by Bruce Schaller and Gorman Gilbert

New York City taxicab issues have fascinated the public while bedeviling politicians and policy-makers for decades. The fascination stems from the combination of the ubiquity of yellow cabs in the city's core areas, the many tales of woe experienced by passengers, and the belief of many that the taxi industry represents a monopoly run amok. This three-part series uses recently compiled information to provide a factual basis for examining service issues and possible solutions. This first paper looks at the key to service quality--the drivers--and shows how service problems stem from inadequate wages and difficult working conditions. The next paper evaluates the widespread notion that the medallion system is the root of service problems. It finds that the growth of taxicab leasing over the last 15 years has had a far more pernicious effect on service quality than has the medallion system. The third and final paper appraises nine strategies for improving service, evaluating the efficacy of city regulation and the potential for market-based policies.








Table 1. Driver Income, 1993 Table 2. Prevalence of Part-Time and Inexperienced Drivers, 1993-94 Table 3. Driver Quality and Amount of Driver Experience


New York City taxicabs have been a subject of public debate and controversy, mayoral commissions, and City regulatory reforms since the birth of the motorized taxicab in 1907. This constant attention is due in part to the simple fact that, for a large swath of often-prominent New Yorkers, the yellow taxi is a part of daily life. Manhattan adult residents hail cabs an average of 100 times a year,(1) and some Manhattanites jump into cabs several hundred times a year.(2) As these figures indicate, taxis are a significant component of the transportation network. In 1993 they carried 34% of all fare-paying passengers traveling by bus, subway, taxi, or for-hire vehicle within Manhattan.(3)

While many issues surround cabs and cab service, to the public the most visible, immediate and perhaps most interesting subject is the taxi driver himself, and occasionally herself. Since all medallion taxis are yellow, the public does not perceive differences among taxi owners, and taxi regulators are an abstraction glimpsed only through the media. Yet every taxi trip includes an encounter with a driver.

New Yorkers are of two minds about their city's 40,000 licensed taxicab drivers. Cab drivers are often seen as hardworking immigrants eager to earn honest livings in their adopted country. But they are also portrayed as foreigners befuddled with the English language, unable to find the Empire State Building, and prone to overcharge, refuse, and abuse passengers. This perception is often reinforced by vivid press accounts of drivers' misdeeds.

These problems highlight an important public policy issue: what can the City do to reduce the incidence of unhappy passenger experiences in taxicabs? This paper explains how the quality of taxi service is shaped by drivers' personal and employment backgrounds, their motivations for joining the taxi industry, and their experiences as drivers. Within this analysis, several stories of larger interest are told: how immigrants do a job few native-born Americans find attractive; how the job shapes the workers and the service they provide; and why economic forces must be the focus of regulators seeking to enhance the quality of taxi service.


For all the differences among drivers and the complexities in the ownership and operation of New York City's 11,787 medallion taxicabs, the job of driving a cab is essentially the same for each driver: cruise the streets and pick up fares. Most passengers are picked up by cruising taxicabs; the remainder are served at taxi stands established at highly trafficked areas. (Medallion cabs have not served telephone prearrangement since two-way radios were removed from them in the mid-1980s.(4) Passengers are primarily Manhattan residents going to and from their homes, workplaces, and such recreational pursuits such as dining, entertainment, and shopping. Out-of-town businesspeople and tourists are also a significant segment of the ridership.(5)

Taxi drivers typically spend most of their work shifts in the city's central core, generally defined as Manhattan from the Battery to 96th Street. This area includes the downtown and midtown business districts, the Theatre District, major shopping areas, most major Manhattan museums, and fashionable residential areas. Nineteen of twenty taxi trips originate within this area, according to a 1988 sample of taxi trip sheets. Since Manhattan trips are short (about 1.9 miles) and quick (averaging 10 minutes), rapid passenger turnover is the key to a profitable day. Drivers average 30 fares in an average shift of 8 to 12 hours, an unusually high number for any taxi industry. A successful driver has learned where best to troll for passengers and how to get to the next destination most quickly.


Who wants to be a New York medallion cab driver? The short answer is: immigrants looking to better their incomes. In a 1991 survey of 2,500 new taxi drivers, 42% wrote "making money" (or similar words) when asked why they wanted to drive a cab.(6) An additional 15% cited their need for work, typically commenting that there is "not much work in my field" or "I was laid off." Only one in three new drivers named nonfinancial attractions of the job, such as liking to drive (16%), liking the independence, flexibility or excitement of driving (9%), or wanting to serve people (3%).

