Amazon says a Mamdani-backed employment mandate could raise delivery costs and put thousands of jobs at risk. City Hall says the company’s subcontracting system shifts responsibility away from the corporation that controls the work. A review of the bill and its supporting records shows that the headline number is a forecast, not a fee, and its outcome depends heavily on what delivery companies choose to do next.
Published at 12:01 p.m. EDT
No New York City household has received a $664 bill because of the Delivery Protection Act. The legislation has not passed, its direct-employment rule is not in effect, and the text does not create a $664 tax or delivery charge.
The number is nonetheless at the center of one of the city’s most consequential fights over online shopping, worker protections and corporate power. Amazon says Intro 0518-2026, a bill backed by Mayor Zohran Mamdani, could add $664 a year in delivery costs for the average city household if last-mile facilities relocate beyond the five boroughs. That equals about $55.33 a month, or $12.77 a week, according to calculations for this article.
Amazon’s warning deserves attention because moving delivery stations farther from customers can add miles, labor time and fuel while reducing the number of packages carried on each route. It does not, however, deserve to be treated as an inevitable price tag. The company’s estimate rests on a modeled response to the bill, particularly a full-relocation scenario, rather than a charge written into the legislation.
That difference is the heart of the story.
The proposal would require operators of covered last-mile warehouses and storage sites to obtain a city license and directly employ workers who perform core warehouse and delivery services. It would generally prohibit third-party contracting for that work. Amazon’s Delivery Service Partner system relies on separate businesses, known as DSPs, that employ drivers who deliver Amazon packages. The bill would force a major restructuring of that arrangement at covered city facilities.
Amazon says the change could make the current network uneconomic and lead it to consider moving operations to New Jersey, Long Island or Westchester. Mamdani and the bill’s sponsor, Council Member Tiffany Cabán, argue that direct employment would attach responsibility to the company that shapes routes, quotas and working conditions.
For New Yorkers, the practical question is not whether the bill carries costs. Meaningful labor and safety rules usually do. The real questions are how large those costs would be, whether relocation is necessary, how much companies would absorb, and what value residents place on the protections the bill seeks to create.
What the bill actually does
The official City Council docket lists Intro 0518-2026 as “Laid Over in Committee” following an April 9 hearing. As of publication, it is a proposal, not a mandate currently governing deliveries. The docket shows 34 sponsors, a sign of substantial political support, but sponsorship is not the same as final passage.
The introduced text would create a licensing system administered by the Department of Consumer and Worker Protection. Each covered facility would pay an annual licensing fee of $500. The bill would require direct employment of core warehouse workers and delivery workers who work more than eight hours in a 30-day period. It also includes annual training, recordkeeping, anti-retaliation rights, notice before most terminations and enforcement provisions.
The direct-employment provision would take effect in the twelfth month after the measure became law. Existing contracts could receive limited extensions, but no later than the twenty-fourth month after enactment. That phase-in matters because it would give companies time to hire workers, renegotiate arrangements, revise routes or challenge the law.
The text also complicates claims that all current drivers would simply disappear. If a facility operator ends a contract because of the law and the subcontractor terminates the affected workers, the operator must offer those workers employment before hiring other people for the same work. The offers must preserve specified rights and provide terms no worse than those in effect at termination.
That is not a guarantee that every existing position, schedule or small business would survive. DSP owners could still lose the enterprises they built, and workers might face disruption even if they receive job offers. It does mean that a forecast of thousands of jobs “at risk” is not identical to a forecast of thousands of workers becoming permanently unemployed.
Where the $664 estimate comes from
Amazon told Fox Business that an analysis by consulting firm AKRF modeled a 267 percent increase in per-package costs for deliveries now handled by city facilities under full relocation. The company also cited projected service declines of 10 percent to 21 percent and said $664 in annual delivery costs could be passed to each city household.
The qualifications matter. “Full relocation” is a scenario in which covered operations move, not a requirement imposed by Intro 0518. The bill regulates facilities that operate in the city, but it does not order Amazon, FedEx or another carrier to leave. Relocation is one possible corporate response alongside direct hiring, network redesign, negotiations over amendments, litigation, reduced service or some combination of those options.
