Amazon has been in the news recently for its new program, Associate Development and Performance Tracker, or “Adapt,” which has dehumanized the most human aspect of a company’s organization: human resources.
Adapt tracks warehouse workers’ productivity and determines how quickly the workers can locate, scan and box packages. Workers who do not meet the required productivity rates will receive automatically generated warnings. After six warnings, Adapt will automatically terminate the worker’s employment, without any human interaction.
Amazon has stated that human supervisors can override the termination; however, this is a reactive solution to the issue of improper automatic terminations, as opposed to a proactive management of the terminations.
Essentially, robots are performing a uniquely human function without any humanity.
This is not to say that Adapt is not a useful and beneficial system for employers. There are many advantages to Adapt and similar technologies. For example, they substantially increase productivity by ensuring that employees are working hard to achieve their targets and they decrease costs by removing underperforming employees. This allows for a more efficient performance management system.
Additionally, the automated termination program eliminates bias by ensuring that no human predispositions or prejudice is involved in employment-related decision making.
However, numerous issues and questions arise when computers take over human functions. There is a real and substantiated concern that employees’ jobs will be replaced by computers. Employee morale can be an issue, as workers may feel indignant that they are being monitored in this way and that their continued employment is in the hands of a computer.
Computers can only measure statistics of productivity; they cannot perform human soft skills to take into account human particularities. Employees should also be given support and further training, if necessary, where performance is an issue. A simple warning letter advising of low productivity rates does help the employee improve his/her performance.
While reducing the risk of bias is a great advantage, a computer cannot recognize the individual factors of an employee’s employment. For example, a computer program cannot recognize whether an issue has arisen as a result of an employee being sick. This could result in a wrongful dismissal claim.
A computer program cannot recognize whether an employee requires accommodation to perform his/her job. A worker who cannot “make rate” and achieve the uniform productivity rate will be automatically fired as opposed to being accommodated.
This exposes the company to a claim that the employee’s human rights have been violated. What if the employee has a physical or mental disability? What if the employee is pregnant? These factors cannot be incorporated into an automatic termination.
An automatic performance management system may also lead to privacy issues and complaints.
Employees have a reasonable expectation of privacy, unless a legitimate business objective is engaged. Monitoring employee performance through Adapt may constitute a breach of privacy. This begs the question, does performance management constitute a legitimate business objective?
Another issue that arises is that the intense pressure to perform in order to hit productivity rates may result in an unsafe workplace, thereby exposing the employer to a violation of occupational health and safety legislation. What if an employee collapses out of exhaustion? What if there is not sufficient attention paid to employees’ safety, as productivity is placed above well-being? What are the working conditions like? Additionally, is employment standards legislation being violated? Are workers being forced to work longer hours than are allowed? Is overtime being appropriately paid? Are averaging agreements, where appropriate, being entered into?
The most significant issue for Canadian employers in implementing an automated termination system is that it likely will not take into account employees’ legal entitlements on termination. Unlike the United States, employees in Canada are not employed at will. As such, they must be dismissed either with just cause or with notice of termination or pay in lieu of such notice.
Would an automatic termination for low productivity rates be sufficient cause to terminate employment for cause? “Cause” in Canada usually involves serious misconduct resulting in material harm to the financial condition of reputation of a company. Low productivity rates, in this author’s opinion, does not amount to just cause.
Even if an employer was confident that low productivity rates did amount to just cause, a court might not agree, which could expose the employer to a claim for punitive damages, mental distress damages and reputational damages on top of wrongful dismissal damages.
If the termination is done without cause, sufficient notice of termination or pay in lieu of such notice must be provided.
This raises numerous additional questions: how much notice is required? Only statutory notice, based on applicable legislation? Does it depend on the employee’s employment contract? How can the computer determine each employee’s contractual rights? Is common law notice included as well? If so, how is that amount determined? Can the computer take into account the various Bardal Factors (Bardal v. Globe&Mail Ltd.  O.J. No. 149) to determine the appropriate notice period? What about severance pay?
It remains to be seen whether an automated performance management system like Adapt can or will be brought into Canada. If it is to be implemented in Canada, employers must ensure that they have the proper policies in place for when and how this type of tracking system will be implemented, as well as the types of recourse available to employees who want to dispute an automatic termination.
There should also be a human component to the process, perhaps involving a supervisor reviewing any warning, discipline or termination letters prior to being sent to employees to allow for a more individualized approach.
Ultimately, while automating termination decisions can increase productivity and efficiency, it raises more questions than it answers. How can a computer take over the innately human functions of human resources? Are we prioritizing efficiency over humanity? This remains to be seen.
Rachel Goldenberg is a content lawyer at LexisNexis Canada.
Photo credit / Andrey Suslov ISTOCKPHOTO.COM
Interested in writing for us? To learn more about how you can add your voice to The Lawyer’s Daily, contact Analysis Editor Peter Carter at firstname.lastname@example.org or call 647-776-6740.
This article was originally published by The Lawyer's Daily -- providing Canadian legal news, analysis and current awareness for lawyers and legal professionals who need a real-time view on the shifting legal landscape.