One of the most publicized traits of New York cabbies is their ethnic diversity. Cab drivers form a virtual United Nations of countries and languages. Nine in 10 new drivers in the 1991 survey were immigrants from a total of 84 countries. Over 4 in 10 were born on the Indian subcontinent (Pakistan, Bangladesh and India). Africa, the Caribbean, the Middle East and republics of the former USSR were other common birthplaces.

Contrary to popular belief, few drivers are newly arrived in the United States. Only 13% of the new drivers in 1991 were in the US for less than one year while one out of two had been in this country for at least six years.

Drivers' levels of education are impressive. In the 1991 survey, 59% of new drivers said they had attended at least some college, and 37% were college graduates. Their most recent occupations, however, did not match their educational achievements. Seven in 10 had most recently worked in a low-skilled job such as livery cab driver, delivery driver, waiter, cook, busboy, baker, store cashier, clerk, or gas station attendant. Conversely, only 9.5% of the new drivers were coming from a skilled or professional job such as carpenter, welder, mechanic, architect, teacher or social worker.

Driving a cab represented a financial step forward for most new drivers surveyed in 1991. Sixty percent expected to earn more driving a cab than they had earned in their previous jobs; only 11% expected to earn less. The majority knew about their new occupation from direct experience and/or friends and relatives. Forty-four percent had driven passengers for-hire previously, primarily livery cars in New York or taxis in their native countries. Seventy-three percent already knew someone who drove a New York cab.

As they prepared to get behind the wheel, most new drivers showed substantial dedication to taxi driving. Three in four expected to drive full-time, and one-half planned to drive for "many years." Three in five hoped to buy a taxicab medallion license eventually. Few new drivers had plans beyond cab driving.

In keeping with their diverse ethnic backgrounds, new drivers speak a plethora of languages. New drivers surveyed in 1991 spoke a total of 60 foreign languages, most commonly Urdu, Punjabi, Arabic, Bengali, and Russian, but also German, Japanese, Twi, Ibibio and Estonian. The median time since learning English was seven years, which one might expect would constitute an adequate amount of time to learn the language. A further look helps to explain why complaints about drivers' English skills are widespread, however. Twenty-eight percent of the new drivers said they had learned English within only the last three years. Furthermore, most new drivers live in immigrant enclaves in New York, and many of their neighbors are from their native countries. Notably, fewer than half (44%) of the new drivers reported speaking English at home in the 1991 survey. With this background, 27% of new immigrant drivers failed their first of two possible tries on a very simple English exam used in the licensing process in 1991. Currently, 32% fail a somewhat more stringent test used now.


Prospective drivers gathered in the week-long driver training course are "eager and intent," in one reporter's words.(7) What happens to these hopeful new drivers? Are their expectations and career plans fulfilled? Why or why not, and with what implications for taxi service?

4.1 Driver Incomes

To answer these questions we first examine driver pay. As discussed above, drivers become drivers for one reason above all others: to make money. How much do drivers earn, do they feel adequately compensated, and with what consequences?

The earnings equation has two closely related parts: how driver earnings are derived and the amounts they earn. The method of compensation is significant because, unlike most of the American labor force, taxi drivers are not paid a fixed salary or hourly wage. Most taxi drivers are "independent contractors" who lease their cabs by the shift or for months at a time. Drivers pay a flat amount for each shift or each week and have exclusive use of the cab for the shift. Driver earnings are the difference between their total revenues (from fares and tips) and their expenses (lease fees and gasoline).

Drivers loathe the leasing system, for two main reasons. First, leasing shifts to drivers all the financial risks created by weather, traffic conditions, variations in taxi demand and in the type of trips garnered in a day. Small variations in demand can make a big difference in the day's profitability. For example, a change of $5.25 in fare revenue (an average trip) changes a fleet driver's take-home income by about 7.5%.(8) Traffic also has a large impact. For example, the fare structure adds $1 to the meter when a cab is stuck in traffic for five minutes compared with $3.10 when a cab spends the same amount of time traveling 30 miles an hour. The acute sensitivity of driver earnings to the number and type of trips creates a pressure and strain that drivers feel all day, regardless of the final revenue tally.

