It could be that the received wisdom on predicting election outcomes is wrong. But what is more likely wrong is the interpretation of how the US economy is actually doing. Those who have studied the numbers say choosing the right economic indicators is key to understanding this election. In other words, it's not so much whether the economy matters; what matters is what you look at.
When Carville put up that sign in 1992, the stock market was a reasonable indicator of the health of the economy.
If the Dow Jones Industrial Average was up, the economy was doing well, which meant the incumbent would reap the benefits.
Today, after the 2008 financial crisis, the stock market is no longer such a reliable indicator. Indeed, if there were a correlation between the stock market and election outcomes, Obama would be reelected in a landslide. The Dow has gone up over 65 per cent since he was inaugurated in January 2009.
Pollsters and political analysts have been trying to find what economic indicators best predict election outcomes, pouring over data from the last 50 years. Statistician Nate Silver looked at 43 indicators and he has come up with a composite index that he tracks on his blog.
He has written that Gross Domestic Product (GDP), a good indicator of the overall health of a country's economy, is too slow to have an effect on voters. Reporting is done quarterly and the figures are published a quarter late.
The unemployment rate would seem to be a good indicator. It's updated frequently, and reported constantly. But what Silver and others have found is that the rate itself has almost no relation with election outcomes. However, the change in the rate during the election year is a very good indicator.
So it's the change, not the actual numbers, which could explain Obama's marginal advantage in the current economic setting, as the unemployment rate dropped below eight per cent in September for the first time since he took office.
Another indicator is the consumer confidence index: the measure of optimism, expressed by consumers in the way they spend and save their money.
The most recent measure put the US index at 70.3 in September, up from 61.3 in August. Increased confidence in the economy can translate into a boom for the incumbent, as satisfied voters won't vote for change.
But indicators are one thing. What are the candidates themselves saying?
Republican Mitt Romney is trying to get voters to look at their wallets.
"Are you better off today than you were four years ago?" is the question Republican Ronald Reagan asked the week before he beat incumbent Richard Nixon in 1980. Romney has been focusing on that question as well. For many voters, the answer is no and Romney is hoping to convince them that his business experience - as the founder of Bain Capital - will allow him to change that, by leading the US out of its economic woes.
Obama's story is more about the past, trying to get voters to look back beyond the immediate crisis to when he took office. He would like people to see the last four years as one of Democrats undoing the destruction left to them by the George W Bush's Republican administration; orchestrating a recovery of a mess they did not create.
To balance the budget, Romney has said he would close tax loopholes so as not to have to raise taxes, an idea that Obama has said does not add up.
Obama wants to raise taxes on the wealthy, but like Romney, has been thin on the specifics of how he would balance the budget.
Ultimately, it may not be their plans and arguments that make a difference but the numbers. Unemployment figures for October are due out on Friday 2 November, four days before election day.