In a new op-ed in The Hill, Lincoln Network’s Garrett Johnson and R Street Institute’s Tony Mills explain why taxpayers should support strengthening the Government Accountability Office—pointing to the auditors’ estimated return-on-investment of $338 for each tax dollar spent in FY2019.
Since 2010, GAO has delivered hundreds of billions in taxpayer savings, including by identifying duplication and fragmentation across federal programs. In March, GAO projected that its work rooting out duplication alone has already resulted in $262 billion in savings with more than $46 billion in additional savings expected.
GAO is delivering these results with a workforce that is considerably smaller than it was 30 years ago. According to the Brookings Institution, GAO’s workforce shrunk from more than 5,000 in 1990 to about 3,000 in 2015.
Looking ahead to the 2020s, Congress’s watchdog could deliver even bigger savings and government improvements by using data analytics to enhance federal oversight.
As I wrote here last month, GAO recently launched an Innovation Lab and hired a chief data scientist, which has the potential to modernize the federal oversight process. After writing that post, I heard some skepticism from fiscal conservatives who doubt that increasing spending on another government agency like GAO is worth the investment.
So I thought it might be helpful to provide a few specific examples of how applying data analytics to federal oversight could be game-changing based on my experience as a former Congressional staffer.
Here are five areas where the effective application of data analytics could address major national problems:
1. Stopping Improper Payments by the Federal Government
Most credit card customers have received a call or text from the bank asking to verify a transaction as it happens. If the government used effective data analytics to curb improper payments, the U.S. Treasury could save more than $100 billion per year. GAO’s auditors estimate that the federal government has wasted $1.4 trillion on improper payments since 2003, which is when federal agencies began tracking them. These calculations are based on sampling and reviews of individual data sets, rather than a comprehensive analysis of all spending.
But that could soon change, as GAO wrote back in October:
“Imagine, for example, being able to throw the analytical power of AI and machine learning at the billions of data points in the Department of Treasury’s general fund and getting to the point where you could identify every improper payment without the need for any statistical sampling at all.”
2. Ending Federal Payments to Dead People
A frustrating example of the government’s struggle to prevent improper payments are the federal benefits paid to dead people. The Social Security Administration (SSA) is responsible for maintaining a database, called the “Death Master File,” of deceased persons that can be used by certain federal agencies to verify that beneficiaries are still alive. But GAO and the SSA Office of Inspector General (OIG) have found problems with SSA’s data collection and management over the years, including more than a million dead people not included in the list. A 2015 OIG review found more than 6 million people had not been identified as deceased even though they were older than 112 years at the time. In 2019, the SSA OIG conducted multiple state-by-state reviews comparing state death records and identified .
Applying data science (including continuously analyzing multiple data sets) could help the SSA improve the management of its death records. In turn, this would also help SSA and other federal agencies stop sending payments to dead people.
3. Monitoring Federal Disaster Assistance Programs to Save Lives and Tax Dollars
Better use of data analytics won’t just help prevent fraud and abuse. It will also improve federal programs, including those focused on helping protect Americans during their greatest time of need.
In October, GAO reported that the federal government has spent more than $450 billion on disaster assistance since 2005, the year Hurricane Katrina hit the Gulf Coast. The Federal Emergency Management Agency’s broad responsibilities include supporting immediate recovery efforts (including saving lives), overseeing long-term recovery projects, delivering federal assistance, and the hazard mitigation process, which aims to lessen the impact of future events.
The DHS Office of Inspector General regularly audits disaster recovery and has questioned the cost of major infrastructure projects. GAO is developing a “disaster resilience framework to support analysis of federal opportunities to facilitate and promote resilience to natural disasters,” and plans to publish it soon. Applying modern data analytics to FEMA’s disaster recovery work could improve efficiency and resilience, including by modeling and analyzing whether mitigation investments actually save lives and property.
4. Preventing Human Trafficking and Child Exploitation
Recent watchdog audits and bipartisan investigations have examined how the Department of Homeland Security’s approach to data management hinders its ability to prevent human trafficking and potential child exploitation. The findings highlight how stronger application of data analytics could help.
A 2016 DHS Office of Inspector (OIG) audit found that “274 subjects of ICE human trafficking investigations successfully petitioned USCIS to bring 425 family members and fiancés into the United States.” In other words, suspected traffickers successfully applied to DHS for immigration petitions (presumably for potential trafficking victims) while under investigation by DHS for human trafficking. The OIG blamed the problem on poor data management and information sharing between two agencies within DHS. The OIG warned that “some human traffickers may remain unidentified and free to abuse other individuals” until the problems are fixed.
A 2019 bipartisan investigation (which I worked on) examined “How the U.S. Immigration System Encourages Child Marriages,” and found that the U.S. Citizenship and Immigration Services (USCIS) approved more than 8,000 spousal or fiancée immigration petitions involving minors over an eleven year period. The review identified “weaknesses in USCIS’s management and administration of spousal and fiance immigration petitions,” which “leave minors vulnerable to fraud, child exploitation, trafficking, and forced marriages,” in part due to the agency’s reliance on paper-files and poor data management.
5. Identifying Root Causes to Address the National Overdose Crisis
The Centers for Disease Control and Prevention reported that more than 70,000 people died from overdoses in the United States in 2017, or more than 190 per day. GAO wrote that this national crisis “has worsened despite dozens of ongoing federal, state, local, and private sector efforts to prevent drug misuse and to treat substance use disorders.”
While there is clearly no simple solution to this major societal problem, the effective application of data analytics could help identify root causes and point policymakers toward solutions. For example, opioids entering the country through international mail have contributed to the overdose crisis, and better data collection and information sharing between the Postal Service and Customs and Border Protection could curb the problem, according to a new GAO report. Overseers could also use data analytics to explore other potential root causes -- such as links between the Social Security Disability Insurance program and high rates of chronic opioid use -- to understand whether Americans are getting appropriate care, and how to prevent overdoses.
Conclusion
Each of these problems are areas of bipartisan, national concern. Effectively applying data analytics could drive billions in taxpayer savings, improve government performance, and even save lives in some cases.
To be sure, solving these problems will require more than oversight. Federal agencies will need to improve government management and update their data management. Congress may need to change laws and use its power of the purse to force executive branch action.
But watchdogs can be a catalyst for change. That’s why fiscal conservatives and everyone who wants good government should welcome strengthening the GAO.