A second problem with leasing stems from the fact that drivers are compensated entirely in cash. While cash may be better than a paycheck for drivers avoiding the IRS, shortcomings in this system seem to outweigh this "benefit." Except for workers compensation, which New York State law mandates, drivers get none of the employee fringe benefits and social safety net protections common to other workers: no Social Security coverage; no unemployment insurance; no disability insurance for sickness or off-hour injury; no health insurance; no pension; and no paid vacations. The lack of these benefits creates a feeling of insecurity and transience among many drivers.

If the method of compensation has serious drawbacks, so does the dollar amount. Table 1 shows average driver take-home incomes under the four most common lease arrangements. These are:

Table 1. DRIVER INCOME, 1993

                            Shift Leasing    Long-term Leasing
                            Fleet    Other  Medallion Medallion
                                               & Car      Only
                             (per shift)         (per week)
Total fare & tip income    $182.81  $173.15   $935.22   $935.22
Lease fees                  82.69     82.22    475.00    310.00
Gasoline                    16.20     16.20    100.76    100.76
Vehicle costs                   0         0         0    132.77
Net per driver per week       n/a       n/a  $ 359.46  $ 391.69
Net per driver per shift $  83.92  $  74.73  $  65.36  $  71.22

Shifts per year               264       264       275       275
Income per year           $22,155   $19,729   $17,974   $19,586

Source: Taxi and Limousine Commission, "Should the Taxi Fare 
Go Up?" March 31, 1994.

Average fare revenue for the four groups ranged from $144 to $154 a shift in 1993. Tips can be conservatively estimated at 15% (good drivers claim 20% or more in tips), bringing total revenue to $170 to $183 a shift. After paying for gasoline and their lease fees, shift lessees take home between $75 and $84 a shift. Long-term lessees take home $359 to $392 a week working an average of 5.5 days a week, yielding $65 to $71 a shift. Annual earnings vary widely because drivers work a widely divergent number of shifts per year. Taking the average number of shifts worked by drivers working "full- time" (i.e., at least 200 shifts a year), drivers take home $18,000 to $22,000 a year (see Table 1).

There is a notably large variation around the averages shown in Table 1. Lease drivers at the 25th percentile of fare revenue (one-quarter of drivers grossed more in fares; three-quarters grossed less) brought in an estimated 29% more in fares than drivers at the 75th percentile.(9) This difference in gross income translates into a 59% difference in take-home incomes after subtracting lease fees, which are fixed, and gasoline costs.

Do these variations hold any promise for the aspirations of individual drivers? We can answer this question in two ways -- for drivers individually and for drivers collectively. First, do income variations represent any type of income growth potential for individual lease drivers? The answer is an emphatic "no." In fact, driver income declines as experience levels rise. Cross-sectional 1990 trip sheet data show that lease drivers with eight or more years of experience had average fare revenue 10% less than those in their second to eighth years as taxi drivers.

Why is this? The key is the impact of longevity on work effort; drivers with greater experience have lower incomes primarily because they work fewer hours each day. Given the difficulty of maintaining a regimen of 10 or 11 hour work days, drivers appear to persevere only by reducing their driving hours.(10) Lower incomes among more-experienced drivers may also be a product of self-selection. The most able and talented drivers, or those willing to work the longest hours, may leave the taxi industry for other job opportunities, leaving a greater proportion of lower-grossing drivers who reduce the averages.

For drivers collectively, incomes are stagnant as well. Despite taxi fare increases in 1987 and 1990 totaling 37%, take-home incomes for fleet drivers were essentially unchanged in 1993 compared with 1986 after adjusting for inflation ($89 per shift in 1986 and $84 in 1993, using 1993 dollars). Rising lease fees and inflation offset the benefits of the higher fare.(11)

There is one possible exception to this dismal picture: owner-drivers. Medallion ownership lets owner-drivers realize the benefits of owning as well as of driving. Once an owner-driver pays off the loan used to finance a medallion purchase, he can take home about $36,000 a year.(12)

The importance of this bounty is shown by driver retention rates. Most drivers who do not buy a medallion leave the industry. As a result, one-half of all full-time drivers with over six years of experience are owner-drivers. The comparable figures are 25% for full-time drivers with 3 to 6 years of experience and under 10% for drivers with fewer than three years of experience.(13)

Unfortunately, the income attainable by veteran owner-drivers is a distant and receding hope for most drivers. The fact is, few drivers ever buy a medallion license. TLC licensed between 4,300 and 6,500 new drivers in each of the last eight years but only 234 drivers bought owner-driver medallion licenses in 1994. In other words, only one new driver in 20 eventually becomes an owner-driver. This ratio is not surprising given the financial resources needed to buy a medallion. Medallion buyers must make a substantial upfront outlay, typically 20% of the medallion price (or $33,000 at the current price of $165,000 for an owner-driver license).

