Transcript
Lars Højsgaard Andersen
UNIVERSIT Y OF COPENHAGEN DEPARTMENT OF SOCIOLOGY PHD THESIS 2015 · ISBN 978-87-7611-913-3
university of copenhagen f a c u lt y o f s o c i a l s c i e n c e s
LARS HØJSGAARD ANDERSEN CONSEQUENCES OF PUNISHMENT IN DENMARK. Five Empirical Essays
CONSEQUENCES OF PUNISHMENT IN DENMARK. Five Empirical Essays
CONSEQUENCES OF PUNISHMENT IN DENMARK Five Empirical Essays
PhD Dissertation Lars Højsgaard Andersen
Lars H. Andersen SOCIO omslag 15 mm.indd 1
01/07/15 09.45
UNIVERSITY OF COPENHAGEN PHD DISSERTATION
CONSEQUENCES OF PUNISHMENT IN DENMARK Five Empirical Essays
Author Lars Højsgaard Andersen Supervisor Mads Meier Jæger
Co-Supervisor Signe Hald Andersen
01 / 30 / 2015
DEPARTMENT OF SOCIOLOGY
PhD Dissertation 2015 © Lars Højsgaard Andersen ISBN 978-87-7611-913-3 Printed by SL grafik, Frederiksberg, Denmark (www.slgrafik.dk)
_________________________________________________________________________
i
SUMMARY Millions of people throughout the world experience punishment each day, and, for example, more than 10 million people were incarcerated on any given day in 2013. Despite these high figures, which should draw in the attention of social scientists, sociology and criminology still have many important but unanswered questions about the consequences of punishment. This dissertation uses econometric methods and register data to provide new empirical evidence on the individual consequences of punishment in Denmark, thereby providing empirical answers to some of these previously unanswered questions. The dissertation consists of an introduction and five empirical essays. The introduction outlines a general framework for analyzing the consequences of punishment, and draws out general implications from the results presented in the dissertation. The first empirical essay analyzes the effect of age at first imprisonment on the Age-Crime Curve. The main result is that being younger at first imprisonment increases the rate of criminal convictions in the short run, but it does not change the overall shape of the Age-Crime Curve. The second essay analyzes the effect of serving a prison sentence at home under electronic monitoring on social welfare benefit take-up. The main result is that, relative to imprisonment, electronic monitoring improves young offenders’ labor market outcomes, and, moreover, it does not leave older offenders worse off. Essays three and four analyze the role of probation and parole officers in shaping the outcomes of probationers and parolees. The third essay decomposes the total variance in criminal recidivism into components attributable to offenders and officers, respectively. The main result is that, in addition to offenders, characteristics of officers account for a non-trivial share of the total variance in rates of
ii _________________________________________________________________________________
recidivism. The fourth essay exploits random assignment of officers to probationers and parolees in Copenhagen, Denmark to identify the effect of officer assignment on labor market outcomes and recidivism. The main result is that, although to a lesser degree than suggested in existing research, officer assignment affects labor market outcomes and recidivism. Finally, the fifth essay analyzes whether having a spouse who has previously committed crime moderates the crime-reducing function of marriage among previously convicted men. Empirical results suggest that this is indeed the case.
_________________________________________________________________________
iii
RESUME En hvilken som helst dag i 2013 sad over 10 millioner mennesker indespærret rundt omkring i verden. Hertil skal man lægge de millioner af mennesker, som oplever andre former for straf, fx alternativer til frihedsstraf, såsom elektronisk fodlænke eller tilsyn. Men selvom så mange mennesker straffes, og selvom alle grene af socialvidenskaberne har kastet sig over emnet, er der faktisk bemærkelsesværdigt mange spørgsmål om de individuelle konsekvenser af straf, som vi stadig mangler at besvare. Denne afhandling, som omhandler de individuelle konsekvenser ved straf, giver empiriske svar på nogle af disse ubesvarede spørgsmål ved at anvende statistiske værktøjer på danske registerdata. Afhandlingen består af en introduktion og fem empiriske essays. Introduktionen beskriver afhandlingens overordnede tilgang til at analysere konsekvenserne af straf og sammenfatter de generelle implikationer af resultaterne i de fem empiriske essays. Det første empiriske essay analyserer konsekvenserne af tidligere indespærring for unge voldsmænd. Hovedkonklusionen er, at tidligere indespærring medfører mere kriminalitet blandt disse unge mænd inden for en kort årrække efter løsladelsen. Det andet empiriske essay analyserer om personer, der afsoner en fængselsstraf i hjemmet med en elektronisk fodlænke, får en bedre arbejdsmarkedstilknytning efter straffen, end sammenlignelige personer, der afsoner deres straf i fængsel. Hovedkonklusionen er, at relativt til et fængselsophold, så klarer unge fodlænkeafsonere sig bedre på arbejdsmarkedet efter, at straffen er udstået. Denne effekt kan dog ikke findes blandt ældre afsonere, for hvem fodlænkeafsoning dog ikke er mere skadeligt end fængsel.
