EXECUTIVE SUMMARY
In recent years, there has been a growing focus on enhancing sustainable logistics in urban environments Known as City Logistics, these initiatives leverage the coordination capabilities of municipal governments to encourage urban delivery companies to engage in collaborative efforts By minimizing truck trips and maximizing truck utilization, these schemes aim to mitigate the negative externalities linked to urban truck traffic.
While existing research primarily emphasizes freight models, transport networks, and urban freight project evaluations, there is a notable gap in understanding the behaviors of urban carriers and freight receivers regarding cooperative multicarrier delivery initiatives Limited studies have explored the behavioral modeling among freight agents, highlighting the need to investigate their interactions and how these relationships influence decision-making and policy implementation Addressing this gap is crucial for a comprehensive understanding of urban freight dynamics.
This report examines a cooperative multicarrier delivery initiative and its impacts at a disaggregate level, structured into two parts The first part analyzes freight delivery patterns in New York City and the factors influencing them, providing insights for the feasibility of implementing freight consolidation centers (FCC) in the city The second part explores the decision-making process for developing urban FCCs through an experimental economics approach, where participants representing various stakeholders receive cash bonuses to simulate real-world decision-making scenarios.
INTRODUCTION
The rapid advancement of emerging technologies has sparked a growing interest in enhancing the efficiency and sustainability of urban logistical operations While urbanization and globalization offer significant conveniences and economic benefits, they also contribute to challenges like traffic congestion and air pollution Consequently, researchers are increasingly focusing on developing effective and sustainable transportation systems In the realm of freight transport, implementing city logistics measures is crucial for achieving environmentally-friendly transport, as trucks significantly impact the environment negatively.
In major urban centers like New York City, the daily delivery and transshipment of goods significantly influence economic competitiveness However, the freight system also contributes to noise pollution and exacerbates traffic congestion To enhance efficiency while minimizing adverse effects, various strategies have been suggested, including the implementation of exclusive truck routes, off-hour delivery schedules, and the establishment of urban freight consolidation centers.
A Freight Consolidation Center (FCC) is defined as a facility that consolidates freight deliveries from outside the city and transships them to local receivers using smaller trucks with full loads, as noted by Zhou and Wang (2013) This approach can significantly reduce the number of truck deliveries, enhance truck load factors, and mitigate congestion and pollution (BESTUFS, 2007; Browne et al., 2005) Additionally, FCCs effectively tackle the "last mile" problem, which is often the costliest segment of delivery due to diminishing economies of scale once a vehicle exits the main road network (Lewis et al., 2010).
Urban freight transport faces numerous challenges, particularly in modeling activities that involve multiple stakeholders Key players in this sector include freight carriers, shippers, residents, administrators, and motorway operators, each with distinct objectives influencing their behavior To effectively evaluate city logistics measures, it is essential to consider the interactions and motivations of these stakeholders, as their diverse goals shape urban freight dynamics.
To effectively assess and implement freight delivery strategies, it is essential to analyze delivery patterns and understand truckers' behavior This report begins by examining the freight delivery patterns in New York City, considering various influential factors Data for this analysis was gathered through a field survey conducted by a research team at the City College of New York from January to June.
In 2013, a comprehensive analysis of freight delivery patterns was conducted through direct observations, which included delivery characteristics like location and vehicle configuration, as well as interviews with truck drivers regarding their tour origins, destinations, start times, durations, number of stops, distances traveled, and major roads utilized Statistical models were employed to analyze these delivery patterns and identify potential influential factors.
The second section of this report examines the decision-making process involved in establishing urban freight consolidation centers (FCC) among various freight agents Utilizing an experimental economics approach, the study simulates the decision-making dynamics among multiple stakeholders, with graduate students participating as representatives of different freight agents To incentivize engagement, participants receive a cash bonus for their involvement, along with potential profits linked to their actions during the game.
FREIGHT DELIVERY PATTERNS IN NEW YORK CITY
Introduction
To effectively implement freight delivery strategies, it is essential to analyze freight delivery patterns and understand truckers' behavior Various studies have explored these patterns, with models typically categorized into two types: commodity-based, which examines the flow of goods, and trip-based, which focuses on vehicle trips.
Chow et al (2010) expanded on the NCFRP 606 report by categorizing freight models into seven distinct groups: economic flow factor models, O-D factor models, truck models, four-step commodity models, economic activity models, logistic models, and vehicle touring models Among these, vehicle touring models stand out for their practical application, as they emphasize vehicle movement and carrier decision-making, providing a clear representation of the transportation system For instance, the truck tour-based microsimulation model in Calgary, Canada (Hunt and Stefan, 2007), analyzes tours based on various factors such as time period, travel purpose, vehicle type, start time, stop location, and stop purpose Key independent variables influencing these characteristics include the number of stops, travel time, zone accessibility, population, and employment.