With rising medallion prices, fewer drivers manage to buy. The number of drivers buying individual medallion licenses has dropped by over 60% since the mid-1980s, when 600 to 800 licenses were transferred each year. Those who do buy a medallion license now wait longer to do so. In 1994, the median amount of time new owner-drivers had spent driving a cab before buying was 5.3 years compared with two years in the mid-1980s.

For those who wait and save, medallion ownership has become a slower route to a decent income. In the mid-1980s buyers typically took out five-year loans. After those five years, the loan would be paid off and the driver's take-home income would rise precipitously. Loans of 15 years are now the norm. Until they pay off the loan, new owner-drivers are financially worse off than lessees by at least one measure. As of the end of 1993 a new owner-driver took home about 15% less per mile driven than did lease drivers.

4.2 Working Conditions

Compensation is the number one concern of drivers, named as the most-desired job improvement by 39% of drivers in a 1994 survey. What are drivers' other major concerns, and what are the prospects for betterment in these areas?

Next on the list is crime, named by 28% in the 1994 survey. Driving a taxi or livery is the most dangerous job in New York City.(14) While TLC mandated protective partitions for nearly all cabs in 1994 and TLC and the Police Department have taken other steps to improve driver safety, safety has been and will continue to be a major concern for drivers in this vulnerable, all-cash business.

Trailing on drivers' list of concerns are improved taxicab vehicle conditions, better treatment from TLC inspectors, better traffic and street conditions, and improving the quality of their fellow drivers. Each of these was named by about 10% of drivers in the 1994 survey. Unfortunately, some of these problems have worsened in recent years. Cars have become older and more deteriorated. Though traffic may have eased with the recession, a day on Manhattan streets is still a grinding experience, and the pavement has taken a beating from icy winters and budget cuts. On the other hand, Mayor Rudolph W. Giuliani's transfer of taxi enforcement duties from TLC inspectors to the police was welcomed by many drivers, who expected fairer treatment from the police. Whether drivers will prefer police oversight in practice remains to be seen.

In sum, driver compensation and working conditions show few immediate prospects for improvement. As an articulate American-born driver summed up the situation, "who else but an immigrant from an impoverished land would be willing to tough it out in this miserable hustle?"(15)

4.3 Transience

After traveling thousands of miles in search of a better life in the United States, immigrant cab drivers clearly hope to fulfill their dreams working as taxi drivers. That many drivers are disappointed seems clear from their next moves. Contrary to drivers' initial plans, they leave the industry in droves. While 77% of new drivers surveyed in 1991 planned to drive a cab full-time, several other driver surveys have found that fewer than one in five drivers actually work full-time (200 shifts or more) in their first year. While 49% of the new drivers in 1991 planned to drive "for many years," only 11% drove full-time in their third year while 62% had either left the industry or drove few or no days at all.

Differences between drivers who stay and those who leave the industry support the notion that drivers tend to leave if their skills provide an alternative. Drivers who in 1994 renewed their licenses for a third year were disproportionately likely to be immigrants who had held lower-skilled jobs and who scored poorly on the TLC English exam. Who was most likely to have moved on? US-born drivers, drivers with perfect scores on the English exam, and drivers previously holding professional jobs.


The wages and working conditions of drivers and rapid driver turnover affect the riding public in many ways. Some are direct and obvious. Anyone familiar with New York cabs has probably noticed how drivers bound out from red lights, speed down the street, weave through traffic, and cut across lanes to stop for hailing passengers. This driving style is popular because one trip can make the difference between a good and mediocre take for a driver's shift. But aggressive driving is also a major problem, at least when measured by public complaints about taxi drivers. With 1,600 complaints a year, traffic violations, often meriting the charge of reckless driving, head the list of formal complaints filed with TLC by passengers and members of the public.

