.8 13.1 14.5 16.0 17.6 19.3

8.3 9.5 10.8 12.1 13.5 14.9 16.5 18.1 19.8 21.7

7.9 9.1 10.3 11.6 13.0 14.4 15.9 17.5 19.1 20.9

9.4 10.7 12.0 13.4 14.8 16.4 18.0 19.7 21.6 23.6

9.0 10.2 11.5 12.9 14.3 15.8 17.4 19.0 20.8 22.8

10.6 11.9 13.3 14.7 16.3 17.9 19.6 21.5 23.6 25.8

10.1 11.4 12.8 14.2 15.7 17.3 19.0 20.8 22.7 24.8

11.8 13.2 14.7 16.2 17.8 19.6 21.5 23.6 25.8 28.2

11.3 12.7 14.1 15.6 17.2 18.9 20.7 22.7 24.8 27.1

13.1 14.6 16.1 17.8 19.6 21.5 23.6 25.8 28.2 30.9

12.6 14.1 15.6 17.1 18.8 20.7 22.7 24.9 27.2 29.7

14.5 16.1 17.8 19.5 21.5 23.6 25.9 28.3 31.0 33.8

4.0 4.6 5.3 6.0 6.7 7.5 8.4 9.3 10.3 11.5

Notes: (1) PR: Public Recognition; (2) Values indicate market shares in%.

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

Table 4 Market shares summary for all industry segments.

Actual Trusted Vendors: 28.2% Carrier Discount

0%

One-Time Incentive

No PR

Simulated Trusted Vendors: 50% 10%

20%

30%

40%

50%

0%

10%

20%

30%

40%

50%

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No Business Support $0 5.2 $1000 6.1 $2000 7.2 $3000 8.4 $4000 9.6 $5000 10.8 $6000 11.9 $7000 13.0 $8000 13.9 $9000 14.8

