One of the key ideas of economics is that incentives matter. This is a simple but deep idea. Incentives dictate outcomes more often than hopes, "what's right", or need.
Understanding homelessness is difficult [1]. There are not very good numbers available and our understanding of homelessness is murky. Homelessness began to rise in the early 1980's. Between the early 1970's and late 1980's the number of beds in U.S. mental hospitals was quartered. This "deinstitutionalisation" has been suspected to cause an increase in the number of homeless [2]. Bottle deposit was introduced roughly in the late 1970's and early 1980's in the United States. The coincidence of bottle deposit introduction and increase in the homeless is interesting. But is it meaningful?
Many factors influence homelessness. In the late 1960's and early 1970's renewal in American cities lead to old and inexpensive housing being replaced by more financially lucrative uses. Containerization
and forklifts and other advances greatly reduce the need for casual labor and greater skills are required to find employment. Innovations in drug technology (crack, meth, etc.) reduces price and increases cost. Social welfare programs constantly change. Time frames for various forces are not crisp - for example deinstitutionalisation started as early as the 1940's. Alcohol and the homeless equate in the minds of many. The variety of potential influences, their overlap, the spread out implementation and diffusion of impacts all make understanding homelessness difficult.The idea that bottle deposits is a key factor in homelessness, is enticing in some ways. Why? There is a correlation there, so the possibility of this being a factor exists. The idea is positive, as it sees the homeless not simply as victims but as economic and hard working people who perform an useful job: would you be up to the task of digging through the trash to earn your paycheck? It is also positive in the sense that it improves the life of the homeless: availability of ready cash is always good, and it is now easier to escape a worse situation such as an abusive home. Viewed in this light deposit programs are a vital social program, one where everyone benefits [3].
Deposit on beverage containers reduces the impact of homelessness by supplying a low skill and ubiquitous job. When costs of something are reduced one should expect to see an increase in its consumption [4].
Notes:
[1] It appears that the homeless rate is roughly the same in Europe as in the US/Canada. For all the claimed differences between the US and Canada, or the US and Europe, most seem to fall into what Freud termed the "narcissism of small differences." If homelessness isn't much different, what true deep and significant differences exist? Most of the differences seem to be small differences of some parameter, yet all the institutions and frameworks are largely similar if not exactly equal.
[2] Of course correlation is not causation. It is best to keep ones mind open to many different correlents before making up ones mind regarding possible/plausible causation. I'm making a claim here regarding a correlation I have not heard discussed before.
[3] One possible negative impact of putting this idea out there is that some cities could change bottle return policy in order to discourage "vagrants" - simply by making it a requirement that bottles be returned by car (for some lame/fake "safety" reason). Of course attaching this idea to the bigger idea could help minimize the likelihood. If Rambo taught us anything - and he has taught us a lot - it is that cities will stop at nothing to enforce "not in my backyard". Rambo taught the city but good, however one battle in a war is nothing.
[4] Call to econometric masters out there - look at data and bottle introduction times, I think this could make a very interesting little paper. Data is sparse, but perhaps deaths of homeless are one recorded number (one would expect a dip in deaths as life got easier, and then an increase as number of homeless increased due to homelessness becoming more attractive) that could tease out information. Some cities may have decent numbers (???). A quick look around suggests that our society doesn't track things like this very well so I leave it up to someone who can arbirtage their knowledge here. Or, actually, we don't seem to track much of anything very well - quick: how many homeschoolers are there? Well, the current estimate post huge movement garning attention is something like 3% of children - however, since it is required by law to send your children to school the exact number should be known. Even for things were we should have data - we track birth/deaths, and kids are suppose to be in school so their absence should be investigaged to ensure they are not chained to a radiator somewhere- we don't).