A second obvious and financially-driven problem is overcharging. The TLC receives about 1,000 overcharge complaints a year and overcharging is clearly more widespread than the complaint data indicate. According to results from TLC's random meter tests of 226 taxicabs operating in Manhattan in 1994, 40% of all taximeters are set at rates above the approved fare. Fast meters overstated the distance actually traveled by the cab by an average of 12%. If anything, these figures understate the amount of overcharging. Some taximeters are rigged so that drivers can add to the metered fare when no one is looking. Overcharges also occur off the meter, such as when a driver demands an exorbitant flat fare (often for airport trips) or additional money for taking more than one passenger.

The causes of a third common complaint, service refusals, are more complex and disputed than the causes of reckless driving and overcharging. In the press, service refusals are often attributed to drivers' alleged racial prejudice and/or desire to avoid becoming a crime victim. This understanding of refusals suggests that policy-makers face a Hobson's choice, that demanding that drivers serve all customers equally forces drivers to risk their lives. Available data paint a much different picture, however. A 1993 survey of 464 complaining riders found that most complainants were white and that the majority were destined for Manhattan locations or low-crime areas of Brooklyn and Queens. Race and fear of crime do not explain refusals of white passengers traveling crosstown.

Why then do drivers refuse? Passenger complainants in the 1993 survey attributed refusals primarily to economics, traffic and geography. The underlying motivation in each case is financial: drivers do not want to dead-head back from an outer-borough destination, they want to avoid rush hour bridge or tunnel traffic, they believe that a trip uptown will be more profitable than a passenger's desired trip downtown. While white passengers cited financial motivations more often than did nonwhites, even nonwhite complainants were more likely to attribute refusals to nonracial motivations than to racial bias.

It should be noted that conclusions based on formal passenger complaints may be misleading since many instances of service refusal involve a taxi driver refusing even to stop for a hailing passenger. These incidents rarely result in a passenger complaint, primarily because it can be impossible to prove that the driver saw the passenger and, in some cases, the driver and/or cab can not be identified. Such passbys may be motivated by racial bias, by fear of the passenger or by fear of the area the passenger may be headed. On the other hand, drivers may be using race as a proxy in their attempts to avoid economically unattractive trips. The latter hypothesis is advocated by one knowledgeable driver.(16)

By creating a strong motivation for drivers to violate traffic laws, drive recklessly, refuse service and overcharge riders, low driver earnings and poor working conditions directly erode the quality of taxi service. Service is affected indirectly as well, the product of many drivers' weak attachment to their jobs. As we have seen, driver pay and fringe benefits are insufficient to keep most drivers in the industry for very long, and many drive as a sideline while they seek better opportunities. The result is that first, second and third-year drivers provide one-quarter of all taxi service, and other part-time drivers provide another one-quarter of all service. (See Table 2.) The swollen ranks of new and part-time drivers work to the detriment of taxi service quality because new and part-time drivers are more likely to leave passengers unhappy.


                                  Pct. of
                               shifts worked
Part-time drivers  1-3 years         11%
                   4-6 years         10
                   7+ years          14
Full-time drivers  1-3 years         15
                   4-6 years         13
                   7+ years          37
Total                               100%

Source: Survey of 2,215 taxicab drivers who renewed their taxi 
driver licenses expiring in July and September 1994. 

This impact is quantified in Table 3, which compares the rate at which different groups of drivers receive summonses for TLC rule violations. Summons records are the best available statistical measure of driver quality. Summonses incorporate reports of service problems from a range of sources: taxi passengers, TLC inspectors, the police, and airport dispatchers. Summonses include the most serious passenger complaints for reckless driving, refusals, discourteous and abusive behavior, and some overcharging complaints.(17) Table 3 reflects data for all summonses, but the same patterns emerge when one looks only at passenger complaints or only at the most serious violations.

As shown in Table 3, full-time drivers are summonsed at a lower rate than part-time drivers, even when the most-experienced part-time drivers are compared with the least-experienced full-time drivers. Among part-time drivers, those with more experience are summonsed at a lower rate than are those with less experience; the same is true for full-time drivers. The biggest gap in summons rates -- a difference of 280% -- is between relatively new, part-time drivers and the most experienced, full-time drivers (7.9 summonses per 1,000 shifts for part-time drivers in their first three years versus 2.1 summonses per 1,000 shifts for full-time drivers licensed for over six years). The clear implication from these data is that a more experienced, more dedicated driver work force would better serve the public.