6.4 7.5 8.6 9.8 11.0 12.2 13.2 14.2 15.1 15.9

6.1 7.2 8.3 9.5 10.7 11.8 12.9 13.9 14.8 15.6

7.4 8.5 9.7 10.9 12.1 13.1 14.1 15.0 15.9 16.8

7.1 8.2 9.4 10.6 11.7 12.8 13.8 14.8 15.6 16.5

8.5 9.6 10.8 11.9 13.0 14.1 15.0 15.9 16.8 17.8

8.1 9.3 10.4 11.6 12.7 13.8 14.7 15.6 16.5 17.4

9.5 10.7 11.8 12.9 14.0 15.0 15.9 16.9 17.8 18.9

9.2 10.3 11.5 12.6 13.7 14.7 15.6 16.6 17.5 18.5

10.5 11.7 12.8 13.9 15.0 16.0 16.9 17.9 19.0 20.2

10.2 11.4 12.5 13.6 14.7 15.6 16.6 17.6 18.6 19.8

11.6 12.7 13.9 14.9 16.0 17.0 18.0 19.1 20.3 21.8

8.6 10.4 12.4 14.6 17.0 19.2 21.4 23.2 24.9 26.3

10.8 12.8 15.0 17.3 19.6 21.7 23.6 25.3 26.7 28.0

10.2 12.2 14.4 16.6 18.9 21.1 23.1 24.8 26.2 27.5

12.6 14.7 17.0 19.3 21.4 23.4 25.2 26.7 28.0 29.2

12.0 14.1 16.3 18.6 20.8 22.9 24.7 26.2 27.6 28.8

14.4 16.7 18.9 21.2 23.2 25.0 26.6 28.1 29.4 30.6

13.8 16.0 18.3 20.5 22.6 24.5 26.2 27.6 28.9 30.1

16.4 18.6 20.9 23.0 24.9 26.6 28.1 29.5 30.8 32.1

15.7 18.0 20.2 22.4 24.4 26.1 27.6 29.0 30.3 31.6

18.3 20.5 22.7 24.7 26.5 28.1 29.6 30.9 32.3 33.8

17.7 19.9 22.1 24.2 26.0 27.6 29.1 30.4 31.8 33.2

20.2 22.5 24.5 26.4 28.1 29.7 31.1 32.6 34.1 35.7

Business Support $0 $1000 $2000 $3000 $4000 $5000 $6000 $7000 $8000 $9000

8.3 9.4 10.6 11.7 12.9 13.9 14.9 15.8 16.6 17.6

7.9 9.1 10.2 11.4 12.5 13.6 14.6 15.5 16.3 17.2

9.3 10.5 11.6 12.8 13.8 14.8 15.8 16.7 17.6 18.7

9.0 10.1 11.3 12.4 13.5 14.5 15.5 16.4 17.3 18.3

10.3 11.5 12.7 13.8 14.8 15.8 16.7 17.7 18.7 19.9

10.0 11.2 12.3 13.4 14.5 15.5 16.4 17.4 18.4 19.5

11.4 12.6 13.7 14.8 15.8 16.8 17.8 18.8 20.1 21.5

11.0 12.2 13.3 14.4 15.5 16.4 17.4 18.5 19.6 20.9

12.4 13.6 14.7 15.8 16.8 17.9 19.0 20.2 21.6 23.3

12.1 13.3 14.4 15.4 16.5 17.5 18.6 19.7 21.1 22.6

13.5 14.7 15.8 16.9 18.0 19.1 20.3 21.8 23.4 25.4

11.7 13.7 16.0 18.3 20.5 22.5 24.4 26.0 27.3 28.6

14.1 16.3 18.6 20.8 22.9 24.7 26.4 27.8 29.1 30.3

13.5 15.7 17.9 20.2 22.3 24.2 25.9 27.3 28.7 29.9

16.0 18.2 20.5 22.6 24.6 26.3 27.8 29.2 30.5 31.8

15.4 17.6 19.9 22.0 24.0 25.8 27.3 28.7 30.0 31.3

17.9 20.2 22.4 24.4 26.2 27.8 29.3 30.7 32.0 33.5

17.3 19.6 21.8 23.8 25.7 27.3 28.8 30.2 31.5 32.9

19.9 22.1 24.2 26.1 27.8 29.4 30.8 32.2 33.7 35.3

19.2 21.5 23.6 25.6 27.3 28.9 30.3 31.7 33.1 34.7

21.8 24.0 26.0 27.8 29.5 31.0 32.5 34.0 35.7 37.5

21.2 23.4 25.4 27.3 28.9 30.5 31.9 33.4 35.0 36.7

23.8 25.9 27.8 29.5 31.2 32.7 34.3 36.0 37.8 39.9

6.9 8.0 9.2 10.4 11.5 12.6 13.7 14.6 15.5 16.3

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

Notes: (1) PR: Public Recognition; (2) Values indicate market shares in%.

13

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

Table 5 Market shares summary for retail establishments.