This study analyzes freight delivery patterns through truck tours by developing statistical models based on data gathered from a field survey in New York City The models focus on key factors such as truck routes, dwell time at each stop, load factor, departure time, and the total number of stops Additionally, the research explores truckers' willingness to utilize Freight Consolidation Centers (FCC).
Data
The dataset analyzed in this study was collected through a field survey conducted by a research team at the City College of New York from January to June.
In 2013, researchers conducted a comprehensive study in various neighborhoods across Manhattan to analyze truck delivery characteristics This included assessing the delivery locations, vehicle configurations, arrival and departure times, and the vehicle owners as indicated on the trucks Additionally, they gathered data on vehicle load factors and the types of commodities being delivered whenever possible The study also featured in-depth field interviews with delivery drivers, utilizing a 21-question survey that explored tour origins and destinations, start times and durations, the number of stops, distances traveled, and major roads used during deliveries.
The study analyzed four key factors: the load factor of used vehicles at the truck's initial stop, associated costs, and the size of the company Additionally, researchers gathered drivers' insights on their employers' potential interest in joining a future urban consolidation center A detailed survey form is included in the Appendix, and the research compiled a total of 94 records, summarizing the variables derived from the survey data.
Table 1 Direct observation data summary
No of valid records Mean Std dev Min Max
LF_obs Observed load factor, i.e., percentage of capacity used 75 0.433 0.321 0 1
Zip Zip code of the company address 69
Loc Location of the company 93
Trip information Arrival Arrival time: 6:00 am to 23:00 70 11:56 2h59 m 6:00 23:00
Dwell Dwell time: derived from the difference between arrival and departure time 62 41.44 48.56 0 240
Commodity type Food 1 if commodity type=food; 0 otherwise 94 0.362 0.483 0 1
Drink 1 if commodity type=drink; otherwise 0 94 0.149 0.358 0 1
Other 1 if commodity type is neither food nor drink; 0 otherwise 94 0.213 0.411 0 1
Trip information LF_first Load factor at first stop 83 0.804 0.228 0.1 1
Derived from T_leave: 0 if leaves between 7pm to 6am; 1 if leaves between 6am to noon; 2 otherwise
T_first Time arriving first stop 64 9:46 3h57m 2:30 18:00
Duration of the entire delivery tour: the difference between leave time and final stop time
N_stop Number of total stops 62 35.5 92.1 1 540
Route Index of truck route 34 1.529 0.992 0 3
Size information Employee 1 if employee>P, 0 otherwise 53 0.623 0.489 0 1
Fleet 1 if fleet size> ; 0 otherwise 37 0.432 0.502 0 1
Willingness to participate joint distribution: 1-definitely not; 2- unlikely; 3-neutral; 4-possible; 5-likely
Models and Results
The spatial distribution of truck tours is influenced by the delivery route, while key indicators such as dwell time, total number of stops, and load factor play crucial roles in determining delivery efficiency Additionally, the start time reflects the temporal distribution of these tours The willingness of respondents to engage in a Freight Consolidation Center (FCC) serves as a vital attitudinal measure for evaluating the feasibility of implementing FCCs Together, these elements provide a comprehensive understanding of truck tours, offering valuable insights into the practicality of the FCC concept.
The interrelated factors in this study necessitate a sophisticated model; however, the small sample size of the dataset limits the development of such complexity To prevent over-parameterization, the research employs simple-form statistical models with a restricted number of controlled variables, each focusing on a specific factor These controlled variables encompass truck and company details, tour information, costs, and commodity types, as outlined in Table 2, which lists the factors for analysis, the models employed, and potential controlled variables.
Table 2 Overview of factors and models
Key indicators Models Controlled Variables
Truck route MNL Duration, total stops, fuel cost, commodity type
Dwell time Duration model Total stops, fleet size, fine cost, mileage, commodity type
Total number of stops Poisson Dwell time, load factor, mileage, commodity type, fleet size
Load factor at first stop Censored linear Truck configuration, total stops, commodity type
Start time MNL Truck configuration, total stops, commodity type
Willingness Ordered logit Number of employees, fleet size, fine cost, commodity type
The small sample size and limited data necessitate a careful balance between the consistency of model behavior and statistical validity This study represents the initial phase of a broader research initiative, showcasing findings from the first round of modeling efforts Future studies will refine the selection of controlled variables and improve model specifications.
A recent survey asked drivers to identify major routes used when entering the city, focusing on four key highways: I-78, I-95, I-278, and I-495 The collected data illustrates the spatial distribution of freight traffic, highlighting the significance of understanding the routes truck drivers choose and their potential connections to other influencing factors Visual representations in Figure 1 display the reported truck routes, while Figure 2 summarizes the frequency of these routes.
