I do agree with Sophable that good attachment in childhood can be really crucial to people's life chances. But I think that attachment can be much more complex than it first appears and that we do not yet fully understand how it works- big mix of nature, nurture, personalities and environmental factors, I suspect. It is certainly not always the case that people who had one/two care givers till the age of three have more secure attachment patterns than those who have experienced group based care at a young age. And attachment is not the only important factor either, imo, though it is crucially important.
I think that the unhappiest and most insecure children I have met personally have all been cared for exclusively by their mothers in their early years, as it happens. But that fact in isolation is a red herring. A cocktail of parental poverty, physical and mental health problems, domestic violence, poor housing, low expectations, lack of play opportunities, poor access to education etc etc etc made their lives hard from the moment of conception. Nursery can be a vital intervention in these sorts of circumstances, to give children an alternative experience and more stimulating environment, to offer support to parents having tough times, to allow parents space away from their children to recharge maxed out batteries as well as, for some, offering a chance to study or take paid work which could lift the family out of poverty. Sometimes having space away from children allows parents to keep things together enough to develop better attachment relationships with their children, paradoxically. That was the idea of sure start nurseries, I believe, and they have been a qualified success, imo, failing only insofar as they have not always reached the most needy who tend to fall off the radar of services. I do think OJ's criticism of Sure Start is a massively cheap shot actually. I can think of a couple of sure start buildings/schemes I know which really are community jewels set within very isolated, rough, poverty stricken areas.