Well its a longterm issue with haldol, not due exactly to the drug itself, to its toxic metabolite, HPP, which like MPTP, is uptaken into dopaminergic neurons, specifically targeting those of the nigrostriatal tract, entering nerve cells in the substantia nigra, and when those kinds of toxins get into cells, they are converted to a charged metabolite, such as HPP+ or MPP+ via monoamine oxidase, IIRC its MAO-b which does this, and then due to being charged, the metabolite cannot leave the cell, and quite literally 'fries' and destroys the nerve cell.
Loperamide (immodium) the antidiarrhea drug also has a metabolite with similar properties, although in adults, loperamide doesn't get to go fuck the brain a new arse ring because its uptaken past the BBB, but immediately gets pumped back out again by P-glycoprotein, a multidrug efflux pump. It cannot be given to newborns or very young children for this reason, tox studies in immature piglets which lack a fully-formed blood-brain barrier have shown lethality when given loperamide.
A chronic, low-dose of such a toxin, acting in the way it does, is still going to cause damage. Personally, I think haldol should have been, and still should be taken off the market entirely. There are enough APs, both typicals and atypicals, with a widely varying selection of activities, that it just has no good reason to remain in clinical use.
http://www.docguide.com/haloperidol-not-olanzapine-linked-brain-shrinkage-schizophrenia-presented-sfnhttp://www.ncbi.nlm.nih.gov/pubmed/15756305 Similar comparative study in rhesus monkeys, chronic treatment, olanzapine and haloperidol, as well as placebo compared.
Look around, IMO the longterm effects of APs are quite scary.