In a capitalist society, your socio-economic mobility is closely tied to your ability to use financial instruments to achieve your goals.
Many of those tools are related to debt, including loans and access to lines of credit. Without the ability to incur debt, our financial muscle is limited to the cash and assets we have on hand.
But today, the emerging influence of algorithms can have an outsized, negative impact on mobility for many people.
The power of credit
In a 2018 report, the Consumer Financial Protection Bureau (CFPB) estimated that 26 million Americans were “credit invisible.” That means they lack any credit history and are invisible to lenders, who need more information to decide whether to approve a loan.
Moreover, according to the report, another 19 million people were “unscorable.” They had some credit history, but the information was insufficient to give them an accurate credit score. This means nearly 1 in 5 of the adult US population has no access to powerful financial tools that enable us to climb the economic ladder.
Debt is an instrument that allows us to defer payments into the future and distribute them over time in exchange for greater purchasing power in the present. A credit card is the most obvious example, but there are other tools we can use for this purpose. Mortgage refinancing, for instance, can be used to not only make homeownership easier but to take cash out on your home’s value.
With debt, you can buy things that improve your quality of life or unlock opportunities. Driving a car gives you the freedom to travel overland. Buying a house offers long-term savings compared to rent and the ability to settle down in the location you desire.
The problem with algorithms
The inability to access the financial tools that leverage the power of debt means many people are limited to what they can pay for with cash.
A few people might prefer that as a lifestyle choice because they aren’t controlling their spending habits. Others are faced with the more urgent concerns of surviving poverty.
Ultimately, though, everyone should have the option to use debt as a means of improving their lives. When some aren’t given that freedom of choice, we have an unequal society. And the CFPB report reveals that minority groups, young consumers, and immigrants are more likely to suffer from credit invisibility, and therefore a lack of socioeconomic mobility.
If human underwriters solely determined access to debt, that might be fixed. People might be following rigid risk assessment policies, but they can use their human intuition and understanding when the situation requires nuance. And they can be held accountable for decisions and outcomes.
However, as in many other industries today, algorithms have supplanted human thinking in large part of decision-making in the finance sector.
Algorithms allow companies to save on time and payroll. But their efficient information processing also offers incentives in finance. Machine-underwritten loans yielded 10.2% greater profits, with a 6.8% lower incidence of default than humans.
However, despite supposed machine objectivity, there’s growing legal evidence that algorithmic efficacy also results in discrimination. The process of machine learning is conducted through human trainers, who can be biased. And the algorithm learns by processing data sets that may be inherently skewed to favor certain groups or characteristics while marginalizing others.
Working against the machine
Unfortunately, algorithms aren’t going anywhere because the business case for AI is too lucrative. And unlike human activity, how algorithms operate is opaque, often referred to as the “black box of AI.”
Developers may be blind to their own biases. The companies that pay them might insist on proprietary rights to their designs, and in any case, the average person lacks the skill to understand how an algorithm works.
Ideally, the realm of finance would subject its algorithms to greater scrutiny. Institutions should constantly observe how the system works and use feedback to determine if certain people or attributes are being discriminated against. Regulatory bodies should hold companies accountable for algorithmic bias, making them swiftly adjust the machine rules towards greater fairness.
That situation might be years away from becoming a reality. Until then, individuals suffering from credit invisibility can use workarounds to gain visibility.
Know your credit score, which factors affect it, and the relative weights given during processing. Start with a secured credit card or store cards, and use them responsibly. Link your bank account to pay eligible bills that help build up your credit file.
Even with an imperfect and unequal system, you can still find a few tools that give you entry and allow you to exercise greater financial mobility.