THE ECONOMICS OF TRADE SECRETS: EVIDENCE FROM THE ECONOMIC ESPIONAGE ACT Nicola Charlotte Searle A Thesis Submitted for the Degree of PhD at the University of St Andrews 2010 Full metadata for this item is available in Research@StAndrews:FullText at: https://research-repository.st-andrews.ac.uk/ Please use this identifier to cite or link to this item: http://hdl.handle.net/10023/1632 This item is protected by original copyright This item is licensed under a Creative Commons License     The  Economics  of  Trade  Secrets:  Evidence  from  the   Economic  Espionage  Act       Nicola  Charlotte  Searle     Submitted  for  the  degree  of     Doctor  of  Philosophy  (Economics)   at  the  University  of  St  Andrews     November  2,  2010        Candidate’s  declarations:     I,  Nicola  Charlotte  Searle,  hereby  certify  that  this  thesis,  which  is  approximately   73,500   words   in   length,   has   been   written   by   me,   that   it   is   the   record   of   work   carried   out   by   me   and   that   it   has   not   been   submitted   in   any   previous   application   for  a  higher  degree       I   was   admitted   as   a   research   student   in   October,   2006   and   as   a   candidate   for   the   degree   of   PhD   in   Economics   in   October,   2007   the   higher   study   for   which   this   is   a   record  was  carried  out  in  the  University  of  St  Andrews  between  2006  and  2010       Date   02/11/10     signature  of  candidate        Supervisor’s  declaration:     I  hereby  certify  that  the  candidate  has  fulfilled  the  conditions  of  the  Resolution   and   Regulations   appropriate   for   the   degree   of   PhD   in   Economics   in   the   University   of   St   Andrews   and   that   the   candidate   is   qualified   to   submit   this   thesis   in  application  for  that  degree       Date   02/11/10     signature  of  supervisor      Permission  for  electronic  publication:       In  submitting  this  thesis  to  the  University  of  St  Andrews  I  understand  that  I  am   giving   permission   for   it   to   be   made   available   for   use   in   accordance   with   the   regulations   of   the   University   Library   for   the   time   being   in   force,   subject   to   any   copyright  vested  in  the  work  not  being  affected  thereby     I  also  understand  that   the  title  and  the  abstract  will  be  published,  and  that  a  copy  of  the  work  may  be   made   and   supplied   to   any   bona   fide   library   or   research   worker,   that   my   thesis   will   be   electronically   accessible   for   personal   or   research   use   unless   exempt   by   award   of   an   embargo   as   requested   below,   and   that   the   library   has   the   right   to   migrate   my   thesis   into   new   electronic   forms   as   required   to   ensure   continued     ii   access  to  the  thesis  I  have  obtained  any  third-­‐party  copyright  permissions  that   may  be  required  in  order  to  allow  such  access  and  migration,  or  have  requested   the  appropriate  embargo  below       The   following   is   an   agreed   request   by   candidate   and   supervisor   regarding   the   electronic   publication   of   this   thesis:   Access   to   printed   copy   and   electronic   publication  of  thesis  through  the  University  of  St  Andrews     Date    02/11/10     signature  of  candidate           signature  of  supervisor   iii     Acknowledgements     Research  Support     I  gratefully  acknowledge  the  Horowitz  Foundation  for  their  generous  support  of   my  final  year  of  doctoral  