There is also no automatic mechanism that would spread an identical $664 charge across every household. A resident who places two online orders a year and a resident who receives several packages a week would not naturally incur the same direct delivery expense. To turn a network-wide cost into a uniform household average, an analysis must make assumptions about order volume, the share of deliveries affected and the degree of pass-through in membership prices, seller fees, product prices or delivery charges.
Amazon’s public statement, as reported by Fox, does not fully explain those assumptions. Until the model’s inputs are independently tested, the $664 figure should be presented as a commissioned projection under a particular operating scenario, not as an audited consumer bill.
One comparison is especially revealing. The statute’s explicit licensing fee is $500 per facility per year. That modest fee cannot by itself produce a citywide household cost of $664. The large number comes from modeled operational changes, especially longer routes after relocation, along with assumptions about how those costs reach customers.
Follow the funding
AKRF prepared its impact assessment for the Five Borough Jobs Campaign, a business advocacy organization opposing the bill. New York Focus reported that the campaign paid the consultant $52,500 for the study.
The source of that campaign’s money is part of the public record. A January through June 2026 filing with the New York State Commission on Ethics and Lobbying in Government lists five Amazon contributions totaling $5,001,377 to the Five Borough Jobs Campaign. The filing also lists smaller contributions from FedEx, Prologis Management and the Trucking Association of New York.
New York Focus further reported that the Five Borough Jobs Campaign spent millions under the banner of the New York Delivers coalition. The coalition has featured local delivery businesses and workers in its opposition campaign. Amazon spokesperson Steve Kelly told the outlet that the company was proud to stand with its partners against legislation it believes would harm consumers, small businesses and employees.
Funding does not prove an analysis is wrong. Consultants frequently work for clients with a stake in the outcome. It does tell readers whose question the report was designed to answer and why independent review is essential. The proper label is an industry-commissioned analysis financed by a campaign overwhelmingly backed by Amazon, not a neutral city forecast.
That disclosure is particularly important because the $664 figure has migrated from a scenario in a report into advertisements, political arguments and news headlines. Repetition can make a projection sound like a settled fact even when the underlying policy has not passed.
The worker and street-safety case
Mamdani announced his support on August 10, saying the bill would impose “commonsense regulation” and give delivery workers “dignity, stability and a safe workplace.” In its official statement, City Hall argued that large corporations exercise substantial control over routes, productivity demands and workplace expectations while leaving formal employment to contractors.
The administration’s case is not based on rhetoric alone. A November 2025 report from the city comptroller’s office examined 18 large last-mile facilities that opened between 2017 and 2022. Fourteen, or 78 percent, had increases in injury-causing crashes within a half-mile after opening. Across all sites, injuries in that radius rose by an average of 16 percent, the report found.
The same report analyzed federal injury records and calculated a total recordable incident rate of 8.3 injuries per 100 full-time-equivalent workers at identified city last-mile facilities from 2022 through 2024. The report said DSPs had a reported injury rate above nine per 100 workers.
Those findings support the city’s claim that the industry presents serious safety concerns. They do not, on their own, prove that opening a warehouse caused every nearby crash. A before-and-after comparison can capture other changes in traffic, development and travel patterns. The data establishes a troubling association that justifies scrutiny, while a stronger causal claim would require additional controls and analysis.
It is also fair to ask whether direct employment alone would reduce crashes. Licensing, training, route design, vehicle technology and enforcement could each affect safety. A credible final bill should connect its requirements to measurable outcomes, publish data after implementation and permit revisions if the rules fail to produce safer streets.
Amazon’s case for its delivery partners
Amazon says Intro 0518 would threaten more than 40 New York City DSP businesses and more than 5,000 people they employ. In written testimony submitted to the City Council, the company said DSP owners make decisions about hiring, fleets and capacity. It reported that city DSP drivers had earned an average of nearly $24 an hour in wages since January 2025 and said full-time employees must be offered health coverage.
