Spatially joining polygons centroid within

[DavidMatthews2]

I have two polygon feature classes that I want to do a spatial join between and sum certain fields. The target feature  has larger polygons (regions) covering the second feature of smaller polygons (property parcels). I would like to use the centroid of the property parcels to do the join as they are not always fully contained in the regions but 'has their centre in' only joins the opposite way, with the target feature being the smaller polygons. 

Is there anyway to do this without creating a point feature from my parcels and using 'contains' as my join?

Thanks

[Jan_PeterGlock1]

Thank your for your comment Marianne. In fact I made a mistake stating that "A's centroids are always within B (but B's centroids are not always in A)". It is just the other way round: B's centroids are always in A but A's centroids not in B.

My aim is to have the attributes of B and the geometry of A. For "have their centroid in" i get "if A's centroid falls into B." As said, this does not work for me.

I am working with two different representations of city blocks from different sources, where one of them (A has the geometry I need) and B has information on the building density.

[JohannLall2]

This is an analysis that pretty much every government needs to do at some point. We have zoning, or regions, or sectors, or whatever, and we need to know which parcels fall into which zone but the edges usually don't match exactly

[RobertStevens]

Because the join  is  made in order to dissolve and/or ignore the joined features in a suitable way (sum, average etc) and ultimately to retain just one copy of the larger geometry.  With disparate data sources there is a problem when boundaries of two polygonal sets do not quite match. Neither completely within nor intersect address this issue, whereas having the centroid of the smaller polygons with the larger does. Clear?