Figure 1 Truck routes in New York City
Nearly half of the 34 drivers interviewed reported traveling on I-95 for their routes, with fewer utilizing other major highways To examine the relationship between truck routes and various tour characteristics, a multinomial logit (MNL) model was developed Discrete choice models are prevalent in transportation and freight research; for instance, Garrido and Mahmassani (2000) employed a space-time MNP to predict freight demand across different seasons, yielding accurate forecasts Additionally, Rich et al (2009) utilized a weighted logit model to analyze mode choice and freight crossings, effectively decoupling agents and shipments In this study, the route was designated as the dependent variable, while duration, total stops, fuel costs, and commodity type were treated as independent variables, with MNL results summarized in Table 3.
Table 3 Estimation results of the truck route choice model
Longer tour duration is associated with a higher likelihood of using I-95, and lower likelihood of others, especially I-495 Apparently, trucks taking I-95 tend to make fewer stops
Trucks traveling on I-495 make more stops but benefit from lower fuel costs due to shorter delivery routes, despite higher fuel prices compared to alternative routes.
Dwell time refers to the total duration a truck spends making a delivery or pickup at a stop, serving as a key indicator of freight delivery efficiency and its external impacts During this period, trucks occupy limited road resources, and if left idling, they contribute to increased fuel consumption and emissions According to the survey data presented in Figure 3, most dwell times range from 10 to 30 minutes, with only a small number exceeding 30 minutes.
Dwell time is examined using hazard-based duration models, commonly utilized in biostatistics (Kalbfleisch and Prentice, 1980; Fleming and Harrington, 1991), and these models are also applied to analyze unemployment duration and business cycles (Kiefer, 1988) However, their application in transportation remains limited, with notable exceptions Bhat (1996) explored factors influencing shopping activity duration through a duration model based on grouped data, revealing that parametric baseline forms could lead to biased duration dependence estimates Similarly, Nam and Mannering (2000) employed hazard-based duration models to assess highway incident duration using data from Washington State’s incident response program, finding that detection, response, and clearance times significantly impacted overall incident duration.
The dwell time is analyzed using a cox proportional hazard duration model and the estimation results are shown below:
Table 4 Estimation results of the dwell time model dwelltime (minute) N_stop 1.162 *** (2.24)
Prob > chi2 0.0018 t statistics in parentheses Coefficients are hazardous ratios in duration model
Frequent stops by trucks are positively correlated with shorter dwell times, suggesting that trucks that make more stops are likely to have reduced duration Additionally, high fine costs and long-distance tours contribute to this trend of shorter dwell times.
The total number of stops is a crucial metric in assessing truck tours, influenced by fleet operations and the types of goods delivered To analyze this, variables related to the fleet, delivery routes, and commodity types are considered as explanatory factors According to the survey data presented in Figure 4, the majority of trucks make fewer than 30 stops, although there are some outliers that exhibit significantly higher stop frequencies.
Figure 4 Distribution of total number of stops
Count data models are employed for analyzing datasets characterized by small, non-negative integer values, such as 0, 1, 2, or 3 These models effectively handle the unique properties of count data, making them suitable for various analytical applications.
Transportation involves various factors that can be quantified through count data, including queue length, household vehicle ownership, and accident frequency on road segments Hellström (2006) introduced a bivariate count data model to analyze household tourism demand, while Ulfarsson and Shankar (2003) utilized a negative multinomial model to forecast traffic accident occurrences Their findings indicated that the negative multinomial model outperformed alternative models, such as the negative binomial and random-effects negative binomial models.
This study uses a Poisson model to analyze the total number of stops, and the results are summarized in Table 5
Table 5 Estimation results of the total number of stops model
A high load factor is linked to fewer stops, likely due to its association with bulk goods that require limited stops Specifically, a coefficient of -0.4890 indicates that a 1% increase in load factor results in approximately a 0.5% decrease in total stops Additionally, longer dwell times and increased tour mileage correlate with fewer stops In terms of commodity types, trucks delivering drinks exhibit a strong positive relationship with stop frequency, indicating they tend to make more frequent stops Conversely, fleet size negatively impacts stop frequency, suggesting that trucks within larger fleets make fewer stops.
Load factor is the percentage of a delivery vehicle's carrying capacity utilized at departure from the depot, indicating the vehicle's efficiency A higher load factor typically correlates with increased delivery efficiency, influenced by factors such as vehicle configuration, commodity type, and delivery tour patterns In this survey, 40 out of 83 responses indicated that trucks are fully loaded upon leaving the depot, while only 16 responses reported a load factor below 60%.
Figure 5 Distribution of load factor at first stop
The Tobit model, which accounts for censored linear relationships, is employed to analyze the load factor constrained between 0 and 1 Key explanatory variables in this study include truck configuration, commodity type, and delivery tour patterns, represented by the number of stops The estimation results for the load factor model are presented in Table 6.
