While the latest trend is to over-estimate Perry's strengths as a primary candidate, Romney still remains institutionally robust. More interestingly, it's entirely possible that a Romney-Obama race could change the political map as we know it.
Romney raised $18 mi in the first quarter of this year. That is nothing to blush at, although it pales in comparison to the $86 mi raised for the president's re-election efforts. That said, money isn't everything, and outside groups like American Crossroads that are determined to smear the president haul mighty amounts. But before Romney can face Obama, he has to win the nomination.
Whoever wins the early primary elections will not only get a polling boost (if the well-documented historical trend will stick) but will also set a narrative defining themselves and the other primary candidates. For Romney, as a monied, institutional wonk-type, who suffers from significant baggage that wedges the conservative GOP base against him, Romney has decided to not worry too much over Iowa. That's a reasonable move, because, according to Nate Silver, Iowa's Republicans are the most conservative in the country.
Still Romney, has to do well somewhere in the early primary states, because otherwise money will stop flowing in and voters could cease to consider him a serious candidate.
Timeline of the early primaries via Wikipedia:
February 6 – Confirmed date of the Iowa caucuses[129]
February 14 – Expected date of New Hampshire primary
February 18 – Confirmed date of the Nevada caucuses
February 28 - Confirmed date of the South Carolina primary
Romney, as someone whose demographic strengths supposedly lie in the substantial Mormon vote, moderate Republicans, Wall Street, and looking credible and serious to independents (as compared to the rest of the increasingly shrill GOP) needs to win New Hampshire and probably Nevada. According to Silver's analysis, New Hampshire is the 4th least-conservative Republican primary electorate. So if Romney's baggage or the charisma of a Perry or Bachmann costs him a win there, Romney almost certainly is done.
It's pretty much impossible for Romney to win the nomination without some conservative support, because even in New Hampshire, 55% of 2008 Republican voters identified as conservative; in Nevada, it's 75%, and Romney needs places besides the NE.
Romney benefits from a fractured primary electorate. As long as the more conservative candidates (Perry, Bachmann, Gingrich, Cain, Santorum) are relatively close in support, Romney can benefit from the "electability" argument enough to pull ahead of the pack. Unlike the Democratic primaries, the Republican process awards winner-take-all delegates in some cases, (see Brooklynbadboy's comment below) so Romney can win contests closely and still end up the nominee.
Romney's strategy, then, is to attack Perry until Perry's support drops down to around parity with Bachmann, and possibly even aide the other rightists. The problem for Romney is that as candidates like Santorum lose steam, growing weary of campaigning, watch the money dry up, and fall into low single-digit support, they will drop out and whomever becomes the strongest conservative candidate will suck up most of the lesser candidates support. Romney can only benefit substantially from knocking out Jon Huntsman, who probably isn't getting a lot of primary votes anyway.
States with large Mormon populations where Romney could do very well include Nevada, Oregon, Colorado, Utah, Idaho. Massachusetts and Michigan could constitute favorite son states again. Romney needs the rest of New England, since these are individually small states: Maine, Vermont, Connecticut, Rhode Island. Romney could do well in the purple, upper Midwest.
While Romney's religion, healthcare legislation in Massachusetts, and other moderate credentials will damage him with the primary electorate, the very prospect that he could beat Obama in swing states, especially if the economy receives more bad news or a scandal hits the White House, makes or should make Republican mouths water.
Maine and New Hampshire are very dissatisfied with the president, and Romney's the only viable Republican nominee who can exploit that. I find the claim of Romney competing in Massachusetts accurate but not especially convincing: for the first half of 2011, Obama's approval in MA was at 57%. Even if Obama continues to bleed support, not approving of the president does not convey support for a Republican, at all; some MA voters are to the left of Obama. What's more, most of the Republican candidates have very high disapprovals and low approvals themselves, and it's still early.
Romney will strongly make appeals to Jewish voters, Independents, seniors and "Former Democrats". Ostensibly, Romney would move back towards the center once the nominee, but he'd possibly choose a Bachmann or Perry as VP to cement evangelical and conservative support. Healthcare may alienate some conservative voters but the reality is, most independents aren't furious about the issue anymore. The election will be about the economy, and Romney's strengths lie in looking competent and reasonable, the way Obama did after the financial crisis of September 15, 2008.
This map takes the states where Gallup found Obama's approval below average and considers them red. Traditional swing states where Obama polls at average approval show up in gray. His strong approval states remain blue. If Romney can't win the below average approval states, he can't win...
A scenario where Romney wins the general election could look like this:
The West is not only moving against Obama in some ways, it's also comprised of the Mormon Belt. Romney's moderation and religion would be emphasized in order to win. That said, the southwest has a strongly growing latino population and Romney would be expected to lose in more latino areas. One commentator took issue with this example, where Obama loses Nevada and gains Arizona. With a typrical Republican, I'd agree this is essentially impossible. However, Obama has decent approval ratings in Arizona and the state has far more latinos than Nevada. The other unspoken point is that if you switch the two states, Romney still wins. If Romney isn't fighting hard in Nevada, he's probably not going to win anyway.
The rust belt. While the choice of giving Michigan to Romney is somewhat arbitrary, (and not especially likely to me, given the support for the Big 3 that Obama staked political capital on) Obama is losing support among unions, working class whites and in the upper midwest. It's also possible he could lose Wisconsin. This map is just one scenario. Romney will be competitive and fighting for votes there, and in OH, PA, FL, IN, MO, CO, NV, and the south.
The south and New England could fracture enough to produce an ideological realignment. In this case, Romney takes New Hampshire and Maine, competes effectively but loses Massachusetts, and loses in some key southern states. Obama's best chances for southern wins are Virginia (I presume Obama keeps Virginia barring a landslide loss) North Carolina, Mississippi, Georgia, and Florida. Winning only a few of these could permanently alter perceptions in the country about liberal and/or black politicians, and southern voters, even if Romney becomes president. It's important to realize that Obama is polling so well in the south compared to Republicans because southern voters, while conservative, are also more authoritarian, and Obama has the advantage of incumbency.
If Obama has an exceptionally bad night, Romney could win in a landslide, but liberal states like Hawai'i, California, Oregon, Vermont, and the mid-Atlantic would probably remain loyal to Obama, even where the president's approval lags. Romney will not be liberal enough for voters in these states.