Electronic voting is an entire field of research. As with cryptography, do not come up with your own methods.

Here are some problems with your method: allows coercion (is not receipt-free), violates vote secrecy (attacker can just force voter to reveal password), allows ballot stuffing, allows partial results (no voting under equal circumstances), probably completely relies on server for "security" of passwords, salted hashing orly?, hash tied to voter by timing info, ...

There are voting schemes (the more sophisticated ones are cryptographic, but interestingly there are non-cryptographic voting schemes like Punchscan that have interesting properties) that address these. But it gets fairly complicated fairly quickly, and at some point normal voters won't be able to make sense of them. The cryptographic voting schemes also tend to get computationally unwieldy, e.g. by requiring huge mixnets.