                               Summonses per
                               100,000 shifts
Part-time drivers  1-3 years         7.9
                   4-6 years         4.9
                   7+ years          4.4
Full-time drivers  1-3 years         3.3
                   4-6 years         3.0
                   7+ years          2.1

Data are for drivers surveyed in July and September 1994 and reporting 
a number of shifts worked in previous year; summonses for period 
between June 1993 and July 1994.


The key to improving the quality of taxi drivers is to increase the number of proficient career drivers through better financial rewards, better working conditions and greater security from crime. By both their words and actions, drivers make it clear that a higher and steadier wage and better employee fringe benefits would make taxi driving decisively more attractive. Since the level of drivers' commitment to their job is critical to the quality of service, the task of improving driver compensation and working conditions should take center stage in efforts to upgrade taxi service.


1 Bruce Schaller, "New York City Taxicab Fact Book, 3rd Edition." NYC Taxi and Limousine Commission, May 1994, page 1.

2 A 1994 survey conducted for the New York Metropolitan Transportation Authority found that 15% of Manhattan adults took at least 5 taxi trips in the last week, which would project to 250 or more taxi trips annually.

3 "Taxicab Fact Book", page 2.

4 About 3,000 cabs had two-way radios before the City mandated their removal. The City acted in order to maximize the amount of street hail service provided by the limited number of taxicabs. Another industry known as "for-hire vehicles", which includes corporate-account black cars and neighborhood car services, responds to telephone reservations.

5 Bruce Schaller, "Hailing From New York." NYC Taxi and Limousine Commission, November 1993, pages 11-13.

6 Data in this section are from survey reported in Bruce Schaller, "Who's Driving New York?" NYC Taxi and Limousine Commission, April 1992.

7 Emily Sachar, "A Driver's Course," New York Newsday, November 16, 1993.

8 This is based on an average fare revenue of $154 per day in 1993 and net of $82 per day. See Table 1 later in the paper for additional details on driver income levels.

9 This datum is based on 773 trip sheets of lease drivers who worked 8 to 12 hour shifts.

10 Because only cross-sectional data are available for lessees, we can not say that gross incomes decline over time for individual drivers. Time-series data for owner-drivers, however, shows that owner-drivers reduce their working hours as they age. Taximeter readings collected at 4-month intervals since Fall 1989 show annual gross revenues for owner-drivers active since 1990 dropped by 14.3% over a 3.4 year period.

11 These data are for fleet lessees only. Few other shift lessees existed in the mid-1980s and thus no time-series comparisons can be made. Due to changes in the provisions of long-term leases, incomes for long-term lessees cannot be compared directly for 1986-93 period. Stagnation of driver incomes is discussed at greater length in the next paper.

12 Annual income of $36,000 for owner-drivers is based on their grossing an average of $169 in fares per shift for each of 276 shifts a year. The $169 figure is higher than for lessees. Owner-drivers, whose cabs are generally single-shifted, typically work longer days and bring in more fare revenues per day than do double-shifting lessees.

13 This analysis is based on an unpublished survey of 2,215 taxicab drivers who renewed their taxi driver licenses expiring in July and September, 1994.

14 There were 36 homicides of taxi drivers and chauffeurs in 1992, compared with 24 homicides of grocery clerks. New York City Health Department, "Occupational Fatalities in New York City during 1992," undated, page 1. [Note, however, that there were no homicides of medallion cab drivers in the two years after TLC mandated installation of protective partitions in taxicabs in 1994.]

15 William Mersey, "Wanted: a New Deal for Cabbies", New York Times, June 19, 1994.

16 William Mersey, "Why Cabbies Pass Up Blacks," New York Times, January 1, 1995.

17 Overcharge complaints where the passenger seeks a refund for the overcharge but declines to attend a hearing are not reflected in summons data.

Published in TRANSPORTATION QUARTERLY, Vol. 49, No. 4, Fall 1995.
Copyright 1995 Eno Transportation Foundation, Inc.

The second and third papers in this series:

Home | Press | Publications | Taxi Research