Specifically, more on Mark Harrison's Christmas post.
The problem with Mark's thoughtful and open-hearted argument is this: it is undercut by admitting, and accepting as valid, some people's dislike of mixing with those of other cultures. But as soon as you have done that, the economic case for migration looks like this:
NET BENEFIT OF MIGRATION = MARKET BENEFIT + NON-MARKET LOSS
Here term 1, the market benefit, is large and quantifiable and regularly rehearsed by economics professors; but term 2, the non-market loss from people being exposed to cultural diversity, is unquantifiable. So now we no longer know whether migration is a benefit or a loss. Also, since apparently most people in the UK would prefer less migration, term 2 may be large.
Mark also admits that people who do not like diversity cannot avoid it by simply moving to more culturally uniform neighbourhoods, since this uniformity may swiftly be lost. Indeed, that is one reason to have national borders, since they preserve some level of uniformity, while allowing free movement (and the resulting gains from trade) within the itnerior.
There is an unavoidable choice here. You can take people's preferences for cultural unity seriously. If so, then it seems that the majority preference is for less immigration, and it is hard to see why that should not be respected - the most basic characteristic of any club is its membership criterion, and if this should not be decided democratically, what should be?
Or, you can think of people's views about immigrants not as raw preferences, but as mistaken - perhaps even toxic - beliefs. If so, then they do not come into the utilitarian calculus, any more than the public's mistaken beliefs about the costs of foreign trade. That is arguably a more liberal and optimistic position, but also a less democratic one.