Most policy discussions about homelessness invariably focus on how to address the “root cause.” This is reasonable enough. If we attack the root of a problem, we expect the branches to wither and die. Presumably, then, we need only to identify the root cause of homelessness to design an effective policy solution.

But is there a root cause?

In some respects, this may be a loaded question. Conflicting answers are not necessarily a matter of truth competing against fallacy. Rather, they reflect wholly different ways of conceptualizing social problems.

In the social sciences, there are two basic approaches to any question. One is essentially sociological, which is to say that the focus is on aggregates. Through this lens, the objects of inquiry are collective entities, such as nations, cities, ethnicities or religious groups. The idea is that we can identify patterns in these aggregates and draw conclusions about the individuals they comprise.

Gregg Colburn and Clayton Aldern adopted this approach in their study of homelessness and gave their answer to the question of root cause in the title of their book, “Homelessness Is a Housing Problem.” They examined cities, not individuals, to identify the factor that best corresponded with homelessness, concluding—rather unsurprisingly—that housing costs provide the strongest explanation for disparities in homeless rates.

Their answer has been the prevailing position among policy experts for decades. It has provided the rationale for the Housing First policies that the federal government has centered its homelessness efforts on, and states hoping to win federal grants can only do so by conforming to the Housing First model.

The operating assumption of the Housing First philosophy for many who espouse it is that once people have stable housing, their other problems will automatically improve. Housing is the root, and issues such as mental illness and substance abuse are mere branches. Policies consequently prioritize rapid rehousing and financial assistance and deprioritize wellness services and emergency shelters.

The alternative approach is individualistic. Under this method, the objects of inquiry are individuals, from whom we can draw conclusions about whatever aggregates they form. Although this perspective is somewhat passé among academics, it helps us answer questions that the sociological approach cannot. Housing may explain disparities in homeless rates between cities, for example, but it cannot help us identify which individuals within a given city are most at risk of becoming homeless.

Although the root of individual homelessness can vary, the strongest predictor is substance abuse. Addiction becomes especially common when we narrow our focus to the chronically homeless whom we find in so-called “tent cities.”

The spread of fentanyl has made it impossible to conceal the role that substance abuse plays in driving homelessness, and it is increasingly clear that housing alone cannot solve the issue. A two-year study from the University of Pennsylvania found that more than half of all fatal overdoses among homeless New Yorkers occurred in supportive housing, challenging assumptions many who support Housing First make.

Not only has Housing First failed to remedy homelessness, but even if it had succeeded, it would have meant little more than letting people die indoors rather than on the sidewalk. If San Francisco or Portland resolved its housing crisis without addressing substance abuse, it would merely look like West Virginia, where homelessness is rare, but poverty and addiction are rampant.

Most homeless individuals require other forms of intervention besides housing support; accepting that drugs are a root cause does not falsify the claim that homelessness is also a housing problem. We can shepherd people through recovery, but most will struggle to achieve independence in cities where median rents exceed $3,000. Housing still needs to be more available and affordable.

Homelessness, in short, stems from many roots and is therefore a problem that can never be solved by policies that attack a single “root cause.” The Housing First experiment has proven a failure of the past decade. It is time we try a more holistic approach to ending homelessness that can address not only housing but also substance abuse and mental illness at the same time.