14

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

wholesale trade (NAICS 42); and arts, entertainment, and recreation (NAICS 71). Small businesses in these segments are ideal targets for U-OHD programs. The data also reveal that a sizable number of responses (10% to 40%, depending on the industry sector) are ‘‘neutral” to the idea of accepting U-OHD. This is very encouraging, because it suggests that a well-managed UOHD program that builds a good track record could win over a large group of participants. Since the ‘‘neutral” and the ‘‘willing” are about 50% of the total number of establishments, the potential market for U-OHD is indeed large. The behavioral models show that willingness to accept U-OHD could be influenced by a one-time incentive, carrier discount, business support, public recognition, and the availability of trusted vendors. In terms of policies, the most effective are the one-time-incentive and the carrier discount; followed by public recognition and business support. However, in terms of influence, the availability of a trusted vendor is the most important variable. The one-time-incentive is key, as without it, receivers are less likely to take the risk and try U-OHD. However, the data show that the one-time-incentive has a limit, as it increases willingness to participate only up to $4000, which increases to 36.46% from 20.11% (no incentive). Beyond that point, further increases in willingness to participate in U-OHD do not materialize (for the range of values studied in the research). The shipping discounts provided by the carriers play an important role as they increase willingness by 20.62% (from 20.43% to 41.05% if a 50% discount is provided). In contrast with the one-time incentive, the discount consistently increases for the range of values considered. Obviously, securing an ongoing discount that could be in place for a long time is bound to be more enticing to businesses than a one-time incentive. The availability of a trusted vendor has a significant impact on willingness to accept U-OHD, increasing it by 11.87%. These results highlight the potential of creating Trusted Vendor Certification Programs that would train carriers on how to do U-OHD safely and in compliance with community-sensitive practices. Independent groups could manage such programs, which could help ameliorate receiver concerns. Business support services and public recognition do provide modest, but meaningful, increases in willingness, in the range of 4.53% and 3.22%, respectively. In addition, these programs have other benefits, as they could help transform the nature of the relations between public and private sectors, and create a more collaborative and business friendly environment. The models suggest that market participation in U-OHD would range between 3.0% and 24.0% with the current proportion of trusted vendors (30.31%), and between 5.1% and 33.8% if that proportion increases to 50%. For the specific case of retail establishments, the estimated U-OHD market share ranges from 5.2% to 25.4% with the current proportion of trusted vendors (28%), and from 8.6% to 39.9% if the proportion increases to 50%. In essence, targeting retail establishments and implementing strategies to raise the proportion of trusted vendors are key strategies to reach significant shifts to U-OHD. It should be noted, however, that the model used in this paper may be too restrictive, as indicated by the parallel line test. Future work should look into alternative model forms to provide more informative results. The picture that emerges is that, although there is no single policy that could accomplish a large shift to U-OHD, the public sector has a number of policy levers to foster U-OHD as part of a multi-layered, multi-stakeholder collaborative approach. These would include: (1) public sector provision of a one-time incentive, a public recognition program to recognize participants, and business support services to help participating business run their operations better; (2) carriers providing shipping discounts to receivers of U-OHD; and (3) the creation of a Trusted Vendor Certification Program. Exploiting these mechanisms—together with regulations to ensure that local communities are not negatively impacted—could go a long way to foster U-OHD. These policies, in helping to advance U-OHD, have the potential to increase economic competitiveness, reduce congestion, improve environmental conditions, enhance livability, and increase quality of life in urban areas. Of course, notwithstanding its potential benefits, OHD is not a panacea. The challenges of business inertia, noise, and pollution impacts at night must be overcome to implement a successful OHD program. Inertia is the result of the pressures of running day-to-day business operations in competitive environments, which makes it difficult for businesses to change behavior, as required by OHD. Overcoming this requires a significant effort of stakeholder engagement. Noise at night presents other challenges; particularly in low-rise cities with cobblestones. In these environments, the low noise delivery practices that are effective in cities like NYC do not help much (Holguín-Veras et al., 2013a). There are also concerns that OHD could lead to local increases in air pollution, because the cooler night air reduces the dispersion of pollutant produced by solar heating in the daytime hours (Sathaye et al., 2010). However, there should be no doubt that OHD reduces total environmental pollution. The estimates of pollution produced by the same truck in the regular and off-hours indicate that the amount of CO2 emitted in off-hours—for the same segments of the network—was between 20% and 75% lower than in the regular hours (Holguín-Veras et al., 2013a, 128). OHD is best suited for highly congested urban areas where these local impacts could be satisfactorily addressed. A solid analysis of local conditions is key to determining whether OHD is worthy of consideration for a given urban area.

Acknowledgments The research reported in this paper was supported by the Commercial Remote Sensing and Spatial Information Technologies’ project ‘‘Integrative Freight Demand Management in the New York City Metropolitan Area: Implementation Phase,” which is part of the United States Department of Transportation’s Office of the Assistant Secretary for Research and Technology (USDOT’s OASRT); and by the New York City Department of Transportation (NYCDOT). The authors would like to also acknowledge the support and guidance from Mr. Caesar Singh and Mr. Vasanth Ganesan (USDOT’s OASRT). Their support is both acknowledged and appreciated. Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

Standard Deviation All Industry Segments, No Business Support, Trusted Vendors: 50% 10%