research    I  would  also  like  to  thank  the  Russell  Trust   for  supporting  my  summer  training    Finally,  I  would  like  to  thank  the  Centre  for   Research  in  Industry,  Enterprise,  Finance  and  the  Firm  (CRIEFF)  at  the  School  of   Economics  &  Finance  at  the  University  of  St  Andrews  for  their  scholarship  which   made  the  first  three  years  of  my  doctoral  research  possible     Personal       I  would  like  to  thank  my  family  for  supporting  and  encouraging  my  education  I   would  also  like  to  thank  my  fellow  PhD  student  and  partner,  Kayhan  Jafar-­‐ Shaghaghi,  for  his  support  of  my  studies  and  calming  presence    My  sister,   Catherine  Searle,  has  been  both  supportive  and  provided  an  inter-­‐disciplinary   perspective  on  doctoral  studies  Special  thanks  go  to  my  namesake,  Nicola   Healey,  for  her  diligent  proofing  (any  errors  are  entirely  my  own)  and  friendship     Academic     First  and  foremost,  I  would  like  to  thank  my  supervisor,  Professor  Gavin  C  Reid     Professor  Reid  has  dedicated  countless  hours  to  the  supervision  of  this  thesis       All  aspects  of  my  research  have  benefited  greatly  from  Professor  Reid’s  attention   to  detail,  mentoring  and  career  guidance    I  have  been  a  student  of  Professor  Reid   for  five  years,  initially  as  a  taught  postgraduate  and  now  as  a  research  student,   and  his  teaching  and  supervision  capacities  have  shaped  my  approach  to   Economics  and  research  profoundly    Professor  Reid  has  been  a  patient  and   exceptional  supervisor  and  I  hope  that  this  thesis  reflects  his  invaluable  input       iv   I  would  also  like  to  thank  Professor  David  Ulph,  my  second  supervisor,  for  his   valuable  input  on  policies  and  advice    Furthermore,  I  would  like  to  thank  the   School  of  Economics  &  Finance  for  being  an  exceptionally  supportive   department    Comments  during  PhD  presentations,  Brown  Bags  and  informal   conversations  have  all  helped  shaped  the  direction  of  my  thesis    Caroline  Moore,   Angela  Hodge  and  Eliana  Wilson  together  ensure  the  smooth  running  of  the   school    My  fellow  PhD  students  have  been  a  constant  source  of  support  and   problem  solving    Academic  support  has  come  from  many  sources  within  the   school  and  special  thanks  go  to  Professor  Felix  Fitzroy,  Dr  Arnab  Bhattacharjee   and  Dr  Geethanjali  Selvaretnam     Outwith  St  Andrews,  I  would  like  to  thank  Connie  Luthy  and  the  Southern   Methodist  University  Cox  School  of  Business  for  hosting  me  during  my  research   trip  to  Dallas,  Texas    My  thanks  also  go  to  Dr  Jeffery  Dubin,  of  the  University  of   California  at  Santa  Barbara,  for  his  feedback  on  my  empirical  work    I  would   particularly  like  to  acknowledge  the  tutoring  of  Professor  Laurent  Maderieux   and  his  colleagues  at  the  2007  Transatlantic  Summer  IP  Academy  with  Bocconi   University  and  the  European  Patent  Office    The  Academy  provided  me  with  a   much-­‐needed  foundation  in  law    I  also  gratefully  acknowledge  the  fantastic   tutoring  at  the  World  Intellectual  Property  Organization  (WIPO)  Summer  School   and  my  excellent  group  of  classmates     Finally,  I  would  like  to  dedicate  this  thesis  to  my  grandmother,  Olwen  Searle,   who,  “should  have  been  a  doctor.”         