Table 6 Estimation results of the load factor model load
Summary 1 left-censored, 56 uncensored, 4 right-censored
Summary
This study analyzes freight delivery patterns in New York City using six statistical models based on a field survey of truck drivers The models assess various independent variables, including route characteristics, dwell time, total stops, load factor, start time, and drivers' perceptions of their company's willingness to utilize a Freight Consolidation Center (FCC) Although the conclusions have limitations, the findings highlight the models' effectiveness in identifying freight patterns and evaluating policy intervention feasibility Notably, the willingness model predicts FCC participation based on specific characteristics, while the load factor model reveals that small combination trucks carrying drinks may have unused capacity, indicating potential for FCC involvement Additionally, the start time model indicates that trucks with frequent stops typically begin their routes between 6 am and noon, suggesting challenges in shifting these tours to off-peak hours due to the need for behavioral changes among receivers.
To enhance the study, increasing the sample size and incorporating additional variables is essential, which can be accomplished through improved handling of missing data and the integration of diverse data sources Furthermore, refining the selection of controlled variables and model specifications is necessary, adhering to a more stringent model development process However, the applicability of the models remains constrained by certain limitations.
16 small sample size, the findings from this preliminary study will provide important reference for future large-scale data collection and analysis, eventually facilitating the city’s freight policy design.
DECISION-MAKING PROCESS FOR FCC DEVELOPMENT
Introduction
Implementing Freight Consolidation Centers (FCCs) presents challenges due to the involvement of multiple stakeholders in the freight transport system Lindholm (2010) developed a Sustainable Urban Transport Plan (SUTP) freight transport model based on various FCC case studies, identifying key elements such as types of goods, vehicles, facilities, and infrastructure, as well as external factors like financial constraints, land use concerns, and pollution issues Additionally, factors such as location and organization type significantly impact FCC development; a strategically chosen location can enhance delivery efficiency while minimizing traffic congestion and pollution Conversely, an inadequate location can exacerbate freight transport challenges The nature of the organization also plays a crucial role; self-sustained FCCs with no government funding operate independently, whereas those receiving government support face more complex stakeholder interactions This study explores how these factors influence FCC development and the utilities of involved stakeholders.
Existing studies by the FCC primarily focus on specific case studies and logistics supply chain analyses A literature review by Browne et al (2005) indicates that an FCC can be highly effective if it has adequate funding, strong public engagement, and minimal congestion and pollution Panero et al (2011) presented detailed FCC case studies from Europe, exploring their applicability to the U.S Notable examples, including La Petite Reine in France, Heathrow Airport in the U.K., and the Tenjin Joint Distribution System in Japan, have been extensively analyzed regarding their operational efficiency, financial viability, and social impacts, providing valuable insights for this research.
In the field of logistics and supply chain studies, Kayikci (2010) introduced a conceptual model for FCC location decisions, utilizing a combination of fuzzy-analytical hierarchy process (AHP) and artificial neural network (ANN) methods to identify the most influential factors and select optimal locations Additionally, Moon et al (2011) developed a joint replenishment and consolidation freight model, demonstrating through mathematical modeling and four algorithms that a quasi-stationary policy resulted in lower total costs compared to a stationary policy.
Method
This study employs experimental economics to explore the FCC development problem, a method that examines human behavior in a controlled laboratory setting While some critics argue that laboratory conditions do not accurately reflect real-world decision-making and may not represent the entire population, experimental economics offers significant advantages, including replicability and control over the decision-making environment.
1993) Replicability allows other researchers to reproduce the experiment and validate the results
Researchers can manipulate the experimental context to explore key factors influencing decision-making However, FCC research often relies on diverse data types that may not be readily accessible When data is lacking, the experimental economics approach is essential for generating synthetic data and analyzing decision outcomes.
Experimental economics has demonstrated its relevance in freight transportation, as evidenced by Holguin-Veras and Thorson (2003), who examined the urban freight market by having participants simulate competing truck companies aiming to maximize profits Their study effectively aligned estimated stops, load factors, and time durations with theoretical expectations Similarly, this research will utilize experimental economics to analyze the FCC decision problem by deriving profit functions from empirical data for four key stakeholders: carriers, operators, government, and residents In this experimental setup, these stakeholders, representing real-world collaboration, strive to maximize their bonuses through strategic bidding while making necessary compromises for group consensus Eight scenarios will be tested to assess the impact of organizational type, location, and carrier size on stakeholders and the FCC development decision The findings will shed light on critical factors influencing the FCC development process and offer valuable insights for future research.
In certain FCC schemes, receivers directly pay for FCC services; however, in most cases, they do not This discrepancy leads to challenges in the distribution of benefits and costs among stakeholders.
In this study, it is assumed that FCC operator charges rent from carriers only, and that receivers are not involved
The experiment is limited to local actors, overlooking the influence of various layers of governance, including state, multi-jurisdictional, and national governments, which may interact and impact one another.