20%

30%

Retail Segment, No Business Support, Trusted Vendors: 50%

Carrier Discount One-Time Incentive

0%

40%

50%

0%

10%

20%

30%

40%

50%

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

$0 $1000 $2000 $3000 $4000 $5000 $6000 $7000 $8000 $9000

0.48 0.53 0.58 0.64 0.69 0.74 0.78 0.81 0.83 0.84

0.56 0.61 0.66 0.71 0.75 0.79 0.82 0.83 0.84 0.85

0.54 0.59 0.64 0.69 0.74 0.78 0.81 0.83 0.84 0.85

0.62 0.67 0.72 0.76 0.79 0.82 0.83 0.84 0.84 0.84

0.60 0.65 0.70 0.74 0.78 0.81 0.83 0.84 0.84 0.84

0.68 0.72 0.76 0.80 0.82 0.83 0.84 0.83 0.82 0.81

0.66 0.71 0.75 0.79 0.81 0.83 0.84 0.84 0.83 0.82

0.73 0.76 0.80 0.82 0.83 0.83 0.83 0.81 0.80 0.77

0.71 0.75 0.79 0.82 0.83 0.84 0.83 0.82 0.81 0.79

0.77 0.80 0.82 0.83 0.83 0.82 0.80 0.78 0.75 0.72

0.76 0.79 0.82 0.83 0.83 0.83 0.81 0.80 0.77 0.74

0.80 0.82 0.83 0.83 0.81 0.79 0.76 0.73 0.70 0.66

0.34 0.36 0.36 0.34 0.30 0.26 0.23 0.23 0.26 0.29

0.40 0.39 0.37 0.32 0.27 0.24 0.24 0.26 0.30 0.34

0.38 0.39 0.37 0.33 0.28 0.24 0.23 0.25 0.29 0.33

0.42 0.40 0.36 0.30 0.26 0.25 0.26 0.30 0.33 0.37

0.42 0.40 0.36 0.31 0.26 0.24 0.25 0.29 0.32 0.36

0.44 0.40 0.34 0.29 0.26 0.26 0.29 0.33 0.36 0.40

0.44 0.40 0.35 0.29 0.26 0.26 0.28 0.32 0.35 0.39

0.44 0.38 0.32 0.28 0.27 0.29 0.32 0.35 0.39 0.42

0.44 0.39 0.33 0.28 0.27 0.28 0.31 0.35 0.38 0.41

0.43 0.36 0.31 0.28 0.29 0.31 0.34 0.38 0.41 0.45

0.43 0.37 0.31 0.28 0.28 0.31 0.34 0.37 0.40 0.44

0.40 0.34 0.30 0.29 0.31 0.33 0.36 0.40 0.43 0.47

Min

0.48

Max

0.47

Max 0.85 0.23

All Industry Segments, Business Support, Trusted Vendors: 50%

Retail Segment, Business Support, Trusted Vendors: 50%

Carrier Discount One-Time Incentive

0% No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

No PR

PR

$0 $1000 $2000 $3000 $4000 $5000 $6000 $7000 $8000 $9000

0.59 0.64 0.69 0.73 0.77 0.81 0.83 0.84 0.84 0.84

0.67 0.71 0.75 0.79 0.81 0.83 0.83 0.83 0.82 0.81

0.65 0.70 0.74 0.78 0.81 0.83 0.84 0.84 0.83 0.82

0.72 0.76 0.79 0.82 0.83 0.83 0.83 0.81 0.80 0.77

0.70 0.74 0.78 0.81 0.83 0.83 0.83 0.82 0.81 0.79

0.76 0.79 0.82 0.83 0.83 0.82 0.80 0.78 0.76 0.72

0.75 0.78 0.81 0.83 0.83 0.83 0.82 0.80 0.77 0.75

0.79 0.81 0.83 0.83 0.81 0.79 0.77 0.74 0.70 0.66

0.78 0.81 0.83 0.83 0.82 0.81 0.78 0.76 0.72 0.69

0.81 0.82 0.82 0.81 0.78 0.75 0.72 0.68 0.64 0.60

0.81 0.82 0.83 0.82 0.80 0.77 0.74 0.70 0.66 0.62

0.82 0.82 0.80 0.77 0.74 0.70 0.66 0.61 0.57 0.54

0.41 0.40 0.36 0.31 0.27 0.24 0.25 0.28 0.32 0.36

0.44 0.40 0.34 0.29 0.26 0.26 0.29 0.32 0.36 0.39

0.43 0.40 0.35 0.30 0.26 0.25 0.28 0.31 0.35 0.38

0.44 0.39 0.33 0.28 0.27 0.28 0.32 0.35 0.38 0.42

0.44 0.39 0.33 0.28 0.26 0.28 0.31 0.34 0.38 0.41

0.43 0.36 0.31 0.28 0.28 0.31 0.34 0.37 0.41 0.44

0.43 0.37 0.31 0.28 0.28 0.30 0.33 0.37 0.40 0.43

0.41 0.34 0.30 0.29 0.30 0.33 0.36 0.39 0.43 0.47

0.41 0.35 0.30 0.28 0.30 0.32 0.35 0.39 0.42 0.46

0.38 0.32 0.29 0.30 0.32 0.35 0.38 0.41 0.45 0.50

0.39 0.33 0.29 0.29 0.31 0.34 0.37 0.41 0.44 0.49

0.35 0.30 0.30 0.31 0.34 0.36 0.40 0.43 0.48 0.54

Min

0.54

Max

0.54

10%

20%

30%

40%

50%

0%

Max 0.84 0.24

10%

20%

30%

40%

50%

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

Notes: (1) PR: Public Recognition; (2) Values indicate standard deviation of market shares in%

15

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005

Appendix A. Standard deviations for the Monte Carlo simulations of trusted vendors