v   Abstract     This  thesis  reports  on  the  economic  analysis  of  trade  secrets  via  data  collected   from  prosecutions  under  the  U.S  Economic  Espionage  Act  (EEA.)    Ratified  in   1996,  the  EEA  increases  protection  for  trade  secrets  by  criminalizing  the  theft  of   trade  secrets    The  empirical  basis  of  the  thesis  is  a  unique  database  constructed   using  EEA  prosecutions  from  1996  to  2008    A  critical  and  empirical  analysis  of   these  cases  provides  insight  into  the  use  of  trade  secrets     The  increase  in  the  criminal  culpability  of  trade  secret  theft  has  important   impacts  on  the  use  of  trade  secrets  and  the  incentives  for  would-­‐be  thieves    A   statistical  analysis  of  the  EEA  data  suggest  that  trade  secrets  are  used  primarily   in  manufacturing  and  construction    A  cluster  analysis  suggests  three  broad   categories  of  EEA  cases  based  on  the  type  of  trade  secret  and  the  sector  of  the   owner    A  series  of  illustrative  case  studies  demonstrates  these  clusters     A  critical  analysis  of  the  damages  valuations  methods  in  trade  secrets  cases   demonstrates  the  highly  variable  estimates  of  trade  secrets    Given  the  criminal   context  of  EEA  cases,  these  valuation  methods  play  an  important  role  in   sentencing  and  affect  the  incentives  of  the  owners  of  trade  secrets    The  analysis   of  the  lognormal  distribution  of  the  observed  values  is  furthered  by  a  statistical   analysis  of  the  EEA  valuations,  which  suggests  that  the  methods  can  result  in   very  different  estimates  for  the  same  trade  secret         A  regression  analysis  examines  the  determinants  of  trade  secret  intensity  at  the   firm  level    This  econometric  analysis  suggests  that  trade  secret  intensity  is   negatively  related  to  firm  size  Collectively,  this  thesis  presents  an  empirical   analysis  of  trade  secrets     Key  Words:  Intellectual  Property,  Trade  Secrets,  Economic  Espionage  Act,   Damages,  and  Firm  Size     vi   Table  of  Contents   CHAPTER  1   :  INTRODUCTION   1.1   Terms  of  Discussion   1.1.1   Intellectual  Property .2   1.1.2   Patents   1.1.3   Trade  Secrets   1.2   Intellectual  Property  and  Innovation   1.3   The  Economic  Espionage  Act  (EEA)   1.4   Outline  of  Thesis   1.4.1   Acronyms 12   CHAPTER  2   :  LITERATURE  REVIEW:  PATENT  AND  TRADE  SECRET   LITERATURE   13   2.1   Introduction 13   2.2   The  Study  of  Intellectual  Property .15   2.2.1   Basis  of  Intellectual  Property  Systems 15   2.2.2   The  Study  of  Intellectual  Property 20   2.3   Policy  Issues  and  Current  Debates 24   2.3.1   Current  Policy  Debates:  Patents 24   2.3.2   Current  Policy  Debates:  Trade  Secrets 31   2.4   Theoretical  Models  in  Economics 34   2.4.1   Patents 34   2.4.2   Economic  Models  of  Trade  Secrets 40   2.5   Empirical  Analyses 53   2.5.1   Patents 53   2.5.2   Empirical  Analysis:  Trade  Secrets 56   2.6   Conclusion 63   CHAPTER  3   :  THE  THEFT  OF  TRADE  SECRETS:  EVIDENCE  FROM  THE   ECONOMIC  ESPIONAGE  ACT  OF  1996 66   3.1   Introduction 66   3.2   Trade  Secrets 69   3.3   The  Economic  Espionage  Act  (EEA)  of  1996 .70   3.3.1   The  Property  Liability  Debate 74   3.4   Database  Construction 78   3.4.1   Data  Issues 79   3.5   Composition  of  Defendants 83   3.5.1   Relationship  to  the  Victim 83   3.5.2   Nationality 84     vii   3.6   Composition  of  Victims  and  Stolen  Trade  Secrets 85   3.6.1   Industrial  Sectors  of  Victims 85   3.6.2   Subject  Matter  of  Stolen  Trade  Secrets 89   3.7   Criminalization  