O slideshow foi denunciado.
Utilizamos seu perfil e dados de atividades no LinkedIn para personalizar e exibir anúncios mais relevantes. Altere suas preferências de anúncios quando desejar.

20160714 StrategyNZ - One-day Workshop - Simon Wakeman Powerpoint(2016)

2.058 visualizações

Publicada em

The StrategyNZ one-day workshop explored how New Zealand might better prepare and publish government department strategy documents on Thursday, 14 July 2016.

Publicada em: Governo e ONGs
  • Seja o primeiro a comentar

  • Seja a primeira pessoa a gostar disto

20160714 StrategyNZ - One-day Workshop - Simon Wakeman Powerpoint(2016)

  1. 1. Innova&on  &  the  performance  of     New  Zealand  firms   Simon  Wakeman   NZAE  conference   1  July  2016   1  
  2. 2. Low  investment  in  knowledge-­‐based  capital  may  explain  up  to  40%  of  the   produc&vity  gap  rela&ve  to  the  OECD  average   2   Source:  NZPC;  de  Serres  et  al.  (2014)  
  3. 3. Interna&onal  comparisons  show  NZ  near  top  of  OECD  rankings  in  genera&ng  science   but  lower  ranked  in  commercialising  innova&on   NZ’S  AVERAGE  RANKING  IN  TOP-­‐LEVEL  CATEGORIES  OF  OECD  STATISTICS   3   7   11   12   20   19   26   Science  base   Human  resources   Internet  use  for   innovaCon   Business  R&D  and   innovaCon   Knowledge  flows   and   commercialisaCon   Entrepreneurship   Source:  OECD  (2012).  Compara?ve  performance  of  na?onal  science  and  innova?on  systems  
  4. 4. Use  Sta&s&cs  NZ’s  Longitudinal  Business  Database  to  study  rela&onship  between   innova&on  and  produc&vity  for  New  Zealand  firms   DATA  SOURCES   Innova&on  ac&vity   - Business  Opera?ons  Survey  (BOS)   - Intellectual  Property  Office  (IPONZ)   Produc&vity  es&mates    (Fabling  &  Mare,  2015)   Financial  data   - Annual  Enterprise  Survey  (AES)   - IR10  tax  returns   Access  to  the  data  presented  was  managed  by  Sta?s?cs  New  Zealand  under  strict  micro-­‐data  access  protocols   and  in  accordance  with  the  security  and  confiden?ality  provisions  of  the  Sta?s?c  Act  1975.  These  findings  are   not  Official  Sta?s?cs.  The  opinions,  findings,  recommenda?ons,  and  conclusions  expressed  are  those  of  the   author/researcher,  not  Sta?s?cs  New  Zealand  or  the  New  Zealand  Produc?vity  Commission.   Longitudinal   Business   Database   R&D  ac&vity   - Business  R&D  Survey   - Business  Opera?ons  Survey  (BOS)   Government  assistance   - Payments  from  government   programmes  (incl.  R&D  grants)  
  5. 5. LBD  contains  a  range  of  measures  reflec&ng  different  aspects  of  innova&on   MEASURES  OF  INNOVATIVE  ACTIVITY   5   Sample  contains  responses  to  Business  Opera?ons  Survey  (2005-­‐2013).   0.0%   1.0%   2.0%   3.0%   4.0%   5.0%   6.0%   7.0%   8.0%   9.0%   10.0%   0%   5%   10%   15%   20%   25%   30%   35%   40%   45%   Engaged  in  R&D   Product  development  expenditure  (R&D/design/ markeCng/other)   Filed  patent   Registered  trademark   Any  innovaCon   Product  innovaCon   (new  to  world/NZ/firm)   Sales  from  new  products   Process  innovaCon   OrganisaConal  innovaCon   MarkeCng  innovaCon   %  total  expenditure/sales   %  firms  engaged  in  ac&vity    
