Most Americans oppose businesses refusing to serve gay people - Reuters/Ipsos poll

By Andrew Chung and Lawrence Hurley

WASHINGTON (Reuters) - Nearly three quarters of U.S. adults believe that businesses should not have the right on religious grounds to deny services to customers based on their sexual orientation, a Reuters/Ipsos opinion poll released on Monday showed.

The findings of the poll, conducted Friday to Monday, were issued on the same day the U.S. Supreme Court ruled in favour of a baker from Colorado who had refused to make a wedding cake for a gay couple, citing his Christian beliefs.

In the poll, 72 percent of respondents said business owners, because of their religious beliefs, should not be allowed to refuse to serve customers based on sexual orientation, while 14 percent said they do have that right. Another 9 percent said businesses have the right "only in certain circumstances" and 6 percent said they do not know.

Like other recent surveys of American adults, the Reuters/Ipsos poll also found that the number of Americans who support gay marriage has increased in recent years. In the poll, 53 percent of respondents said same-sex couples should be allowed to marry legally, with all the same rights as marriages between a man and a woman. This is up from 42 percent who said so in a similar Reuters/Ipsos poll in late 2013.

The Supreme Court legalized gay marriage nationwide in a landmark 2015 decision.

The poll also asked a more general question about how much leeway businesses should have to reject any customer for religious reasons. Fifty-seven percent of respondents agreed that business owners "never have the right to deny services to customers," even if their denial is based on religious beliefs.

Another 19 percent said businesses "always have the right to deny services" and 24 percent said they can deny services "in only some instances."

The Reuters/Ipsos survey collected responses online in English throughout the United States from 722 American adults. It had a credibility interval, a measure of the poll's precision, of 4 percentage points.

(Reporting by Will Dunham)