iv _________________________________________________________________________________
Det trejde og fjerde essay handler om, hvilken betydning tilsynsførende har for deres klienter, som har fået en betinget dom. Det tredje essay opdeler variationen i tilbagefald til kriminalitet blandt klienterne i én delmængde, der kan tilskrives omstændigheder ved klienten selv, og én delmængde, der kan tilskrives omstændigheder ved den tilsynsførende. Hovedkonlusionen er, at ud over, at en stor del af klienternes tilbagefald til kriminalitet kan forklares af forhold ved klienterne selv, så kan en betragtelig del af tilbagefaldet til kriminalitet også forklares af forhold ved den tilsynsførende. Det fjerde essay analyserer, hvad det betyder for de betinget dømte, hvilken tilsynsførende de tilses af. Hovedkonklusionen er, at det faktisk betyder noget for de betinget dømtes arbejdsmarkedstilknytning og tilbagefald til kriminalitet, hvilken tilsynsførende de tilses af, men at betydningen er noget mindre, end den eksisterende forskning på området fremstiller det. Det femte og sidste essay analyserer, om den kriminalitetssænkende effekt af ægteskab, som literaturen gentagne gange har analyseret, også findes for tidligere dømte mænd, samt om den eventuelt afhænger af partnerens egen kriminalitetshistorie. Hovedkonklusionen er, at begge dele er tilfældet.
_________________________________________________________________________
v
ACKNOWLEDGMENTS Never write ‘‘it goes without saying’’. Why would you possibly want to waste words telling your reader something which there is no need to tell? This was one of the first things I was taught when I began my PhD studies back in 2012, and it is still one of those things that I keep in mind as I write academic papers today. But today, following 2½ years of dissertation work, my first academic publications, living for a while in the United States, getting married and having the joy of one more paternity leave, saying the stuff that goes without saying seems like the most important part of this dissertation. I have a lot to say that goes without saying, because so many people have offered me help, advice, and support during the process of writing this dissertation. First, I thank my supervisors, Mads Meier Jæger and Signe Hald Andersen, for excellent supervision and guidance. Second, I thank the Rockwool Foundation Research Unit for funding my PhD studies, and I especially thank Torben Tranæs for always believing in my ideas and abilities. Third, I thank Peer Skov, Peter Fallesen, and Rasmus Landersø, my friends and colleagues, for making each day of my PhD studies quite a bit more entertaining, and for helping me out with methodological headaches (and for inducing other fun-related headaches). Fourth, I thank Britta Kyvsgaard at the Danish Ministry of Justice for, back in 2005, directing my inexperienced research attention to the field of criminology, and for continuously being a great mentor. And speaking of mentors, a special thank goes out to Christopher Wildeman. Chris invited me over to Yale University for a semester; we collaborated (and still collaborate) on
vi _________________________________________________________________________________
research papers; he taught (and still teaches) me how to write academic papers and give presentations; and we had (and still have) great fun. Most importantly, though, he restored in me the urge to tell good sociological stories, for which I am very grateful. Also, during our stay in New Haven, Chris and his family took great care of my family and me, and I thank Chris, Cilla, and the kids for that. I would like to thank a few persons at the Danish Prison and Probation Service for providing me with access to those incredibly good data on which this dissertation is based, and for helping me understand those data. Every empirical researcher ought to have a data pusher like Hans Graunbøl who, whenever I call him with strange and unclear data inquiries, keeps his calm, explains how registration processes take place regarding that particular topic and how this matters for the nature of the data, and provides me with access to data. Anita Rönneling, Anne Bergstedt, Estrid Pass, Pernille Poulsen, and Pernille Lape spend hours answering my inquiries about nitty-witty details of probation and parole. I thank you for this. In addition to work-related acknowledgments, I thank friends and family. I thank my parents for always leading me down the right roads, and I thank my friends, especially Anders, Rune, Calle, and Krølle, for making those roads so much fun. I also thank Sidse’s family for support and for helping out with the kids. The acknowledgment that goes the most without saying, and therefore is the most important one, concerns my beautiful wife Sidse and our two lovely kids, Karl-Louis and Marie. Der findes ikke ord til at beskrive min taknemmelighed, I er intet mindre end formidable. Selv når børnene er syge, forældrene forjagede, og arbejdet både dragende og krævende, er det hjemme hos jer, det hele giver mening.
_________________________________________________________________________
vii
Last, I am grateful for receiving travel grants from (in Danish): Augustinus Fonden, Christian og Ottilia Brorsons Rejselegat for yngre videnskabsmænd og –kvinder, Etly og Jørgen Stjerngrens Fond, Knud Højgaards Fond, Oticon Fonden, Rockwool Fondens Forskningsenhed, Torben and Alice Frimodts Fond. Without financial support from these grants my family and I would not have been able to visit Yale University. I feel confident that everyone I just thanked actually already knew how much I appreciate their help and support. But sometimes you need to say the things that go without saying, even if they don’t do your gratitude any justice in writing.
viii _________________________________________________________________________________
_________________________________________________________________________
ix
CONTENTS
Summary Resume
i iii
Acknowledgments
v
List of Tables
xi
List of Figures
xii
1
INTRODUCTION
2
FIRST IMPRISONMENT AND THE AGE-CRIME CURVE
27
3
THE EFFECT OF ELECTRONIC MONITORING ON SOCIAL WELFARE DEPENDENCE
63
4
DECOMPOSING RECIDIVISM VARIANCE INTO PROBATION AND PAROLE OFFICERS AND THEIR CLIENTS
115
5
MEASURING THE EFFECT OF PROBATION AND PAROLE OFFICERS ON LABOR MARKET OUTCOMES AND RECIDIVISM
147
6
THE EFFECT OF MARRIAGE AND SPOUSAL CRIMINALITY ON RECIDIVISM
189
Bibliography
1
221