  6. 6. R&D  expenditure  is  not  a  good  proxy  for  innova&on  output  in  all  cases   R&D  ACTIVITY  VS  INNOVATION  ACTIVITY                       TYPE  OF  INNOVATION  EXPENDITURE  BY  SECTOR                       6   Ever  engaged  in   innovaCon       No   Yes   Ever  engaged   in  R&D    No     27.5%   54.7%    Yes     1.5%   16.3%   Sample  contains  firms  responding  to  Business  Opera?ons  Survey  Innova?on  module  in  2007,  2009  &  2011.   Sample  contains  firms  responding  to  Business  Opera?ons  Survey  (2005-­‐2012).   0%   2%   4%   6%   8%   10%   12%   14%   Primary   Manufacturing   Services   R&D  expenditure   Design  expenditure   MarkeCng  expenditure   Other  expenditure  
  7. 7. Use  approach  developed  by  Fabling  &  Mare  (2015)  to  es&mate  firm  produc&vity   MEASURING  FIRM  PERFORMANCE   7   Labour   (L)   Raw  materials/   inputs  (M)   Capital  (K)   ProducCon   Output   (GO)   ( ) ( ) ( ) ( )ln   ln ln ln   where   , { , , } r r s it j r it rs it i r s r t i r G O X X X r s L K M α β δ γ ε ≠ += + + + ∈ ∑ ∑∑ ( )∑ ( ) ( ) ( )MFP: Empl  ln( oyment:   Valued  added:   Labour  productivity  ln ln ln ln where   , { , , } :   ) it it it it it it r r s it it r it rs it it rr s r GO GO MFP GO X X X r s L K M L M M L β δ ≠ − ∈ − − − = ∑∑ ∑
  8. 8. 0 20 40 60 80 % dif Employment Value-added output Labour productivity MFP output type R&D activity Product innovation new to world Product innovation Process innovation Organisational innovation Marketing innovation RELATIVE  OUTPUT  LEVEL  INNOVATORS  VS.  NON-­‐INNOVATORS  BY  INNOVATION  MEASURE   On  average  innova&ng  firms  larger  but  less  produc&ve  than  non-­‐innova&ng  firms  in  year   in  which  innova&on  occurs   8   Chart  shows  difference  in  performance  measure  predicted  from  series  of  OLS  regressions  of  output  level  in  year  0  on  innova?on  in   year  0  with  controls  for  year  and  industry.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011  with   performance  measures  up  to  2012.  Observa?ons  weighted  by  BOS  sampling  weights.  
  9. 9. 0 10 20 30 % 1 2 3 4 5 6 7 8 9 10 decile of mfp Product innovation (new to world) 1 2 3 4 5 6 7 8 9 10 decile of mfp Product innovation (new to NZ) 1 2 3 4 5 6 7 8 9 10 decile of mfp Product innovation (new to the firm) 1 2 3 4 5 6 7 8 9 10 decile of mfp Process innovation 1 2 3 4 5 6 7 8 9 10 decile of mfp Organisational innovation 1 2 3 4 5 6 7 8 9 10 decile of mfp Marketing innovation PROPORTION  OF  FIRMS  ENGAGED  IN  INNOVATION  BY  DECILE  OF  PRODUCTIVITY  IN  YEAR  0   Most  produc&ve  firms  least  likely  to  innovate   9   Chart  shows  likelihood  of  firm  innova?on  predicted  from  probit  regression  of  innova?on  ac?vity  in  year  0  on  produc?vity  decile  in   year  0.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011.  Observa?ons  weighted  by  BOS  sampling   weights  
  10. 10. Use  difference-­‐in-­‐differences  approach  to  measure  impact  of  innova&on  on   produc&vity   EMPIRICAL  APPROACH   •  Regress  change  in  firm  output/producCvity  (lnYin  -­‐  lnYi0)  on  innovaCon  in  year  0  (I0)   10   0 0 0ln –ln              in i I i X iY Y I X tα β β ε= + + + + •  Compare  within-­‐firm  differences  in  output   ! controls  for  unobserved  firm  characterisCcs   affecCng  output  levels   •  Control  for  observed  firm  characterisCcs  (age,   size,  industry,  etc.)   ! controls  for  observed  firm  characterisCcs   affecCng  changes  in  output   •  For  MFP,  use  2-­‐year  moving  average   ! accounts  for  measurement  error   •  Weight  observaCons  by  BOS  sampling   weights  Cmes  predicted  gross  output   ! quanCCes  corresponds  to  aggregate   economic  output/producCvity   •  Do  not  instrument  for  innovaCon  (cf  Crepon,   Duguet  &  Mairesse,  1998)      
