Joint physical custody has lots of problems, and even when ordered, judges are often reluctant to enforce it. Disagreements usually inure to the benefit of the mother in these cases, and I have seen as many joint custody fathers get screwed as noncustodial fathers. Most parents don't have money to litigate, and many litigate and go broke fighting over custody. This is truly a disaster for the entire family, but it happens way too frequently.
I have the answer: the "time-shift shared parenting order". I have long pushed for a "time-shift shared parenting order" (or in father's rights parlance, a "delayed gratification order"). It gives primary custody to one parent until the halfway point from the current time until the time the child goes to college. At this halfway point, primary custody automatically flips to the other parent -- at a date specified in the original divorce decree (no additional litigation necessary). So, if you had two kids, ages 5 and 7, the 7 year old would go to the other parent when he/she reaches the age of 7+ ((18-7)/2) = age 12 1/2. The other child would go to the other parent at age 5+((18-5)/2) = 11 1/2. The exact date of custody transfer will be written into the original decree so it is immediately enforceable without litigation. If the judge we concerned about separating the children, he could average those two ages to come up with a custody change when the oldest child reaches age 12.
The time-shift shared parenting order will work, and is easily saleable because:
1. The parent who does not get "first shot" knows they will get their turn, and so they don't feel totally devastated and robbed. We won't see so many dejected fathers dropping out of society, quitting work, committing suicide, etc.
2. Courts usually give custody to mothers because of adherence to the "tender years" doctrine (whether legal or not, it is still as operative as ever because folks believe that young kids need mothers). Under the time-shift shared parenting order, we are more than happy to go along with the tender years doctrine (or any other sort of belief), for we will get our turn too under the time-shift order.
3. The time-shift order will tend to move custody of children to fathers when kids get into their teen years, are harder to displine, when the male figure is most important for bringing them to maturity, keeping daughers from getting preganant, and sons from getting into trouble.
4. By building it into the order, both parents are aware of what will happen in the future, and can plan for it accordingly.
5. By building it into the order, kids know what will happen and will be ready for the transition when it comes. There is plenty of time to socialize the change, so they are ready for it.
6. If one parent moves away, it still does not affect the change of custody. It might be painful for the kids and the noncustodial parent, but the other parent will get his turn.
7. The time-shift order avoids all possible feminist objections and judicial concerns about parents "getting along". They simply don't have a leg to stand on opposing it.
8. It will greatly reduce perennial litigation. We simply won't have millions of pariahed responsible parents who have no avenue but to keep litigating the custody order.
9. When the wide body of studies are invoked showing how fathers are better at dealing with teenagers, and saving society from the wide variety of social problems, we have a real political and judicial selling point.
10. In order for this to work across state or jurisdictional lines, the original decree must state that if a modification of custody is filed in another jurisdiction, the terms of the custody modification must be made according to the custody statute of the original court of jurisdiction. This prevents parents from moving to antifamily states and filing for a custody modification. It is common in contracts to state terms of jurisdiction, and there is no reason why we can't maintain statutory jurisdiction to ensure consistency when parents move.