To simplify the problem, we only consider carriers, operators, government and residents from local area
Experiment Design
The experiment aims to evaluate the impact of various FCC conditions on stakeholder decisions, providing valuable insights into the practical decision-making process of FCC development Involving four key players—carriers, operators, government, and residents—the experiment consists of eight sessions, each corresponding to a distinct FCC proposal characterized by specific organizational types, locations, and carrier sizes Each session comprises ten rounds of bidding, with sessions deemed undesirable if no agreement is reached within these rounds.
Participants in the game receive a base payment of $20, but their total earnings depend on performance-based bonuses Each round offers the chance to earn a $1.00 bonus for maximizing profit; if a player's bid is profitable but not optimal, the bonus is calculated proportionally based on actual versus maximum profit For instance, if the maximum profit is $1000 and the player earns $800, they would receive an $0.80 bonus, contingent on achieving consensus among all stakeholders This incentive structure reflects real-world dynamics, where stakeholders prioritize their own benefits and only support proposals that serve their interests If consensus is not reached, no bonuses are awarded.
To analyze the interactive relationships among stakeholders, we define a profit function tailored to each stakeholder The initial parameter values for these functions are primarily derived from the La Petite Reine (LPR) freight consolidation center case study conducted in France (Panero et al.).
Established in 2001, this urban distribution system has become one of the largest in Europe, handling nearly 250,000 parcels annually Utilizing small cargo cycles for the transshipment of goods from trucks to local receivers within a 15-mile radius, it enhances efficiency while minimizing congestion and pollution The operation relies on minimal government support and offers competitive rental rates, with additional revenue generated through advertising on cargo cycles Overall, this successful facility serves as a model for effective urban distribution, providing valuable empirical data.
This study treats carriers as a single entity with a unified delivery route, assuming uniformity among all carriers involved Based on the work of Arnott et al (1993), the profit function for these carriers is defined by the cost savings achieved through reduced delivery distance and time, minus the total rent paid to the FCC It is further assumed that, in the absence of the FCC, a truck would need to traverse the city in a circular route to complete all deliveries, with the circle's radius reflecting the average delivery distance to the FCC Consequently, the profit function for carriers can be formulated accordingly.
N = number of deliveries per year
D = reduced distance traveled per delivery (km)
T= reduced travel time per delivery (hour) r c = rent for carrier ($/parcel)
= unit cost related to delivery distance, including fuel cost, insurance and maintenance ($/km) The estimated parameter values are presented in Table 10, below
Table 10 Parameter Values for Carriers’ Profit Function
5*1.59*55/24.22 $/hour (BLS, 2012 and Holguín-Veras, 2010)
0.49 $/mile (Holguín-Veras and Polimeni,
∆ 15km/(25km/h)=0.6h (Panero, et al, 2011)
180 deliveries per week * 52 weeks60 deliveries per year (Panero, et al,
The original values have been adjusted to current figures using the inflation factor from the Bureau of Labor Statistics (BLS, 2012) To convert the commuter's value of time to that of truckers, a factor of 55/24 is applied (Holguín-Veras, 2010) The distance traveled by cargo cycles from the FCC and the corresponding travel time are denoted as ∆ For consistency, key parameters are sourced from the La Petite Reine (LPR) freight consolidation center case study conducted in France (Panero et al., 2011) In this experiment, rent is designated as the variable for the carrier.
The operator's revenue primarily derives from rental income, as noted by Panero et al (2011) Additionally, in scenarios with significant public involvement, government financial incentives may contribute to the operator's overall revenue The operational costs of the FCC are determined by labor expenses, which are a key component of the total operational budget.
20 total costs When there is only one operator, the profit function for the operator could be expressed as: o o o o /
The operator profit (P o) is influenced by several factors, including the financial incentive for the operator (f o) and the labor cost percentage (l) in the total operation cost Additionally, the number of employees (n) and the wage rate for the operator (w o) play significant roles, along with the total working hours per year (h) and the rent for the operator (r o) calculated per parcel Understanding these variables is essential for optimizing operational efficiency and profitability.
The estimated parameter values are presented in Table 11, below
Table 11 Parameter Values for Operator Profit Function
Parameters Values n 10 (Panero, et al, 2011) h 13h/day*6 days/week*50 weeks900 hours/year (Panero, et al, 2011) l 0.83 (Panero, et al, 2011)
In a case study of La Petite Reine's freight consolidation center in France, key parameters such as financial incentives, rent, and wage rates were identified as variables for the operator in the experiment This approach highlights the significance of optimizing these factors to enhance operational efficiency and drive better outcomes in freight consolidation.
In this study, "government" encompasses the collective public agencies dedicated to serving the entire region affected by the FCC The government's profit is derived from minimizing externalities related to pollution and congestion (Panero et al., 2011) Major pollutants, including carbon dioxide and nitrogen oxide, are quantified in monetary terms to assess their impact.
The quantification of congestion externalities can be effectively addressed through congestion pricing and the redistribution of toll revenue, as suggested by Mirabel and Reymond (2011) It is assumed that the congestion resulting from truck deliveries corresponds to the tolls paid by these trucks Consequently, the government's profit function can be articulated to reflect this relationship.
