concurring in result.
I write separately to express my concern about — and to distance my vote to affirm from — that portion of the majority opinion which concludes that because of “[t]he parents’ agreement that there was a change of circumstances that required or necessitated a change in custody” the agreement “freed the judge from the constraints of the two-step analysis ordinarily required in modification proceedings and allowed him to treat this case as an original custody determination.”
Parents are at liberty to decide custody between them without intervention of the courts. However, when they cannot agree and petition the courts to decide custody, our test is whether the change in circumstances is a change that requires or necessitates a change in custody. I have urged adherence to the two-step analysis. Eg., Johnson v. Johnson, 480 N.W.2d 433, 438 (N.D.1992) [VandeWalle, J., concurring specially]; Anderson v. Anderson, 448 N.W.2d 181, 183 (N.D.1989) [VandeWalle, J., concurring specially]. When the change in circumstance must be one which so adversely affects the child so as to “compel or require a change in custody to foster the best interests of the child,” Johnson v. Schlotman, 502 N.W.2d 831, 834 (N.D.1993), I do not understand how the judge could not examine that change nor do I agree that it would authorize the trial court to treat the case as an original custody determination. We do not allow the parents’ stipulation of child support to free the judge from the constraints of the child support guidelines. E.g., State of Minn. v. Snell, 493 N.W.2d 656 (N.D.1992).
The lack of permanence because of the constant shifting of the child since the date of the original divorce is obviously due to Patricia’s life style for, as the trial court found, *691“Her jobs since separation all lasted less than a year. Likewise, she has changed homes and cities several times since the separation. She has established no stable living arrangements of long duration.” Roland’s life style was in stark contrast. It appears it was so permanent as to be “rigid” and that the trial court’s characterization of Roland as “being a stickler for detail and very rigid in the ‘rightness’ of his ways,” as compared to the trial court’s characterization of Patricia as being “best able to give love and affection without jeopardizing the relationship with the non-eustodial parent” is the keystone for the trial court’s decision to award custody to Patricia.
Notwithstanding the finding that “neither parent had any permanence as a family unit,” I believe the best interests of the child are better served by adhering to the two-step analysis rather than permitting the trial court to proceed as if this were an original custody determination. The trial court did make findings on changed circumstances and found that the prior judgments were unworkable and not being complied with. The issue is whether the changes in circumstances require a change in custody. I realize that this makes it more difficult to affirm the trial court’s decision in this matter. After examining the evidence record, I would have awarded custody to Roland for the greater part of the year and to Patricia during the summer. However, what I would have done at the trial level is not the standard which guides our functions as appellate judges.
In this case, our review is governed by the “clearly erroneous” standard of Rule 52(a), NDRCivP. That is a standard deferential to the decision of the trial court. I have recently indicated that in cases of child custody “in which there is not clearly a ‘right’ or a “wrong’ solution,” “this court should not attempt to substitute its judgment for that of the trial court, even if we disagree with the result_” Barstad v. Barstad, 499 N.W.2d 584, 589 (N.D.1993) [VandeWalle, C.J., dissenting]. I have also expressed by belief that a continued pattern of circumvention of visitation may justify a change of custody in the best interests of the child. Blotske v. Leidholm, 487 N.W.2d 607, 611 (N.D.1992) [VandeWalle, J., concurring specially]. That applies to other patterns which jeopardize the relationship with the non-custodial parent.
I agree the trial court should be affirmed.
MESCHKE, J., concurs.