Transit agencies may contract for a portion or all of their operations needs. Smaller, newer transit agencies without historical relationships with unionized labor are more likely to contract all of their service. Larger, older agencies with standing relationships and long histories with unionized labor typically contract only a portion of their labor, if any at all. Reducing the number of union contracts would be politically difficult. <ref>Iseki, Hiroyuki, Amy Ford, and Rachel J. Factor. [http://trb.metapress.com/content/1k55102377427762/ “Contracting Practice in Fixed-Route Transit Service: Case Studies in California.”] 2006.</ref> When contracting out just a portion of services, an agency can choose to contract for labor, [[advertising]], policing, technology services, and maintenance of vehicles, as well as operation of entire lines.
The prevalence of contracting for public transit services has grown since the 1980s with many examples of its success in reducing costs, especially in providing [[cost-effective ADA service]]. About 18 percent of all vehicle hours, including both fixed-route and demand-responsive services, are provided through contracted services.<ref>Kim, Songju and Martin Wachs.[[media:Access_TransitContracts.pdf|“Transit and Contracts: What’s Best for Drivers?”]] 2006.</ref> Demand-responsive services are much more likely than standard bus services to be contracted - at about 66 percent and 6 percent, respectively.<ref>Transportation Research Board. [[media:TRB_ContractingReport.pdf|”Special Report 258: Contracting for Bus and Demand-Responsive Transit Services. A Survey of U.S. Practice and Experience.”]] 2001.</ref>
Of the savings achieved through privatization of transit operations, the vast majority appear to come at the expense of labor rather than from an increase in productivity. Contracted workers earn 38% less per hour and 34% less per year than their publicly employed counterparts. Benefits were the most severely differentiated form of compensation with contracted workers receiving 58% less in benefits than public employees. <ref name="Kim 2005">Kim, Songju (2005), “The Effects of Fixed-Route Transit Service Contracting on Labour”, University of California, Berkeley, unpublished dissertation.</ref> These reductions in benefits might also be considered one of the 'ripple effects' that accompany privatization,
which tends to mean lax labor rules and a better negotiating position for management. <ref>McCullough III, William Shelton, Brian D. Taylor, and Martin Wachs. [[media:TransitServiceContracting.pdf|”Transit Service Contracting and Cost-Efficiency.”]] 1998.</ref>
One study of data from the [http://www.ntdprogram.gov/ntdprogram/ National Transit Database] found a few important differences between private and public providers, in terms of driver compensation. First, privatized systems pay drivers less, and offer fewer benefits than public agencies. By offering reduced benefits and wages, private transit operators can attain higher labor efficiency - they can offer the same level of service at a lower cost. However, contracted transit workers work more overtime than publicly employed transit workers, which can undercut some of the labor savings. Next, private contractors have higher insurance and training costs in part because they have a higher rate of driver turnover.<ref>Kim, Songju and Martin Wachs.[[media:Access_TransitContracts.pdf|“Transit and Contracts: What’s Best for Drivers?”]] 2006.</ref> [[File:Metro210crash.jpeg|right|thumb|350px|A Los Angeles Metro bus awaits repairs after a crash. Source: http://www.fta.dot.gov/about/region9_5806.html]]
Comparably higher turnover rates contribute to a larger proportion of inexperienced drivers and contracted labor has been associated with a decline in service quality. Contracted transit service has been found to have up to 70 percent more collisions and 34 percent more mechanical breakdowns than through comparable publicly provided service.
<ref>Nicosia, Nancy (2002), “Essays on Competitive Contracting: An Application to the Mass Transit Industry”, University of California, Berkeley, unpublished dissertation.</ref>