∆CO2=reduced carbon dioxide (ton)
∆NOx=reduced nitrogen oxide (ton) a = CO2 price ($/ton) b = NOx price ($/ton)
N = number of deliveries per year f = financial incentive for the government ($)
The estimated parameter values are presented in Table 12 below
Table 12 Parameter Values for Government Profit Function
Parameters Values t $30+$150=$180 (Holguin-Veras and Polimeni, 2006)
200 kg=0.2 ton (Panero, et al, 2011) a 20 $/ton (Johnson, L et al., 2011) b 300 $/ton (FERC, 2012)
In this study, we assume that 5-axle trucks are utilized for peak hour deliveries, incurring a toll cost of $30 and a parking fine of $150 (Holguin-Veras and Polimeni, 2006) To ensure consistency, key parameters are derived from the La Petite Reine (LPR) freight consolidation center case study conducted in France (Panero et al., 2011) Additionally, the financial incentive is established as a variable for government intervention in this experiment.
Residents near freight container terminals (FCC) often oppose their development due to valid concerns While FCCs generate local employment opportunities, reflected in wage rates, they also bring challenges such as increased noise and diminished community safety and vibrancy These negative externalities are likely to influence land prices in an open market Lin and Ben (2009) employed an enhanced hedonic price model to analyze these impacts.
Industrial land agglomeration, exemplified by FCC construction, significantly influences land prices, a concern for residents who are directly impacted by the value of their property The profit for local residents is derived from both wage income and fluctuations in land prices, highlighting the interconnectedness of local economic factors.
= land price elasticity ($/percentage change in industrial land)
P 0=original land price ($/m 2 ) n= number of employees w r = wage rate ($/hour) h= working hour per year (hour)
The estimated parameter values are presented in Table 13, below
Table 13 Parameter values for residents profit function
2*1.2287=$54.3/m 2 (Abelairas-Etxebarria, P., & Astorkiza, I., 2012 and Google Finance, 2012)
Note: Land area value comes from La Petite Reine (LPR) freight consolidation center case study in
France (Panero, et al, 2011) w r * indicates wage rate is set as the variable for residents in the experiment
Figure 8 illustrates the interactions among four players, highlighting how a carrier interacts with the operator through rental bids Both the carrier and operator submit bids, denoted as r*, which ultimately influence the profits of both carriers and the operator, as indicated by the double arrow Similar interactions occur between the operator and other players For simplicity, it is assumed that carriers, the government, and residents do not engage in direct interactions with one another.
Figure 8 Interactions between Freight Agents
Experiment Implementation
In a recent experiment, four graduate students were randomly assigned roles as carrier, operator, government, and resident to explore decision-making dynamics Each participant aimed to maximize their profit by selecting values for three key variables: rent charged by the operator, financial incentives provided by the government, and wage rates determined by the operator and resident Consensus among the group was achieved when the bids for all variables differed by no more than 5% Participants had up to two minutes per round to discuss and adjust their bids, with a total of ten rounds to reach agreement; failure to do so indicated an unfavorable scenario for establishing a Federal Communications Commission (FCC).
This study explores eight scenarios to evaluate various organizational types, location choices, and carrier sizes in relation to service delivery The type of organization influences the partnerships and financial incentives between operators and government entities, categorized as private (no financial incentive), public (full financial incentive), and public-private partnerships (PPP) which offer partial financial incentives (Panero et al., 2011) Given the effectiveness of PPPs, this research focuses on comparing private organizations with PPPs The location of service delivery, as defined by the FCC, plays a critical role; urban areas face challenges like congestion and pollution despite shorter delivery distances, while suburban areas reduce these externalities and save on travel distance The study contrasts outskirt locations, which imply longer travel distances, with center locations that suggest shorter last-leg distances Additionally, "carrier size," determined by the number of truck deliveries managed by the FCC, reflects the carriers’ acceptance and utilization of the FCC services, with two size categories considered: small and large.
250,000 parcels and 9,360 deliveries, or large size with 750,000 parcels and 28,080 deliveries (Panero, et al, 2011) The complete scenario information is summarized in Table 14
To ensure nonnegative profit, it is crucial to consider the variable range, which varies based on the rent levels observed in different case studies (Panero et al., 2011) Additionally, the estimated freight volume and delivery counts for large carriers are derived from Nemoto's 1997 study Wage rates are set according to the minimum wage standards established by the Department of Labor in 2012, alongside a consideration for employees earning a high annual income of $100,000.
Results
The experiment was carried out successfully, with four players The bidding values and group consensus results are summarized in Table 15
Table 15 Bidding Values and Group Consensus Results
* Number of group consensus achieved
Note: The “*” indicates group consensus is reached
Table 15 reveals that consensus was reached multiple times in all scenarios except the third The variation in the number of group consensus instances appears to be influenced more by player experience than by differing FCC conditions, as players exhibited a conservative approach.