  11. 11. -6 -4 -2 0 2 4% 0 1 2 3 4 years change in employment -6 -4 -2 0 2 4% 0 1 2 3 4 years change in value-added output -6 -4 -2 0 2 4% 0 1 2 3 4 years change in labour productivity -6 -4 -2 0 2 4% 0 1 2 3 4 years change in MFP Yes No -10 0 10 20% 0 1 2 3 4 years -10 0 10 20% 0 1 2 3 4 years -10 0 10 20% 0 1 2 3 4 years -10 0 10 20% 0 1 2 3 4 years RELATIVE  GROWTH  BY  OUTPUT  TYPE  OF  R&D  ACTIVE  VS  NON-­‐ACTIVE  FIRMS   Firms  engaged  in  R&D  ac&vity  have  faster  employment  and  output  growth  than   non-­‐R&D-­‐ac&ve  firms  but  produc&vity  growth  not  different   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  output  from  year  0  to  year  t  on  innova?on   in  year  0.  Regressions  include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   11  
  12. 12. -4 -2 0 2 4% 0 1 2 3 4 years change in employment -4 -2 0 2 4% 0 1 2 3 4 years change in value-added output -4 -2 0 2 4% 0 1 2 3 4 years change in labour productivity -4 -2 0 2 4% 0 1 2 3 4 years change in MFP Yes No -10 -5 0 5 10 % 0 1 2 3 4 years -10 -5 0 5 10 % 0 1 2 3 4 years -10 -5 0 5 10 % 0 1 2 3 4 years -10 -5 0 5 10 % 0 1 2 3 4 years RELATIVE  GROWTH  BY  INNOVATION  ACTIVITY  OF  INNOVATORS  VS  NON-­‐INNOVATORS   Similarly  firms  engaged  in  innova&on  grow  faster  in  size  but  not  produc&vity   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  output  from  year  0  to  year  t  on  innova?on   in  year  0.  Regressions  include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   12  
  13. 13. -6 -4 -2 0 2% 0 1 2 3 4 years Product innovation -6 -4 -2 0 2% 0 1 2 3 4 years Process innovation -6 -4 -2 0 2% 0 1 2 3 4 years Organisational innovation -6 -4 -2 0 2% 0 1 2 3 4 years Marketing innovation Yes No -4 -2 0 2 4 % 0 1 2 3 4 years -4 -2 0 2 4 % 0 1 2 3 4 years -4 -2 0 2 4 % 0 1 2 3 4 years -4 -2 0 2 4 % 0 1 2 3 4 years RELATIVE  MFP  GROWTH  BY  INNOVATION  TYPE  OF  INNOVATORS  VS  NON-­‐INNOVATORS   Only  firms  engaged  in  marke&ng  innova&on  experience  significantly  higher   produc&vity  growth   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   13  
  14. 14. -6 -4 -2 0 2 4 % 0 1 2 3 4 years 2005 -6 -4 -2 0 2 4 % 0 1 2 3 4 years 2007 -6 -4 -2 0 2 4 % 0 1 2 3 years 2009 -6 -4 -2 0 2 4 % 0 .2 .4 .6 .8 1 years 2011 Yes No -5 0 5 10% 0 1 2 3 4 years -5 0 5 10% 0 1 2 3 4 years -5 0 5 10% 0 1 2 3 years -5 0 5 10% 0 1 years RELATIVE  MFP  GROWTH  BY  YEAR  OF  PRODUCT  INNOVATORS  VS  NON-­‐INNOVATORS   Returns  to  innova&on  vary  drama&cally  by  year   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   14  
  15. 15. -6 -4 -2 0 2 4 % 2005 2006 2007 2008 2009 2010 2011 2012 year all No Yes RELATIVE  MFP  GROWTH  OVER  TIME  OF  PRODUCT  INNOVATORS  VS  NON-­‐INNOVATORS   Product  innovators  experience  higher  MFP  growth  during  economic  recovery  but   possibly  lower  growth  during  economic  downtown   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   15  
  16. 16. RELATIVE  MFP  GROWTH  BY  YEAR  OF  PROCESS  INNOVATORS  VS  NON-­‐INNOVATORS   Process  innovators  appear  not  to  do  significantly  becer  or  worse  than  non-­‐innovators   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   16   -6 -4 -2 0 2 4% 0 1 2 3 4 years 2005 -6 -4 -2 0 2 4% 0 1 2 3 4 years 2007 -6 -4 -2 0 2 4% 0 1 2 3 years 2009 -6 -4 -2 0 2 4% 0 .2 .4 .6 .8 1 years 2011 Yes No -10 -5 0 5% 0 1 2 3 4 years -10 -5 0 5% 0 1 2 3 4 years -10 -5 0 5% 0 1 2 3 years -10 -5 0 5% 0 1 years
