Popular search terms:
Published:   |   Last Updated: February 8, 2024

Getting Rid of the Kryptonite: The IRS Should Quickly Implement Scanning Technology to Process Paper Tax Returns

NTA Blog: logo

Since the beginning of the COVID-19 pandemic, the IRS has fallen nearly a year behind in processing paper tax returns. As of March 18, 2022, the paper return backlog stood at nearly 15 million. Most taxpayers receive refunds, so return processing delays generally mean refund delays. Refund delays have caused frustration for many and financial hardship for some, potentially including evictions, utility shutoffs, and the inability to afford food and medicines.

In other cases, a taxpayer’s inability to furnish proof of current filing and a tax transcript due to processing delays has adversely affected loan applications, including applications for mortgages, personal or business loans, and even student financial aid.

When I released my annual report in January, I said that paper is the IRS’s Kryptonite and the IRS is buried in it. The reason paper returns are so challenging is that the IRS still has not implemented technology to machine read them, so each digit on every paper return must be manually keystroked into IRS systems by an employee.

It doesn’t have to be that way. During the past two decades, state tax agencies have been using scanning technology to automate the processing of paper tax returns. During that time, the IRS has considered, rejected, proposed, reconsidered, partially implemented, and deferred the question of whether to implement scanning technology.

Yesterday, I issued a Taxpayer Advocate Directive (TAD) directing the IRS to work with the tax software industry to implement 2-D barcoding for next filing season. The TAD also directed the IRS to implement optical character recognition (OCR) or similar technology for next filing season if possible or, if not, for the following filing season. In this blog, I will provide an overview of the problem and what I think the IRS should do to solve it. For additional details, you can access the TAD here.

The Problem: IRS Employees Must Keystroke Each Digit From Paper Tax Returns Into IRS Systems

Although the COVID-19 pandemic has caused delays for some taxpayers who e-filed their returns, the overwhelming majority of lengthy delays have been experienced by taxpayers who filed original returns on paper or who have filed amended returns, which are generally processed as paper returns even when submitted electronically. Last year, the IRS received nearly 17 million paper Forms 1040, over 4 million Forms 1040-X, and millions of paper business returns.

The delays in processing these returns result from the IRS’s archaic data intake process. The IRS’s submission processing function today evokes images of what data transcription looked like in the 1960s – prior to the information age. Employees manually transcribe all paper tax returns. Transcription consists of keystroking each digit and each letter on the return. For a moderately complex return, several hundred digits may need to be transcribed. For longer returns with more forms and schedules, the number of digits may approach or exceed 1,000 digits.

In the year 2022, this doesn’t just seem crazy. It is crazy.

The Solution: Implement Scanning Technology to Machine Read Paper Returns

Fortunately, there are two types of scanning technology that would allow the IRS to “machine read” paper returns and reduce the need for manual data entry. Scanning technology would speed return processing, substantially reduce or eliminate transcription errors, and enable the IRS to reassign employees from data entry jobs to other positions, ultimately saving tens of millions of dollars in labor costs. There are currently two leading types of scanning technology: (i) 2-D barcoding and (ii) OCR.

2-D Barcoding and 20 Years of IRS Indecision

When a customer buys a product at a grocery store or a pharmacy, the product is typically marked with a barcode that can be scanned at the checkout line. Similarly, when a taxpayer prepares a tax return using tax return software, software companies generally can place a 2-D barcode on the return that encodes the return data in a machine-readable form. The IRS could then scan the barcode – like the supermarket or pharmacy does – and convert the data into a digital form that would allow the IRS to process the return like an e-filed return.

2-D barcoding technology is well established. In 2002 – fully two decades ago – we reported that 17 states were using 2-D barcoding for returns prepared with tax return software but filed on paper, and we recommended the IRS consider doing so as well. At the time, the IRS disagreed with our recommendation to incorporate 2-D barcodes onto Forms 1040, stating that doing so would undermine the goal of transitioning taxpayers to e-filing. In 2003, however, the IRS had begun working with tax return software developers on a 2-D project for other tax forms, and the IRS has implemented 2‑D barcoding for certain forms, including Schedules K-1.

In 2014, the IRS reversed its position regarding 2-D barcoding for Forms 1040, requesting that Congress provide it with authority to require taxpayers who prepare their returns with software but file them on paper to print their returns with a scannable 2-D barcode. In 2018, the House-passed version of the Taxpayer First Act contained a provision to require 2-D barcoding. By that time, the IRS had changed its position again. We have been told the IRS requested the provision be removed from the legislation to give it flexibility to adopt alternative scanning technologies like OCR. Congress removed the provision from the final version of the Taxpayer First Act that was enacted, but the IRS still has not implemented any type of scanning technology for Forms 1040.

Thus, 20 years after more than one-third of states were already using 2-D barcoding, 18 years after the National Taxpayer Advocate initially recommended it for Forms 1040, 8 years after the Treasury Department requested that Congress provide the IRS with the authority to mandate 2-D barcoding, and 4 years after Congress sought to grant that authority and the IRS changed its position, the tax system finds itself in a crisis that might not exist, at least to this degree, if 2‑D barcoding or similar technology had been implemented.

Based on TAS’s analysis, 50%-60% of the individual income tax returns submitted on paper and processed over the last two years were prepared with tax return software and would not need to be transcribed if 2-D barcodes were added. While the IRS has announced plans to work through the backlog before the end of 2022, it is not clear it will be able to do so. Implementing 2-D barcoding for paper returns filed beginning in January 2023 will serve as an insurance policy against a continuing backlog next year by reducing the influx of new paper returns that require transcription. Even if the IRS does manage to work through its backlog this year, 2-D barcoding will reduce processing time and costs in future years, which will allow the IRS to reassign submission processing employees to perform other work.

Optical Character Recognition Technology

OCR is a newer technology with different advantages and disadvantages compared with 2-D barcoding. The main advantage is that it can be used to machine read all paper returns, including returns prepared by hand. The main disadvantage is that it is less accurate. For example, a “1” and a “7” may look similar, so OCR may read the digit incorrectly. However, OCR technology should still be more accurate than manual data transcription because an employee not only will have the same difficulty distinguishing between the “1” and the “7” but also will hit the wrong key by mistake from time to time. Last year, IRS employees made transcription errors on about 22 percent of paper returns.

In discussions with state tax agencies, we have been told that some states use both 2-D barcoding and OCR. Where a return is prepared with software and a barcode can be applied, the barcode provides near 100 percent accuracy. Where a return is not prepared with software or if a barcode cannot be read (e.g., where a taxpayer’s toner cartridge ran low or the barcode has smudged), OCR is used, and despite its slightly lower accuracy level, it still reduces the need for manual data transcription and eliminates errors attributable solely to human error in hitting the wrong key.


The IRS should be a leader in tax administration information technology practices, but due to a combination of funding limitations and management challenges over many years, it is not. The archaic keystroking of paper tax return data is a case study. Twenty years ago, at least 17 states had automated that task, yet the IRS still has not managed to do so.

If necessity can be the mother of invention, the unprecedented paper returns backlog resulting from the COVID-19 pandemic should spur the IRS to take immediate steps to automate the processing of paper tax returns – not in two years or three years but beginning with the upcoming filing season. The timely payment of tax refunds for millions of taxpayers is depending on it.

Read the past NTA Blogs

The views expressed in this blog are solely those of the National Taxpayer Advocate. The National Taxpayer Advocate presents an independent taxpayer perspective that does not necessarily reflect the position of the IRS, the Treasury Department, or the Office of Management and Budget.

Subscribe to the NTA Blog