It's easy to think that people are "in it for themselves" as the saying goes, with their policy and political preferences motivated by narrow self-interest. But decades-old research shows that this often is not the case - and a new analysis we've conducted of attitudes about Obamacare backs it up.

We looked after hearing a comment at a polling conference last week echoing others we've noticed about the Affordable Care Act - that the law lacks broad public support because people think it's harming their own health care or increasing their costs. When policy impacts get personal, the thinking goes, look out.

Not so fast. While that's in line with conventional wisdom about the power of self-interest, it's out of whack with a different notion that I've often seen supported in our own survey data - the political scientist Donald Kinder's sociotropic model, proposing that Americans' broader concerns about what's best for society trump their own self-interest in driving political and policy views.

It's testable, and so we did, using data from our most recent ABC News/Washington Post poll late last month. We included a variety of predictors - demographics, partisanship, ideology and four questions on Obamacare, measuring views of its impact, both personally and nationally, on health care overall and on health care costs. We ran two regressions, one predicting support for the law, another, just for fun, predicting the role of views on Obamacare on House vote preference this fall. (A regression measures the independent relationship between two variables when others are held constant.)

The results are pretty interesting, and they do support the sociotropic model.

Our first analysis shows that views on the ACA's influence on the quality of care - rather than its impact on cost - are the strongest correlates of support for the law. And it's attitudes about the law's effect on the entire healthcare system in the country, rather than its effect on the quality of care that individuals experience personally, that by far are the most important.

Further, neither personal nor national costs of the law (as opposed to its impact on care) are significantly related to support for the ACA in our model. (Other predictors are significant, of course; e.g., Democrats are more likely to support the law, conservatives less likely to do so, controlling for other factors.)

In modeling the relative role of attitudes toward the ACA on House vote preference, the predictive power of views of the law's effect on the quality of personal care disappears entirely, while views on its impact on the healthcare system overall remain significant, and strongly so - following only political party identification and ideology as correlates of vote preference in this model. (Specifically, those who think the quality of healthcare in the country has gotten worse as a result of the ACA are more likely to vote Republican, again controlling for other factors.)

In this case, views of the ACA's impact on national costs of health care also significantly predict vote intentions, albeit more weakly than do views of its effect on the health care system overall. Views of the law's impact on personal costs, however, again are not a significant predictor.

It's fair to say that there's some overlap between sociotropic and self-interested views - i.e., the sense that what's good for the country is good for you, and vice versa. But they're not the same thing; views of the national vs. personal impacts of Obamacare on the health care system correlate at .59, and views of its personal vs. national impacts on cost correlate at .42. A correlation of 1 is a perfect relationship, so these are substantial matches, but far from exact ones.

In percentage point values, Americans were more apt by a 20-point margin in our ABC News/Washington Post poll to say the ACA is making the overall health care system worse rather than better. They also were 15 points more likely to say it's making their own quality of care worse, not better. Perceptions on costs were even more negative, with the public 39 points more apt to say the law is increasing rather than decreasing their own health care costs, and 47 points more likely to say it's raising, not lowering, health care costs nationally.

It's also worth noting that correlations don't indicate causality. We posit here that views of the ACA are informed by views of its impact; the obverse may likewise be true. Regardless, our main point holds - contrary to what may be popular belief, views of the law's effect on society at large are more salient than views of its personal impact

The regression analyses we're reporting underscore how these attitudes - in particular, those on Obamacare's perceived impact on the health care system nationally - relate both to skepticism about the law and the Democratic Party's challenges this fall.

Research Analyst Damla Ergun produced the modeling for this report. Click here for details, including the regression outputs.