  17. 17. -6 -4 -2 0 2 4% 2005 2006 2007 2008 2009 2010 2011 2012 year all No Yes RELATIVE  MFP  GROWTH  OVER  TIME  OF  PROCESS  INNOVATORS  VS  NON-­‐INNOVATORS   Returns  to  process  innova&on  do  not  vary  significantly  by  year   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   17  
  18. 18. RELATIVE  MFP  GROWTH  BY  YEAR  OF  ORGANISATIONAL  INNOVATORS  VS  NON-­‐INNOVATORS   Similarly  returns  to  organisa&onal  innova&on  in  2009  significantly  higher  but  not  in   other  years     Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   18   -6 -4 -2 0 2 4% 0 1 2 3 4 years 2005 -6 -4 -2 0 2 4% 0 1 2 3 4 years 2007 -6 -4 -2 0 2 4% 0 1 2 3 years 2009 -6 -4 -2 0 2 4% 0 .2 .4 .6 .8 1 years 2011 Yes No -4 -2 0 2 4 6 % 0 1 2 3 4 years -4 -2 0 2 4 6 % 0 1 2 3 4 years -4 -2 0 2 4 6 % 0 1 2 3 years -4 -2 0 2 4 6 % 0 1 years
  19. 19. -6 -4 -2 0 2 4% 2005 2006 2007 2008 2009 2010 2011 2012 year all No Yes RELATIVE  MFP  GROWTH  OVER  TIME  OF  ORGANISATIONAL  INNOVATORS  VS  NON-­‐INNOVATORS   Firms  doing  organisa&onal  innova&on  during  economic  downturn  experience   significantly  higher  returns  during  economic  recovery   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   19  
  20. 20. RELATIVE  MFP  GROWTH  BY  YEAR  OF  MARKETING  INNOVATORS  VS  NON-­‐INNOVATORS   By  contrast,  marke&ng  innovators  in  2005  experienced  lower  produc&vity  growth  but   in  2007  experienced  higher  growth   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   20   -10 -5 0 5% 0 1 2 3 4 years 2005 -10 -5 0 5% 0 1 2 3 4 years 2007 -10 -5 0 5% 0 1 2 3 years 2009 -10 -5 0 5% 0 .2 .4 .6 .8 1 years 2011 Yes No -10 -5 0 5 10 15 % 0 1 2 3 4 years -10 -5 0 5 10 15 % 0 1 2 3 4 years -10 -5 0 5 10 15 % 0 1 2 3 years -10 -5 0 5 10 15 % 0 1 years
  21. 21. -10 -5 0 5% 2005 2006 2007 2008 2009 2010 2011 2012 year all No Yes RELATIVE  MFP  GROWTH  OVER  TIME  OF  MARKETING  INNOVATORS  VS  NON-­‐INNOVATORS   Firms  that  had  introduced  marke&ng  innova&on  did  significantly  becer  during   economic  downturn   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   21  
  22. 22. -20 -10 0 10% 0 1 2 3 4 years 2005 0 1 2 3 4 years 2007 0 1 2 3 years 2009 0 .2 .4 .6 .8 1 years 2011 new to world new only to NZ new only to firm not new RELATIVE  MFP  GROWTH  BY  YEAR  OF  PRODUCT  INNOVATORS  BY  NOVELTY  OF  PRODUCT   Innova&on  followers  did  becer  than  innova&on  leaders  in  2005  &  2007  but  innova&on   leaders  did  becer  in  2009   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   22  
  23. 23. -10 -5 0 5 10% 2005 2006 2007 2008 2009 2010 2011 2012 year all new to world new only to NZ new only to firm not new RELATIVE  MFP  GROWTH  OVER  TIME  OF  PRODUCT  INNOVATORS  BY  NOVELTY  OF  PRODUCT   Innova&on  followers  performed  becer  during  economic  downtown  but  innova&on   leaders  performed  becer  during  the  recovery   Charts  show  coefficients  from  OLS  regressions  of  change  in  2-­‐yr-­‐MA  of  MFP  from  year  0  to  t  on  innova?on  in  year  0.  Regressions   include  controls  for  base  year  and  firm  characteris?cs.  Bars  show  95%  confidence  intervals.  Sample  contains  firms  in  BOS  2005-­‐2011   with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.   23  