and  Detection 90   3.7.1   Comparison  to  Civil  Actions 90   3.7.2   Defendants’  Costs  and  Benefits 95   3.7.3   Detection  and  Reporting:  The  Impact  on  Victims 97   3.8   Cluster  Analysis 100   3.9   Illustrative  Case  Studies 104   3.9.1   Cluster  1:  Copyrightable  Trade  Secrets  in  Service  and  Retail  Sectors 105   3.9.2   Cluster  2:  Potentially  Patentable  Trade  Secrets  in  Manufacturing  and  Construction 110   3.9.3   Cluster  3:  Cases  Where  Neither  Patents  nor  Copyright  Were  Available 115   3.10   Conclusion 121   3.11   Appendix:  Use  of  PACER  in  Database  Construction 124   3.11.1   Example  Docket  from  PACER:  USA  v  Dorn 126   3.11.2   Sample  Document  from  PACER:  Wu  Indictment .129   3.11.3   EEA  Database  Extract 131   CHAPTER  4   :  DAMAGES  VALUATIONS  OF  TRADE  SECRETS 132   4.1   Introduction 132   4.2   Key  Concepts  in  Damages  Valuations  of  Trade  Secrets 136   4.3   Overview  of  Literature 138   4.4   Principles  of  Damages  and  Dispute  Valuation 145   4.4.1   Damages  in  Tort  and  Contract  Law 147   4.5   Value  Estimation  of  Trade  Secrets  in  EEA  Cases:  Survey  of  Models  and  Methods 151   4.5.1   Points  of  View  and  Concepts  of  Time  in  the  Models 153   4.6   Income  Models 156   4.6.1   Key  Components  of  Income  Models 157   4.6.2   Unjust  Enrichment 159   4.6.3   Lost  Profits 166   4.6.4   Reasonable  Royalty 174   4.7   Cost  Models 180   4.7.1   Research  &  Development .180   4.7.2   Replacement  Costs 182   4.7.3   Actual  Damages 184   4.8   Market  Models 185   4.8.1   Market  Guideline  Transaction  Approach 186   4.8.2   Market  Models:  Fair  Market  Value  Example 187   4.9   Discussion  of  Relative  Merits  of  Models 187   4.10   Conclusion 189     viii   CHAPTER  5   :  STATISTICAL  EVIDENCE  AND  ANALYSIS  OF  EEA  VALUATIONS   191   5.1   Introduction 191   5.2   Distribution  of  the  Value  of  Trade  Secrets 191   5.2.1   Discussion  of  the  Lognormal  Distribution  of  the  Value  of  Trade  Secrets 197   5.3   Analysis  of  the  Value  of  Trade  Secrets  Based  on  the  Valuation  Method 198   5.4   Comments  on  the  use  of  Models 202   5.4.1   Income  Models 202   5.4.2   Cost  Models 204   5.4.3   Market  Models 204   5.5   Statistical  Analysis:  The  Range  of  Estimates 205   5.5.1   Statistical  Analysis:  Cross  Referencing  Method 208   5.5.2   Implications  of  Difference  in  Means  in  the  Valuations 215   5.6   Conclusion 217   CHAPTER  6   :  THE  DETERMINANTS  OF  TRADE  SECRET  INTENSITY:     EVIDENCE  FROM  THE  ECONOMIC  ESPIONAGE  ACT 218   6.1   Introduction 218   6.2   Theoretical  Overview 218   6.3   Specification  of  the  Model 223   6.4   Data 225   6.4.1   Continuous  Variables .227   6.4.2   Dummy  Variables .227   6.5   Functional  Form 230   6.6   Analysis  and  Results 235   6.6.1   Consideration  of  Other  Independent  Variables 237   6.6.2   Consideration  of  Other  Functional  Forms .240   6.6.3   Test  for  Heteroskedascity 241   6.7   Robustness  testing 242   6.7.1   Outliers 243   6.7.2   Robust  Regression 243   6.7.3   Trimming:  Kernel  Density 244   6.7.4   Trimming:  Trimmed  Least  Squares 246   6.8   Endogenous  Switching:  Endogenous  Selection  and  Sample  Selectivity 247   6.8.1   Heckman  Correction .248   6.8.2   Truncated  Regression 249   6.9   The  Role  of  Outsiders 251   6.10   Elasticity  Analysis  Across  Multiple  Forms 254   6.11   Conclusion 257     ix       301         302       303       304       305       306       307       308       309       310       311       312       313       314             315