26 the first two scenarios, and became too drastic in scenario 3, and eventually grew more experienced and rational in the last several scenarios
The Federal Communications Commission (FCC) influences bidding prices and profit margins through factors like organizational type, location selection, and carrier size An analysis of various scenarios, taking these control factors into account, reveals significant differences in bidding outcomes and profitability, as summarized in Table 17.
Table 16 Impacts of Different Factors on Bidding and Profits
Average bidding prices ( : $/parcel, : million
1 vs 2 (outskirt, small carrier size)
:0.42 vs 0.20 :1.80 vs 0.86 :-0.45 vs 0.82 :1.05 vs 1.02 decreases increases
3 vs 4 (center, small carrier size)
: NA vs 0.68 : NA vs 0.66 : NA vs 12.49
:NA vs 0.28 :NA vs 1.18 :NA vs 0.39 :NA vs 0.52
5 vs 6 (outskirt, large carrier size)
:1.23 vs 0.85 :5.40 vs 4.72 :1.94 vs 2.94 :0.62 vs 0.74 decreases increases
7 vs 8 (center, large carrier size)
:0.26 vs 0.82 :5.40 vs 4.67 :0.74 vs 0.70 :0.49 vs 0.75 decreases slightly decreases
1 vs 3 (private, small carrier size)
: 3.10 vs NA : 0 vs NA : 27.06 vs NA
:0.42 vs NA :1.80 vs NA :-0.45 vs NA :1.05 vs NA
5 vs 7 (private, large carrier size)
Lower Lower Lower Lower Lower
6 vs 8 (public- private, large carrier size)
: NA vs 1.45 : NA vs 0.00 : NA vs 12.05
:NA vs 0.26 :NA vs5.40 :NA vs 0.74 :NA vs 0.49
Higher Higher Higher Note: The “NA” indicates data is unavailable
Table 17 illustrates that the shift from purely private organizations to public-private partnerships (PPP) results in a notable decline in government profits Specifically, government profit drops by 12.6%, decreasing from $5.40 million to $4.72 million when the organizational type changes while the FCC remains in the outskirts with a small carrier size (comparing scenarios 5 and 6) In an extreme scenario (1 vs 2), where only the organizational type changes while the FCC is still located in the outskirts and the carrier size is large, the government profit plummets by 52.2%, from $1.80 million to $0.86 million This finding aligns with the negative correlation between financial incentives and government profit as indicated in equation (3).
The operator's profit significantly increases in private-public partnerships, with gains ranging from 51.5% in scenarios 5 vs 6 to approximately 282% in scenarios 1 vs 2, aligning with the positive correlation suggested in equation (2) Notably, local residents also benefit from this partnership model through higher wages, as the financial incentives for operators are partially allocated to labor costs, particularly evident in scenarios 5 vs 6 and 7 vs 8 Overall, the findings indicate that public sector involvement in FCC organizations enhances benefit redistribution, making FCCs more appealing to both operators and local communities However, it is important to note that increased operator profits do not automatically result in lower rent, as evidenced by a 15.4% rent increase in scenarios 5 vs 6.
6) to 29.0% (scenarios 1 vs 2) From the perspective of policy design, using financial incentives does not seem to effectively lower rent for carriers
The study indicates that freight consolidation centers (FCCs) situated on city outskirts are more appealing to carriers due to reduced travel distance and time, as evidenced in various scenarios Although central locations can offer extremely low rents, which can attract carriers, this strategy ultimately diminishes operator profits and leads to lower wages and resident profits Outskirt locations not only minimize travel for trucks but also tend to offer lower rental costs and exert less impact on local communities However, this does not imply that all FCCs should be placed in suburban areas; cities can also benefit from establishing FCCs in urban settings by repurposing brownfields and enhancing short line railroad usage Overall, the findings suggest that, for nearly all stakeholders, outskirt locations are generally more favorable than central ones.
The analysis reveals a positive correlation between carrier size, which serves as a proxy for FCC utilization rates, and both total carrier profit and government profit Larger carriers experience increased cost savings and enhanced government profits, while operators also benefit from higher revenues linked to greater FCC utilization However, carrier size does not significantly affect rent or financial incentives, with rent remaining stable across various scenarios Financial incentives fluctuate, showing a decrease of 26.8% in one scenario and an increase of 16.7% in another Ultimately, while all stakeholders gain from higher utilization rates, operators emerge as the primary beneficiaries, gaining increased negotiation power in the market.
Table 17 summarizes the effects of organizational type, location choice, and carrier size on bid prices and profits It is evident that the optimal conditions for FCC development involve a public-private partnership, a location on the outskirts, and a larger carrier size.
(Change from private to private- public)
(Change from central to outskirt)
(Increase) NI NI NI NI
Note: “ ” means positive impact, “ ” negative impact and “NI” no significant impact.
