This paper will be presented at the CityDAO Academic Conference, taking place virtually on Sept 3-4; please see link for details and to register.
Education is a building block of human flourishing, a foundation from which we see innovation and scientific discovery emerge. We all have a desire to understand the world around us; education provides the key to unlocking this door.
In its current Web2 form, the internet has gifted humanity with historically unparalleled access to knowledge and the ability to acquire skills. The amount of knowledge contained within something as easy to access as Wikipedia would boggle the minds of those who lived in the pre-internet age. The ease at which we can learn basic skills by simply searching YouTube would seem like a superpower to past generations. Other low-cost, high-quality, and easy-to-access online educational resources include Khan Academy, Coursera, and Udemy.
Yet, like many things with Web2, something is missing; in advancing the mission of giving the world access to essential educational tools, Web2 has gone far, but it can’t traverse the last mile of the journey.
Given the ability to access high-quality educational content at a low cost via the internet, one would imagine education is becoming more accessible, cheaper, and improving in quality. However, none of this is true.
Globally, education is not accessible, and most children worldwide don’t achieve minimal learning standards. According to Unicef, “Of the roughly 1 billion children and adolescents who go to school every day, over 600 million are unable to reach minimum proficiency levels in reading and mathematics.”
We also see alarming increases in educational costs, particularly when it comes to post-secondary education. According to CNBC, college costs have increased by more than 25% in the last ten years. Another estimate suggests the price of a college degree in the US has increased by 1200% since 1980; source. We also see similar increases in primary school costs. However, these costs are mainly reflected in housing prices in neighborhoods close to high-quality schools, making them more challenging to measure. Additionally, administrative costs have soared; looking at post-secondary expenses in the US; we see that “Sector-wide, per-student spending on administration increased by as much as 61% from 1993 to 2007”; source.
Finally, measured educational outcomes per dollar spent have seen almost no improvement for decades; the paper “The Productivity Collapse in Schools” by EA Hanushek outlines this problem in great detail. While learning tools improve, increases in administrative costs counter these improvements. While educational content costs go down, the cost to administer it increases by the same percentage. In economic terms, there has been no increase in productivity.
The last mile problem refers to a situation in goods transportation where the final segment from producer to consumer is the most expensive mile to cover. It is indeed a perplexing and troublesome problem. A product could be produced, ordered, and delivered 99% of the way by utilizing state-of-the-art technologies, yet, the whole process has no value if the good can't make the final 1% of the journey.
In these situations, a huge premium can be offered to anyone who can deliver the goods the remaining distance. Importantly, if someone has a monopoly on the ability to deliver the goods the last mile, the premium can be very high. It is a disturbing problem as all cost-saving techniques and technologies utilized in the first 99% of the journey won't reduce the overall price if those delivering the goods the last mile can extract the value.
I will argue the above analogy captures education today. While education has improved in unimaginable ways, the price remains high while accessibility is limited given a type of last mile problem; there remains a costly link in education's journey from producers to consumers.
To understand the nature of the problem, we need to realize that supplying education has two main components. Education is the process of learning and testing. Learning is complex, but at its most basic level, it can be described as the process of your mind understanding something it didn't before. Testing is a process of verifying that outcome. The testing process involves the accumulation of grades, transcripts, degrees, diplomas, certificates, etc. For the sake of simplicity, I will call all of these "verified credentials."
Obtaining verified credentials is education’s costly last mile; it’s the final step in the process that takes the least time, but this final link explains the high and rising cost of education. Simply, learning is becoming easier and cheaper; proving you did so is not. But why are verified credentials so valuable; why can’t we bypass this costly link?
Verified credentials represent information. In 1970 a branch of economics began that has been broadly termed "Information Economics." In brief, the area of research explores why information is critical for markets to function. The scope of study primarily started with a famous paper titled "The Market for Lemons," where the author argued that in markets where it is challenging to verify the quality of products, only low-quality products would be supplied. The author used the example of a used car market; a bad used car is called a "lemon." See this video for an easy-to-understand visual.
The reasoning is simple enough. If you are a buyer, why would you pay top dollar for something that might be low quality? If you're a seller, why would you bring a top-quality item to a market if no one will offer you a fair price? Without a trusted method to separate the good from the bad, we only get the bad.
