Adopting the Ask Once Policy in the US
This post is in response to an assignment for the course DPI-662: Digital Government: Technology, Policy, and Public Service Innovation at Harvard Kennedy School. The prompt is as follows.
The U.S. government is thinking of adopting an “ask once” policy — allowing them to collect and integrate citizen data across federal departments/ databases, with huge potential benefits for seamlessness and govt. analytics — but requires data sharing between various agencies. Write a memo with the challenges and risks of adopting this policy and your position on whether it should be implemented.
“Other software companies have incorrectly assumed that the future will look like the past, forming their strategies based on assumptions about a world that no longer exists. A focus on targeted analytical tools and optimizing specific functions within complex organizations is insufficient. We believe that software must connect the entire enterprise.”
- Palantir’s S-1 Filing
Ask Once should be a North Star
Adopting the “Ask Once” policy would require the US government to build an omniscient identity database where, as an example, individual government issued IDs are automatically tied to public and private sectors databases including those belonging to the IRS, US Treasury Department, intelligence agencies, and healthcare institutions to name a few. Some arguments for the policy are that Ask Once and associated digital identity initiatives could dramatically improve public sector service quality and save the US taxpayer billions of dollars annually in fiscal savings and increased efficiency, in addition to potential ancillary benefits such as electoral integrity, increased efficiency in healthcare, and improved ease of doing business.
However, as we were reminded in Netflix’s The Social Dilemma, “nothing vast enters the world without a curse.” Ask Once comes with its expected share of baggage encompassing cyber security and privacy concerns, government manipulation big-brother style, legal and regulatory concerns, inherent risks in overhauling 60-year-old digital infrastructure, political inertia, and the list goes on. With that said, my view is that Ask Once should be a “North Star” for the US that is implemented in multi-year stages. As part of this digital renaissance, the US government would be well-served to build internal capacity to create more robust policy frameworks, build trust with the American people, and work through the inevitable animated debates around States rights vs. Federal oversight.
“We are working towards becoming the default operating system across the U.S. government.”
- Palantir’s S-1 Filing“
Leaving it to the private sector
In recent US history, the private sector has led innovation. For the plethora of arguments against Ask Once, a closer review of the issue reminds us that we have across various government institutions already outsourced the aggregation of existing government databases to private companies such as Palantir. Palantir contracts with dozens of US government institutions including the CIA, FBI, ICE, DHS, and the CDC to connect and analyze databases not only within the government sphere but also “databases containing telephone, email addresses, financial data, call transaction records, and social media information…used by the federal government to make determinations that affect people’s lives with little to no oversight” according to Jeramie Scott, senior counsel for the Electronic Privacy Information Center (EPIC). Over the years, Palantir’s lobbying efforts have skyrocketed and a good chunk of their business comes from government clients.
The risks of relying on companies like Palantir to do the government’s innovation on its behalf versus building capacity inhouse outweigh the benefits. Take for example the program Palantir is rolling out to hundreds of police departments to grant them access to granular surveillance networks without giving customers (the US government) much clarity on how the software algorithm works. Below is a sample of the types of information that are aggregated and analyzed by Palantir Gotham, used by some law enforcement agencies across the US.
As it stands Palantir seems to have laid the foundation for a monopoly on surveillance data that the US is becoming increasingly dependent on. Palantir’s CEO Alex Karp said it best, “[the CIA] may not like us. Well, when the whole world is using Palantir they can still not like us. They’ll have no choice…That’s de facto how we got the FBI, and every other recalcitrant place.”
Recommendation: build for the future and start small
If we learned anything from the Treasury’s $1.4 billion Coronavirus stimulus payment to dead people — due to lack of access to the Social Security Administration’s full set of death records — is that cross-agency collaboration is necessary and often leads to a well-functioning government. Whether its healthcare, business operations, identification cards for public transportation, we must build for the future.
The US faces monumental challenges in implementing Ask Once such as the public’s general distrust of the US government, existing 1960's tech infrastructure, and the political and governmental structures that severely hinder national progress. However, starting small, for example tackling one system at a time and calibrating along the way, may be the way to go. Adoption may be slow but steady. When it comes to digital innovation, the US cannot be left behind.
“In order to register my daughter to go to school here, I had to bring a photocopy of my electricity bill. To prove I live here. What!?”
- Former President of Estonia on moving to Silicon Valley