Summary
This study employs experimental economics to examine the factors affecting stakeholder profits in freight consolidation center (FCC) development decisions Four key players—carriers, operators, government, and residents—bid on rent, financial incentives, and wages in various scenarios, aiming to maximize their profits and reach consensus While organizational type, location, and carrier size do not significantly impact consensus due to player inexperience, they do influence individual profits Specifically, outskirt locations and larger carrier sizes yield higher profits for carriers and operators, while also benefiting residents through increased profits, rent, and wages The study offers a valuable framework for analyzing FCC development and the effects of various factors, although it acknowledges limitations such as simplified profit functions and reliance on the LPR case study for parameter values Variations in freight volume and delivery numbers based on FCC size and location may lead to differing profit structures Additionally, the study notes that subtle interactions between players, such as the lack of direct communication between local residents and the government, may not be fully captured, and players' experience levels can also affect results.
30 its influence on their bonuses until after a few iterations Several test iterations could be run in the future experiment
Future research will focus on creating more theoretically grounded cost functions and robust empirical parameter values It is essential to enhance player training and address "burnt" samples resulting from misunderstandings during experiments Additionally, incorporating a wider range of factors and participants is crucial, as FCC development decisions are influenced by various elements like local economies and transportation conditions The analysis prototype established in this study facilitates an initial exploration of the FCC development decision-making process The insights gained will aid practitioners in understanding stakeholder interactions, and with further refinement, this framework has the potential to significantly enhance FCC planning and decision-making, ultimately contributing to the advancement of sustainable freight transportation systems.
CONCLUSIONS
The report analyzes cooperative multicarrier delivery initiatives and their impacts at a disaggregate level, focusing first on current freight delivery patterns in New York City A survey was conducted to gather comprehensive data on truck configurations, delivery routes, load factors, operational costs, and truckers' willingness to adopt FCC Statistical models were developed to explore the relationships among these factors, particularly those affecting dwell time, load factor, departure time, and the total number of stops—critical indicators of FCC efficiency and feasibility The findings will offer valuable insights into freight delivery patterns in New York City, serving as a vital reference for the city's freight policy development.
Future work will focus on gathering more detailed data, as the existing dataset is limited in sample size and freight delivery information, with several incomplete responses Additionally, testing alternative models and refining current ones will enhance the comprehensiveness of the results.
This article explores the role of experimental economics in identifying factors that affect stakeholder profits in freight consolidation center (FCC) development By defining profit functions for key stakeholders—carriers, operators, government, and residents—four players engage in bidding on rent, financial incentives, and wages to optimize their profits Eight scenarios are analyzed to uncover influential factors and suitable conditions for FCC decision-making Findings indicate that public-private partnerships can reduce rent and increase wage rates, resulting in higher profits for carriers, operators, and residents Additionally, a central location is shown to decrease rent and wages, impacting overall profit distribution.
Larger carrier sizes enhance profits for all stakeholders involved, creating a strong financial incentive for collaboration In conclusion, the ideal conditions for the development of FCC include public-private partnerships, strategic outskirt locations, and the utilization of larger carrier sizes.
Future work will focus on enhancing profit functions for all players, as the existing models are overly simplified By incorporating additional variables, we can better reflect the complexities of the market Furthermore, future games could explore the dynamics of multiple carriers with diverse characteristics and consider various groups of residents to create a more comprehensive simulation.
APPENDIX: SURVEY FORMS
Information to be observed visually:
Location of the vehicle Time of survey Surveyor
1 Number of axles: ; Number of tires:
2 Vehicle Configuration: ; (Samples provided van, single unit, tractor-trailer)
3 Estimated load factor: % of capacity filled with cargo;
4 Name of the company: ; Zip Code:
If possible to be done discreetly, take a picture of the vehicle
Full Load in First stop? Load Varies Every time? Reload or not? Yes: 100%-95%
No, how full: o 95%-80% o 80%-50% o 50%-30% o Less than 30%
Leave depot First Stop Final stop
Already completed Yet to be completed Varies daily or not?
Yes Average Number of stops :
IV Only delivery in Manhattan?
No, other place: o Brooklyn o Bronx o Queens o Long Island o Staten Island o Other: ;
V Can you estimate the average distance you traveled: Miles/day
What is your origin: o I 78 o I 95 o I 278 o I 495 o Other:
Tunnel o Queens Midtown o Lincoln o Holland o Brooklyn Battery or
Bridge o George Washington o Third Avenue o Willis Avenue o Triborough o Queensborough o Williamsburgh o Manhattan o Brooklyn
Name of your company Type of goods
Parking Cost in Manhattan Parking fine Double Park
How often did you receive :
Daily: times Weekly: times Monthly: times
How often you have to double park:
Never 1-5 times/day 5-10 times/day Always
Can you estimate your daily fuel cost: $ ;
How likely do you think your company will use joint distribution if necessary facilities provided?