The takeaway message is that trusted information regarding quality is critical for markets to function. The outcome is frustrating and not optimal for something like used cars, but it doesn't seem like a massive global issue. The problem is that the same logic can be applied to education in the context of labor markets. Learning increases your ability, but what if there is no way to prove this took place? The labor market breaks down, and sadly, the inability to verify credentials erodes the incentive to learn. Skills are required to access the global labor market, but without the ability to verify skills, the global labor market remains out of reach.
The sad reality is that in countries where the educational system isn't trusted, and skills cannot be verified, the labor market will offer a wage in line with the assumption the person isn’t skilled. If the wage offered doesn’t incentivize the accumulation of skills, the incentive is not to acquire them, as acquiring skills takes time and effort. It’s a sort of self-fulfilling prophecy and an equilibrium that is far from optimal.
Verified credentials contain valuable information required for labor markets, you can learn for free, and many do, but the reality is you won’t get paid for the skills you acquired until you can verify you obtained them. But why can't we just produce more verified credentials? Even if we accept they are required for functioning labor markets, it still doesn’t explain their high cost.
Inherently, the problem is the ability to measure learning has not kept pace with the ability to learn. Our ability to measure learning is stuck in a pre-internet world, even though the ability to learn has leaped into a new age. The former rests on internet technology, while the latter rests on a much older and slower evolving technology, networks of trust.
In the intro, I mentioned how easy it is to access free educational resources. For example, Harvard, a top university, has a vast collection of recorded classes and the materials needed to learn a given subject available for free online; see link (note many universities have followed suit). The quality of this free content is very high, allowing you to learn as if you enrolled in the class. How, then, is it still possible for Harvard to charge hundreds of thousands of dollars for tuition while turning away people desperate to pay the price?
The answer is that Harvard is a trusted verifier of information, mainly that a student who attended and completed their program is highly skilled. When you enter the labor market, it is accepted that credentials from Harvard equate to a high skill level. While you can learn for free, the verified credential can only be obtained by formally attending Harvard. You need the Harvard stamp, or one from another top-ranked university, to get out of the lemons market. You don’t pay to learn; you pay to verify you learned.
Trusted verified credentials remain in short supply as building trust takes time and rests on a system of checks and balances that are complicated to produce. Why is Harvard so trusted? It might seem disconnected, but it’s the same reason democracies are trusted more than dictatorships and why you can't bribe a judge in a country with a functioning legal system. It is because they are erected on networks of checks and balances, which is the same as saying they have a degree of decentralization.
The problem is that while trust requires checks on power, power doesn't want to be checked. Centralization doesn't want to be decentralized; it takes a lot of work and time, but what if we don’t want to wait?
Web3 is an experiment in speedrunning trust networks by building them on code and coalescing trust quickly by aligning ownership and stake with incentives in a decentralized fashion. The goal of Web3 is to replace trust created via social networks with trust vested in code. Can we put verified educational credentials on the blockchain and bypass the costly last mile?
Imagine an open test anyone in the world could access that would prove you have a given skill. When the test is taken, the individual connects their wallet, verifying that the person taking the test is who they say they are. Upon completion and a passing grade, a smart contract is triggered that issues a soul-bound token that you passed the test. A verified credential that is immutable and stored forever on the blockchain. The process is immune to bribery, and cryptography ensures the process isn't fraudulent.
The above would be a huge breakthrough and a great starting point, but it isn't a complete solution. We would also need to answer the following questions. How do we know that the person who developed the test can't sell answers beforehand? How would we know the test is a good one? The system still requires incentives to improve the process in a decentralized and trustworthy fashion.
A decentralized education system would need to ensure those with a stake in its success can validate and improve the process. Web3 is about ownership; ownership in a system you want to validate and build upon to ensure that what you own retains its value. It's a positive feedback loop. Web3 is speedrunning much of human history, using code to coordinate and structure complex social interactions, thus bypassing time-intensive processes by assigning ownership and stake to the right people.
Web2 was built on the ability to transfer information, but it can't solve the trust problem. Supplying trusted verified credentials is supplying valuable information. But trust is inherently difficult to produce. Web3 can build trust in the base layer. With this potential in mind, web3 could change education forever, allowing for the completion of the last mile. Ushering in an age where education is low-cost, high-quality, and accessible to everyone in the world who demands it.
In short, Web3 = trust!
Written by Scott Auriat (scotta.eth) – This article was made possible with funding by the CityDAO Research and Education Guild