If demography is destiny, Republicans can't win the presidency by acting more like Democrats. The GOP's best shot in 2016 is not to nominate a moderate. They must nominate a conservative who can attract more conservative voters to the polls, just like President Obama built his own coalition and increased the relative electoral power of each constituent part. Not that it will be easy.
As long as the GOP nominates someone plausible, they start off with 46 percent of the vote and a large chunk of the electoral college. Getting to 270 + 1 electoral votes and then to 50 percent of the popular balloting requires trade-offs and choices.
Where Obama drew in younger voters, unmarried women, black voters, and Latinos, Republicans would be wise to focus, in the short term, on raising turnout among married women, white men over 30, and self-described evangelical Christians.
The ideal Republican strategy is not terribly convoluted. Find and nominate the most acceptable conservative. Find the swing states where demographic composition of the electorate has been volatile and where there is room among those demographic groups to grow the GOP's share. Put the two together. (I would add: If I were Machiavellian, I would urge Republicans to do everything they can to suppress the Democratic vote. I am not Machiavellian, and plenty of Republicans are already doing this.)
Of course, overall demographic trends tilt the balance away from Republicans. The uphill climb begins with an awareness that Democrats will have somewhere between 200 and 250 electoral votes banked by Election Day. I'll start with 226, which leaves Pennsylvania (20 electoral votes) and 30 other states for Republicans to pick up. A lot of states, yes, but the most populous ones are already true blue, and there are fewer large states where Republicans can meaningfully turn out their demographic tranches relative to whatever a strong Democratic presidential candidate might do.
Ohio, Virginia, Michigan, Wisconsin, Pennsylvania, and Iowa are pretty much the only states where there's enough give for Republicans. If the 2016 nominee wins those states, he or she will need at least eight more electoral votes. It is not inconceivable that a GOP nominee can:
(1) Excite the party with the promise of his electability.
(2) Create enough of a contrast with the Democratic nominee to keep the caucus/primary voting base motivated.
(3) Appeal just a bit more to to blue collar white voters and to married women.
Assuming Hillary Clinton will be the Democratic nominee, then the challenge of task two is reduced somewhat. Against Democratic nominee Hillary Clinton, there aren't many scenarios where conservative activists don't unite and stay motivated.
A counter-pressuring challenge, of course, is that Clinton will aim to modify Obama's coalition with one of her own. Her appeal to married women might surpass expectations. It's hard to imagine that she replicates Obama's poor 39 percent showing among white men in 2012. She will exceed that percentage, almost certainly.
What Republicans really can't do is nominate someone who would try to eat into Democratic percentages among younger voters and Latino voters. It's a much taller order to take votes away, and there aren't many states where, unless Democrats massively lose face among these groups, a Republican can take them over. (I see just two: Colorado and Florida.)
So to those who say: The GOP will ONLY win the presidency if it moderates its tone on social issues, I say: not with the electorate as currently constituted. It might be useful, but it is neither necessary or sufficient.
That doesn't mean a GOP nominee needs to learn hard to the right. He or she simply needs to do what Obama did: Find stuff that people want, and promise to give it to them.