  24. 24. -20 -10 0 10 20% 0-5 yrs 5-10 yrs 10-20 yrs 20-50 yrs 50+ yrs age R&D activity Product innovation Process innovation Organisational innovation Marketing innovation Younger  firms  experience  highest  boost  in  produc&vity  growth  following  innova&on   RELATIVE  CHANGE  IN  MFP  OVER  3  YEARS  BY  INNOVATION  TYPE  &  AGE   24   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  3-­‐yr  MA  of  MFP  from  year  0  to  year  t  on   innova?on  in  year  0  interacted  with  firm  characteris?cs  and  controls  for  base  year.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.  
  25. 25. -10 -5 0 5 10% 0-20 20-50 50-100 100+ total employment R&D activity Product innovation new to world Product innovation Process innovation Organisational innovation Marketing innovation Small-­‐to-­‐medium  size  firms  appear  to  do  worse  if  they  innovate     RELATIVE  CHANGE  IN  MFP  OVER  3  YEARS  BY  INNOVATION  TYPE  &  FIRM  SIZE   25   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  3-­‐yr  MA  of  MFP  from  year  0  to  year  t  on   innova?on  in  year  0  interacted  with  firm  characteris?cs  and  controls  for  base  year.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.  
  26. 26. -20 -10 0 10 20% Primary Manufacturing Services sector R&D activity Product innovation new to world Product innovation Process innovation Organisational innovation Marketing innovation Firms  in  manufacturing  sector  experience  strongest  impact  of  innova&on  on  MFP   growth   RELATIVE  CHANGE  IN  MFP  OVER  3  YEARS  BY  INNOVATION  TYPE  &  SECTOR   26   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  3-­‐yr  MA  of  MFP  from  year  0  to  year  t  on   innova?on  in  year  0  interacted  with  firm  characteris?cs  and  controls  for  base  year.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.  
  27. 27. -20 -10 0 10 20% NZ-owned domestic foreign-owned domestic international level of internationalisation R&D activity Product innovation new to world Product innovation Process innovation Organisational innovation Marketing innovation Interna&onally  connected  firms  engaged  in  product  or  organisa&onal  innova&on   experience  higher  MFP  growth,  while  domes&c  firms  have  lower  growth   RELATIVE  CHANGE  IN  MFP  OVER  3  YEARS  BY  INNOVATION  TYPE  &  INTERNATIONAL  FOCUS   27   Charts  show  difference  in  predic?ve  margins  from  series  of  OLS  regressions  of  change  in  3-­‐yr  MA  of  MFP  from  year  0  to  year  t  on   innova?on  in  year  0  interacted  with  firm  characteris?cs  and  controls  for  base  year.  Bars  show  95%  confidence  intervals.  Sample  contains   firms  in  BOS  2005-­‐2011  with  output  measures  un?l  2012.  Observa?ons  weighted  by  BOS  sampling  weights  ?mes  predicted  output.  
  28. 28. Innova&on  macers  for  produc&vity,  but  more  for  some  firms  than  others   FINDINGS   •  R&D  and  innovaCon  associated  with  growth  in  size  but  not  necessarily  with  producCvity   •  Firms  doing  markeCng  innovaCon  show  clearest  returns  overall,  especially  in  2007-­‐2009   (economic  downturn)   •  Firms  doing  product  and  organisaConal  innovaCon  performed  beeer  afer  2009  (economy   recovery),  but  possibly  worse  before   •  Younger  firms,  manufacturing  firms,  and  firms  with  internaConal  connecCons  have  highest   returns  to  innovaCon   CONCLUSIONS   •  More  to  innovaCon  than  R&D   •  Non-­‐technological-­‐types  of  innovaCon  maeer  as  much  if  not  more   •  Market  condiCons  important   •  Some  types  of  firms  get  more  value  out  of  innovaCon  than  others   For  more  informa&on   www.produc&vity.govt.nz     @Nzprocom   28  

×