With unprecedented backlogs of almost 1.5 million pending applications and more submitted each day, the United States Citizenship and Immigration Services (USCIS) faces nearly a year turnaround time to issue immigrants’ work permits. Most of these permits are valid for two years, with some automatically extended for an additional 180 days. However, the ongoing bureaucratic backlog is leaving many immigrants in limbo as they await renewal decisions.
Plagued by office closures and suspension of services at the start of the pandemic, and now faced with significant labor shortages, the public sector’s ability to operate has been greatly impacted. And yet, the question still remains—how did we get here?
The Costs of Backlogs
The reality of expired work permits is costly. While some companies are positioned to place immigrants on unpaid leave until the permit is processed, others are not, resulting in indefinite layoffs, with no promise they’ll be welcomed back. No job and no pay occasionally find some individuals needing to continue to work on an expired permit illegally, despite the risk of deportation or fines.
Let that sink in.
Rent, utilities, groceries, and diapers are among the list of basic necessities on the line for individuals out of work and are a direct result of these backlogs and inefficient processes. Individuals’ livelihoods are being impacted, as are their physical, financial and mental health. And it isn’t just the work permit backlogs that are suffering. The struggles in the USCIS extend to green card applicants too, which ended the 2021 fiscal year with more than 8 million pending cases.
The uncertainty of when these individuals will receive their renewals or verdicts is daunting as immigrants, asylum seekers, and others are left waiting, as backlogs continue to mount.
Where Technology can Play its Hand
The public sector is no stranger to unprecedented backlogs. Pre-pandemic, renewing your passport took roughly 6-8 weeks. But the U.S. Passport Services experienced four-month wait times last summer and continue to be above pre-pandemic norms, currently at 8-11 weeks. What’s more, the Internal Revenue Services (IRS) is navigating 20 times the usual amount of tax filing backlogs. Public servants are inundated with a firehose of data and manual paperwork.
While the Temporary Final Rule (TFR) was announced at the start of May by the Biden administration—helping to avoid gaps in employment for noncitizens by extending the window for certain expired employment authorization and renewal permits—the TFR only buys more time.
For government agencies to efficiently and effectively process millions of forms, applications, and documents to meet the needs of their constituents, more must be done to modernize the public sector. Though document processing seems straightforward, today’s documentation is received in various formats, frequently with poor readability (messy handwriting or low resolution) and high variability. This makes it nearly impossible to reliably and efficiently process and extract data for downstream use.
In fact, one of the most significant issues is how agencies attempt to fix the problems. Outdated, legacy technologies can handle data processing to a certain extent, but they cannot reliably identify, read, and transcribe the diverse layouts, text inputs, and degradations that make up real-world documents. Often, agencies patch various legacy systems together to index and transcribe documents, but this is a fragile ill-advised investment, lacking the flexibility to support new use cases or unexpected surges in volume that have become more commonplace. Legacy technology often fails when faced with messy handwriting, crossed-out lines, or text outside a bounding box, requiring manual reviews and corrections that prevent organizations from achieving peak efficiency. What’s needed is a new approach to automation software—like Human-Centered Automation.
Designed to achieve greater accuracy and reduce processing times, human-centered automation creates a symbiotic work culture between humans with machines, fundamentally improving organizational agility. By providing continuous process improvements and ensuring agencies can manage the growing document and data backlogs, public entities will be better positioned to serve a growing number of constituents with varying needs.
Rising expectations and demands placed on organizations highlight the limits of legacy technology and its lack of evolutionary capability. Moving beyond traditional OCR, NLP and current automation solutions is the only way forward. By enabling collaborative human-machine processing capabilities, at scale, with greater accuracy than previously possible, we can create a winning combination for organizations, employees, and the end consumer.
Upending outdated technology and processes is no small feat, but it must be done—individuals’ lives and well-being are on the line.
The author, Tony Lee, is Chief Technology Officer, at Hyperscience.