Unformatted text preview: Soc 120: Marriage and the Family
Lecture 21: Divorce Effects Prof. Elwert Review Divorce II Explanations Micro things that individuals do that can affect divorce Age at marriage Parental divorce Fire alarm... 1 Preview Micro Explanations (cont'd) Consequences of divorce for spouses Consequences of divorce for children Short term long term Divorce vs parental conflict Selection or causation? Estimating divorce effects A short methodological history Cohabitors are: 1) less sure about future of relaHonship (test the waters) 2) less commiTed to the relaHonship 3) more independent gender role ideologies (more liberal than those who get married straight away) 4) less religious (parHcipate less in religious acHviHes) 5) divorced parents (more careful about jumping into marriage) people who don't they've been cohabiHng might have obtained a divorce had they married straight Micro Explanations II Cohabitation Strong predictor of divorce Those who cohabit have a higher risk of divorce 50% But likely pure selection, not causation did cohabitaHon cause divorce? Cohabitors are less committed to union, more associated independent gender roles, less religious participation, purely due to more likely to have divorced parents selecHon Cohabitors more likely to divorce regardless of cohabitation doesn't maTer if they cohabited or not
get married because But may actually decrease divorce (Elwert) Trial marriage: "bad matches weeded out in cohabitation, better matches sort into marriage. cohabitaHon does Cohabitation prevents divorces by preventing prevent divorces--> formation of potentially doomed marriages one has more Hme No effect for cohabitors who do marry their partner
to decide if they married really want to get Trial marriage: people split up who realize it wasn't meant to be REMEMBER: 1) cohabitaHon predicts divorce 2) associaHon simply due to selecHon and once we subtract out selecHon factor--> no associaHon remains 3) cohabitaHon actually decreases the risk of divorce by prevenHng marriage in the first place 2 Divorce as a Process
long process: marital conflict, separatoin, legal divorce owned stuff together--> had a shared residence and now must split up (high conflict) Distinguish between marital conflict, separation, and legal divorce (Amato 2000) usually takes Conflict/unhappiness often prolonged period awhile Separation physically moving apart many call it trial separaHon--> hold out in hope something will change Divorce legal separaHon Legal separation of assets Legal and physical arrangements for child custody De facto custody still mostly goes to mothers usually rests Custodial parents does Coordination of parental efforts with mother Noncustodial parent (hopefully) pays child support day to day work Co-parenting (coordinate/cooperate) vs. parallel parenting (operate separately)
co-parenHng is not easy so usually parallel parenHng occurs where kid has 2 independent parents mixed message parenHng is harmful to child Consequences for Spouses Crisis period both spouses experience crisis aFer divorce
~ 1-2 years of intense emotional upheaval, ambiguous attachment for both spouses
coping strategies: 1) small house is less expensive than large house (same with different neighborhood) 2) move with parents because they can't support themselves and child independently 3) reenter labor force in rush (some when marriage is ending) Consequences for women Sharp decrease in average economic well-being
money that one spouse pays to the other to make up for the loss in economics (children not included) ~30% decline in standard of living (composition adjusted) Coping strategies: residential moves,living with parents, reenter LF Stress and depression as adjust to being worker and primary parent. Alimony (increasingly rare) does not happen as much Child support
child support: money from the non-custodial parent to the unit that has the child (not every child is enHtled to get child support--> only 60%) 60% if divorced mothers supposed to receive support 49% of custodial mothers get anything ($4,650 in 2001 on average) (Sorenson 2003)
over half the women who are supposed to get child support don't receive it 3 Consequences for Spouses Consequences for men Average divorced man suffers economic hardship, too though less than average divorced woman (McManus & DiPrete 2001) Average male standard of living declines 15-20%
reasons for loss of money: Loss of wife s income, economies of scale, child essenHally he is losing dependence (not losing wife's income & child support payments are not equal to payments he made during marriage) for women: one can mother and one can be a wife B UT most men only act as fathers if they are married support payments But men contributing >80% of pre-divorce income gain in relative standard of living Tight coupling of marriage and fatherhood for men Sharp decline in contact with children and change in nature of time spent together (`fun' activities outside the house rather tan daily interaction) about 17% of single However, increase in single father families (17% of single parent families in 2007 parent families in 2007), often coresident with other were run by fathers oFen relatives widowers are over-represented living with other relaHves as well Short Term Consequences for Kids
during crisis period: on average Effects differ by time horizon.kids suffer psychologically Near universal emotional turmoil during crisis period, 1-2 years (e.g. depression, self-esteem, irritability, behavioral problems) Factors Loss of a parent (by divorce) Economic loss Diminished parenting parent that is there is also adjusHng Coercive cycles Multiple transitions (new home, school, friends, parent) Transitions per se may create problems self-reinforcing vicious cycles--> parent is stressed and doesn't reason and kids fight back turning into a cycle 4 Divorce vs. Conflict Pre (and post) divorce marital conflict may harm children Divorce as a process starts off as unhappiness Kids as parent s pawns Studies find that some post-divorce problems start predivorce Consequences of divorce depend on circumstances Psychological wellbeing of children whose parents ended a severely conflictive marriage appears to improve (Amato & Booth 1997) But only ~1/3 of divorces end severely conflictive marriage (physical abuse, frequent yelling in front of children, etc)
try to claim sole custody to hurt other parent or to get physical assets of marriage not separaHon that harms kids but the process that leads kids there if there is a conflicHve marriage then it is more beneficial for it to end but only 1/3 of them do everything before refers to crisis periods Long Term Consequences for Kids
distress does not make kids "sick" long-term--> most kids become normal again Evidence for some but not widespread long-term harm Distress vs. disorder Most children resume normal emotional/behavioral development Adult mental health of adult children of divorce somewhat worse Glass half-empty / half-full 29% vs 13% of kids experiencing vs not experiencing divorce do drop out of HS 7/10 vs 9/10 of kids experiencing vs not experiencing divorce do not drop out of HS BUT divorce is not guaranteed to ruin a child's life
on average relaHvely few suffer from divorce divorce more than doubles for high school drop outs emphasizes the negaHve because we think divorce kids stuff have problems 5 Is it Causation or Selection?
