In many areas we find difficulty in this. The first thing we run into is religious incongruence fallacy or the fact that few people observe their religion in a theologically correct way, most don't know what this means and high theological understanding is also associated with higher likeliness of becoming non-religious. This is because cognitive dissonance becomes far more pronounced the more you understand about a particular religion versus the real and practical behavior you and most others have. Still while religious folks are more than happy to describe them as charitable, compassionate, or good-willed we find no real evidence for this in most cases.
A particularly interesting study on whether or not people would stop to help an injured person. This isn't the full study but it does summarize many of the points. Interestingly those who viewed religion as a quest tended to do far worse than average compared. It was ironic that the way the authors leveraged hurriness (the lack of time) was by making participants late. They were also to be reading the parable of the Good Sammaritan to highlight the incongruence even more:
http://faculty.babson.edu/krollag/or...y_samarit.html
Highlights several things:
http://science.sciencemag.org/content/322/5898/58
Interestingly Moral behavior of religious folks seems to be high while in the presence of other religious folks, lower in the presence of atheists or in anonymous situations.
http://science.sciencemag.org/content/322/5898/58
Studies have shown that American Chrsitians definitely are more charitable... On Sundays... if you include tithing to their church. In general it was also found that the donations to your church acted to insulate one from guilt of not giving at later periods meaning that throughout the week giving less negated the moderate advantage.
http://www.hbs.edu/faculty/Pages/item.aspx?num=37730
Furthermore these effects are shortlived (literally in the process of minutes). Interestingly the level of religiosity had no real effect and may even have had a counter-productive effect as the combination of giving to one's church a long with increasingly general views of the outgroup inhibited people's giving. In short this is mediated more by the situation than the person's beliefs.
http://science.sciencemag.org/content/322/5898/58
An interesting study does seem to show that when rule of law is weak religious folks are less likely to cheat. This is most obviously pronounced within poor communities and destabilized countries than it is in communities like the US. As the relative expectation of justice increases, and the trust in the rule of law increases ipso-facto religious belief decreases leading to the conclusion that religion is a substitute for the rule of law in the absence of it but is quickly abandoned in the presence of it.
https://doi.org/10.1111/j.1467-9280.2007.01983.x
In fact the evidence is quite strong that the rule of law, ethics, and morality in general is an inbuilt human phenomena, in short we're moral because instinctually we're moral and religion is a result of this innate morality in the absence of other systems of moral rule but quickly becomes less and less relevant when moralities are legislated.
https://doi.org/10.1038/nature01963
Some more supporting info:
https://www.sciencemag.org/news/2014...-all-misbehave
http://journals.sagepub.com/doi/abs/...88868309351002
Now when we consider things like violent crime things tend to get very one sided. It should be noted that the most violent countries are also those with the highest religious belief. This is pretty obvious.
http://journals.sagepub.com/doi/abs/...51399020001005
While they find regular attendance is associated with less violence, they also find that the more conservative one is in their religious views the more abusive they are likely to be especially amongst men.
MIT has an interesting book on this and finds repeatedly that Religious observance especially in nations with strong ethical framework for laws, economic prosperity and governance stability is actually maladaptive and leads to far more criminal activity. This is because most religions have far looser standards of ethics than developed nations and often include far more discrimination or intolerance in their dogmas providing a justification for criminality.
https://mitp-web2.mit.edu/sites/defa...6_ind_0001.pdf
In fact this seems to be the case in all criminal activity in at least the developed world, white collar criminals for example are dramatically more religious.
https://www.ncjrs.gov/App/Publicatio....aspx?ID=70827
When we delve into many of the assumptions about atheists we find that not only are they less problematic than the religious but dramatically so. For example Atheist divorce rate is half that of religious folks.
https://secularpolicyinstitute.net/w...on_Atheism.pdf
There's also an idea that atheists lack a sense of wonder and appreciation of the beauty of the universe, this is actually opposite with atheists far more likely to find beauty in the universe and religious folks especially those who are extremely religious to have a pessimistic sense of the universe.
http://www.pewresearch.org/fact-tank...bout-atheists/
Overall Atheists are 10 times less likely than average criminal, which is about 6 times less likely than Protestants. When we account for economic and racial differentiation this number shoots up dramatically with atheists doing even better. We find the same thing with regards to education.
Hell we even find that atheists are more likely to know theology than their religious counterparts:
http://www.pewforum.org/2010/09/28/u...wledge-survey/
The list goes on and on and on. This is more than enough to assert religion has no effect on morality, and it does give tentative indications that it may even be on average a problem for society rather than a virtue.