16

J. Holguín-Veras et al. / Transportation Research Part A xxx (2017) xxx–xxx

References Beijing Traffic Management Bureau, 2010. Delivery Time Restrictions. Retrieved March 11th, 2011, from . Beijing Traffic Management Bureau, 2014. Delivery Time Restrictions. Retrieved July 22th, 2014, from . Ben-Akiva, M., Lerman, S., 2010. Discrete Choice Analysis: Theory and Application to Travel Demand. The MIT Press, Cambridge, Massachusetts. Brant, R., 1990. Assessing proportionality in the proportional odds model for ordinal logistic regression. Biometrics 46 (4), 1171–1178. http://dx.doi.org/ 10.2307/2532457. Brom, M.A., Holguín-Veras, J., Hodge, S., 2011. Off-hour deliveries in Manhattan, New York City: experiences of Pilot Test Participants. Transp. Res. Rec. 2238, 77–85. http://dx.doi.org/10.3141/2238-10. Brown, J.J., Reingen, P.H., 1987. Social ties and word-of-mouth referral behavior. J. Consum. Res. 14 (3), 350–362. Browne, M., Allen, J., Anderson, S., Woodburn, A., 2006. Night-time delivery restrictions: a review. In: Taniguchi, E., Thompson, R. (Eds.), Recent Advances in City Logistics. Elsevier, Amsterdam, pp. 269–281. Cassidy, W.B., 2009. New York delivers at night. J. Commer. 10 (35), 4A. Cassidy, W.B., 2010. New York to expand off-peak truck program. J. Commer. Centro de Inovação em Sistemas Logísticos, 2015. Avaliação do Projeto-Piloto de Entregas Noturnas no Município de São Paulo. Serie Cadernos Tecnicos, 18, from . Changsha Bureau of Public Security, 2013. Announcement about Freight Vehicles Ban in Certain Areas and Routes. Retrieved July 22th, 2014, from . City of New York, 2011. PlanNYC: A Greener, Greater New York: Update 2011. Retrieved 5/1/11, from . City of New York, 2015. One New York: The Plan for a Strong and Just City. From . Domencich, T.A., McFadden, D., 1975. Urban Travel Demand: A Behavioural Analysis. Noth-Holland, Amsterdam. Domínguez, A., Holguín-Veras, J., Ibeas, Á., Dell’Olio, L., 2012. Receivers’ response to new urban freight policies. Procedia-Soc. Behav. Sci. 54, 886–896. Federal Highway Administration, 2012. Federal Grant Opportunity Request for Applications (RFA): Off Hours Freight Delivery Pilot Project. Retrieved July 1, 2012, from . Harris, F., 1915. Operations and Costs. A.W. Shaw, Chicago, IL. Hess, S., Rose, J.M., 2009. Allowing for intra-respondent variations in coefficients estimated on repeated choice data. Transp. Res. Part B: Methodol. 43 (6), 708–719. http://dx.doi.org/10.1016/j.trb.2009.01.007. Holguín-Veras, J., 2008. Necessary conditions for off-hour deliveries and the effectiveness of urban freight road pricing and alternative financial policies in competitive markets. Transp. Res. Part A: Policy Pract. 42 (2), 392–413. Holguín-Veras, J., 2011. Urban delivery industry response to cordon pricing, time-distance pricing, and carrier-receiver policies in competitive markets. Transp. Res. Part A 45, 802–824. Holguín-Veras, J., Aros-Vera, F., 2014. Self-supported freight demand management: pricing and incentives. EURO J. Transp. Logist. 3 (1), 1–24. http://dx.doi. org/10.1007/s13676-013-0041-1. Holguín-Veras, J., Aros-Vera, F., Browne, M., 2015. Agent interactions and the response of supply chains to pricing and incentives. Econ. Transp. (in press). http://dx.doi.org/10.1016/j.ecotra.2015.04.002. Holguín-Veras, J., Marquis, R., Brom, M., 2012. Economic impacts of staffed and unassisted off-hour deliveries in New York city. Procedia Soc. Behav. Sci. Holguín-Veras, J., Marquis, R., Campbell, S., Wojtowicz, J.M., Wang, X.C., Jaller, M.A., Hodge, S.D., Rothbard, S., Goevaers, R., 2013a. Fostering the use of unassisted off-hour deliveries: operational and low-noise truck technologies. Transp. Res. Rec. 2379, 57–63. Holguín-Veras, J., Ozbay, K., Kornhauser, A.L., Brom, M., Iyer, S., Yushimito, W., Ukkusuri, S., Allen, B., Silas, M., 2011. Overall impacts of off-hour delivery programs in the New York city metropolitan area. Transp. Res. Rec. 2238, 68–76. Holguín-Veras, J., Sánchez-Díaz, I., 2016. Freight demand management and the potential of receiver-led consolidation programs. Transp. Res. Part A: Policy Pract. 