kids of divorce would have Effects of divorce on children are hotly debated. Much evidence that at least some (maybe all) of longterm association between divorce and negative outcomes is due to selection, not causation Example: Divorce is associated with lower test scores for kids
parents who have low- economic status are more likely to get a divorce while simultaneously produce kids with low test scores anyway done worse anyway Selection
Divorce Parents low SES Divorce Low scores Causation Low scores never a clear either-or , differenHaHng is difficult Estimating Divorce Effects Separating causation from selection is difficult ...because controlling (i.e. accounting) for selection is difficult Example: Fact: children of divorce have lower educational test scores than kids from two-bio-parents homes. But maybe that s because divorcing families are a priori of lower education, ability, motivation, etc.? How could we possibly control for all possible selection factors? Preview Next slides: intuitive introduction to pretty advanced strategies See my essay on causal inference on the course website (text accompanying remaining slides) Read the genetic studies in Cherlin (p.400-1) and understand how they are very similar to Geronimus and Korenmann (1992) 6 Defining Causal Effect Definition: A causal effect is the difference between the potential outcome under treatment (i.e. divorce) and the outcome under control (i.e. no divorce) We would like to compare the same kid s test scores if parents divorced and if they didn t. to observe we have to If a kid s test scores are lower if parents divorce than if compare what is to parents would not divorce then we say that divorce caused a what isn't reduction in test scores. Problem: We can t observe both potential outcomes at the same time because parents either divorce or they don t! Known as the fundamental problem of causal inference (Holland 1986) Kids in treatment group may differ systematically from kids in control groups due to selection kids who have experience divorce differ from kids who have not experienced divorce even B EFORE divorce problem with staHsHcs: can't have a parent with divorced and not divorced at the same Hme Two Solutions
what is not controlled: -area: kids of divorce live in different places that kids who are not experiencing divorce Two strategies to control for selection: 1. Compare treatment and control kids that resemble each other in every observed respect Comparing apples to apples : i.e., compare divorce kids with high inc, high educ parents to non-divorce kids with high inc, high educ parents, etc. This controls for observed selection. only aspect that is different between children is one has divorced parents and one doesn't 2. Compare kids to themselves before and after divorce Cases as their own controls, fixed effects : i.e. inspect kid s test score trajectory over time Controls for selection due to unobserved factors specific to the child and his/her family 7 Adjusted Group Comparison We could compare the test scores of children who did/didn t experience divorce and who resemble each other in observed characteristics (e.g. parental income, education).apples to apples Obvious problem: Problem of unobserved selection (not captured by observed characteristics) Here, divorced kids have lower test scores. Simple Pre-Post Comparison Compare kids test scores before and after divorce kids that experience divorce are worse off anyway Here, it looks like all of the post-divorce difference already existed pre-divorce => difference appears to be due to selection 8 Problems of Pre-Post Tests Recall: divorce is a process, not a time point. Maybe, marital conflict in the run up to the divorce harmed kids such that the simple pre-post test masks the harm done by divorce?
no vary over Hme suggests that the low test scores aFer divorce seem to exist before--> divorce does not harm test scores Here, it could be that the process of divorce created postdivorce differences in test scores. Long Term Trajectories Aughinbaugh et al (2005) control (1) for many observed characteristics and (2) also look at test score trajectories 5 years pre and post divorce. They find no statistically significant evidence that the trajectories of kids from divorced and intact families vary over time. Test score differences appear to be pre-existing => due to selection. 9 ...
View Full Document
This note was uploaded on 03/21/2012 for the course NUTR SCI 132 taught by Professor Anderson during the Fall '09 term at University of Wisconsin.
- Fall '09