It is common knowledge that money wins elections. If we look at tables containing the candidates and the money they spent, there is a clear correlation between money spent and who wins. This is the whole basis for most of the uproar about sponsorship scandals. But remember : correlation is not causation. Although more money might cause higher chances of victory, it could be that higher chances of victory causes more money. Or it could be that a candidate's appeal to the voting public causes both money and high victory percentages.
Let's first use an analogy to simplify the issue. When do you cheer for a sports team:
A) When they have a sure win
B) When they stand a possibility of loosing
C) When they stand a possibility of winning
D) When they have a sure loss
Most people would answer A, B, and C. If you answered A, it is probably because you wanted to cheer for your favourite team, which is currently dominating. If you answered B, it is probably to encourage your team to play a bit harder to minimise the chances of a loss. If you answered C, it was probably to get the final push going and claim victory. But if you answered D, it is because the team is your favourite, and it is unlikely that you would ever not cheer for it. Now just replace "team" with "politician and party", and "cheer" with money.
But if you don't like the logic type of argument, because the situation is actually more complex, then entreat the statistical argument. In the book Freakonomics, authors Steven D. Levitt and Stephen J. Dubner provide statistical proof that the candidates spending hardly matters at all when it comes to elections. They do this by comparing elections where the same candidates ran against each other (because the candidate is relatively constant). They found that when a candidate slashed their spending in half, they lost only 1% of their vote. (They used the 1000+ Congressional races where two candidates ran together for several times in a row, from 1970 until 2005, when the book was published. You can read the entirety of this argument on pages 8 to 11.)
But if you present the anecdotal evidence of various politicians who were elected "by money", as it would seem, allow me to present counter-evidence. Steve Forbes and Michael Huffington did not win their elections, and neither did Thomas Golisano, who spent $93 million of his own money over three elections, and won 4%, 8%, and 14%.
In conclusion, money does not win elections, and neither do elections win money. The chances of success of a candidate, based on their personality, their appearance, their charisma, and their platform win elections. Money is simply a byproduct of the of the chances, and has very little effect, the same way cheering has very little effect on the game at hand.