84, 109–130. Holguín-Veras, J., Silas, M.A., Polimeni, J., Cruz, B., 2007. An investigation on the effectiveness of joint receiver-carrier policies to increase truck traffic in the off-peak hours: Part I: the behaviors of receivers. Netw. Spatial Econ. 7 (3), 277–295. http://dx.doi.org/10.1007/s11067-006-9002-7. Holguín-Veras, J., Silas, M.A., Polimeni, J., Cruz, B., 2008. An investigation on the effectiveness of joint receiver-carrier policies to increase truck traffic in the off-peak hours: Part II: the behaviors of carriers. Netw. Spatial Econ. 8 (4), 327–354. http://dx.doi.org/10.1007/s11067-006-9011-6. Holguín-Veras, J., Wang, Q., Xu, N., Ozbay, K., Cetin, M., Polimeni, J., 2006. The impacts of time of day pricing on the behavior of freight carriers in a congested urban area: implications to road pricing. Transp. Res. Part A: Policy Pract. 40 (9), 744–766. Holguín-Veras, J., Wojtowicz, J.M., Wang, C., Jaller, M., Ban, X.J., Aros-Vera, F., Campbell, S., Yang, X., Sanchez-Diaz, I., Amaya, J., González-Calderón, C., Marquis, R., Hodge, S.D., Maguire, T., Marsico, M., Zhang, S., Rothbard, S., Ozbay, K., Morgul, E.F., Iyer, S., Xie, K., Ozguven, E.E., 2013b. Integrative Freight Demand Management in the New York City Metropolitan Area: Implementation Phase.‘‘ from . Marcucci, E., Danielis, R., 2008. The potential demand for a urban freight consolidation centre. Transportation 35 (2), 269–284. http://dx.doi.org/10.1007/ s11116-007-9147-3. Marcucci, E., Gatta, V., Scaccia, L., 2015. Urban freight, parking and pricing policies: an evaluation from a transport provider’s perspective. Transp. Res. Part A: Policy Pract. 74, 239–249. http://dx.doi.org/10.1016/j.tra.2015.02.011. Noise Abatement Society, 2013. John Connel Awards. Retrieved January 24th, 2013, 2013, from . Nuzzolo, A., Comi, A., 2014. Urban freight demand forecasting: a mixed quantity/delivery/vehicle-based model. Transp. Res. Part E: Logist. Transp. Rev. 65, 84–98. Pigou, A.C., 1920. The Economics of Welfare. MacMillan and Co., Ltd., London. Sathaye, N., Harley, R., Madanat, S., 2010. Unintended environmental impacts of nighttime freight logistics activities. Transp. Res. Part A: Policy Pract. 44 (8), 642–659. Shenzhen Bureau of Public Security, 2013. Announcement about Truck Routes and Banned Areas. Retrieved July 22, 2014, from . Silas, M.A., Holguín-Veras, J., 2009. Behavioral microsimulation formulation for analysis and design of off-hour delivery policies in urban areas. Transp. Res. Rec.: J. Transp. Res. Board 2097, 43–50. http://dx.doi.org/10.3141/2097-06. Transport for London, 2013. Fleet Operator Recognition Scheme. Retrieved January 24th, 2013, 2013, from . U.S. Environmental Protection Agency, 2013. SmartWay Transport Partnership. Retrieved January 24th, 2013, from . Varian, H., 1992. Microeconomic Analysis. Norton, New York, NY. Wall Street Journal, 2010. Congestion Fight Runs Into Night. Retrieved July 2, 2010, from . Yáñez, M.F., Cherchi, E., Heydecker, B.G., De Dios Ortúzar, J., 2011. On the treatment of repeated observations in panel data: efficiency of mixed logit parameter estimates. Netw. Spatial Econ. 11 (3), 393–418. Yannis, G., Golias, J., Antoniou, C., 2006. Effects of urban delivery restrictions on traffic movements. Transp. Plan. Technol. 29 (4), 295–311.

Please cite this article in press as: Holguín-Veras, J., et al. Fostering unassisted off-hour deliveries: The role of incentives. Transport. Res. Part A (2017), http://dx.doi.org/10.1016/j.tra.2017.04.005