Those figures come from Amazon and should be read as company-reported data. Still, they challenge any simplistic picture of every DSP driver as an independent gig worker. The DSP companies employ drivers, often as W-2 workers, and their owners handle teams and payroll. The dispute is about whether those separate employers are genuinely independent businesses or an intermediary layer in a network substantially controlled by Amazon.
New York Delivers has warned that more than 10,000 jobs across the broader local delivery system could be at risk. That larger number covers more than Amazon’s 5,000-plus DSP employees and should not be presented as an Amazon-only job count. Neither figure establishes how many workers would ultimately lose employment after the bill’s rehiring requirement, corporate adjustments and any amendments are considered.
Amazon also points to safety improvements. Its Council testimony said the serious-crash rate for its city operations improved 35.7 percent from 2024 to 2025 and highlighted driver training, in-vehicle safety systems and more than 800 electric cargo bikes in Manhattan and Brooklyn. Again, these are company claims. To reconcile them with the comptroller’s broader findings, policymakers need comparable definitions, route exposure data and facility-level reporting.
The choice hidden inside the warning
The strongest part of Amazon’s argument is logistical. A delivery station placed close to dense customer clusters can send out shorter, fuller routes. Move that station across a bridge or county line and each vehicle may spend more time reaching its first stop and returning after its last. More miles can mean more fuel, labor and congestion.
The weakest part is inevitability. The company has not said that full relocation is its only feasible response. Its own testimony says it would examine options, including potential relocation. That language leaves open direct employment, partial relocation, changed service levels and negotiations over the bill’s terms.
The consumer result also depends on competitive choices. A carrier can raise delivery fees, alter a membership price, charge sellers more, absorb part of the expense, reduce same-day availability or shift costs into product prices. Rival retailers and carriers may respond differently. The final burden would probably vary by company and by household ordering behavior rather than arrive as one identical annual invoice.
Supporters face their own unanswered question. If the Council believes the industry’s cost estimate is inflated, it should produce or commission a transparent alternative. Calling every price warning a corporate choice does not quantify the expense of direct hiring, training, compliance, insurance or network changes. Worker protections can be worth paying for, but policymakers still owe the public a credible price range.
What happens next
As of September 5, the Council record shows no final vote and no enacted law. The measure remains in the Committee on Consumer and Worker Protection. Cabán and Mamdani are pressing for passage, labor groups are mobilizing behind it, and Amazon and its allies are spending heavily to stop or change it.
Before a vote, the Council should demand three things in public: the detailed assumptions behind the $664 estimate, an independent assessment of relocation and pass-through scenarios, and amendments that protect current workers during any conversion from DSP employment. Lawmakers should also explain how the $500,000 bond requirement for each delivery worker retained as a nonemployee under permitted exceptions would operate in practice.
For households, the honest conclusion is narrower than either campaign slogan. Amazon has identified a plausible risk that moving facilities outward would make deliveries slower and more expensive. It has not demonstrated that every household will necessarily pay $664 a year. City Hall has documented legitimate concerns about injuries, crashes and divided responsibility. It has not yet shown precisely what the transition will cost or which provision will deliver each promised safety improvement.
The fight is therefore not simply Amazon against Mamdani, or convenience against workers. It is a negotiation over who carries the financial and legal responsibility for the last mile: the platform that designs the system, the local contractor that employs the driver, the worker who bears the physical risk, or the household waiting at the door.
The answer will be written in the final bill, in the operational choices companies make, and in whether officials insist on evidence after the headlines fade.
Reporting note and interview provenance
This article was produced from a review of the introduced bill, the City Council docket, the mayor’s public statement, Amazon’s submitted Council testimony, the comptroller’s crash and injury analysis, and state lobbying disclosures. The monthly and weekly equivalents of Amazon’s annual estimate are original calculations for this article.
Sources
- New York City Council, Intro 0518-2026 docket and bill text
- New York City Mayor’s Office, Mamdani endorsement, August 10, 2026
- Amazon testimony submitted to the City Council
- New York City Comptroller, last-mile crash and injury report
- New York State Commission on Ethics and Lobbying in Government, source-of-funding filing
- New York Focus, campaign funding investigation
- Fox Business, Amazon’s September 2 cost warning