Banque nationale de Belgique SA · EURONEXT:BNB

Banque nationale de Belgique SA — Investor Presentation (March 2014)

86 slides from Banque nationale de Belgique SA's investor presentation, March 2014

  1. Slide 1: The Anatomy of French Production Hierarchies

    The Anatomy of French Production Hierarchies Lorenzo Caliendo Yale University - NBER Ferdinando Monte Johns Hopkins University Esteban Rossi-Hansberg Princeton University - NBER March 2014 Caliendo, Monte, Rossi-Hansberg The Anatomy () of Production Hierarchies March 2014 1 / 33

  2. Slide 2: Introduction

    Introduction Firms are heterogeneous in a variety of dimensions I But little is known about where this heterogeneity comes from Some of the observed heterogeneity is the result of organizational di¤erences I The number and knowledge of employees Our aim is to understand empirically how …rms are orga

  3. Slide 3: Related Literature

    Related Literature Model of organization based on Garicano (2000) I Applied to GE in Garicano and Rossi-Hansberg (2004, 2006, 2011) I With heterogeneus …rms in a product market: F Caliendo and Rossi-Hansberg (2012) Few empirical studies on organizational change I Baker, Gibbs, and Holmstrom (1994):

  4. Slide 4: Sketch of Theory in CRH (2012)

    Sketch of Theory in CRH (2012) Agents employed in a …rm act as production workers or managers Production workers: I Generate a production possibility that can yield 1 unit of output I For output to be realized the worker needs to solve a problem F Problems are drawn from F (z ) with F 00 (z ) < 0 I

  5. Slide 5: Sketch of Theory in CRH (2012)

    Sketch of Theory in CRH (2012) Managers I Specialize in solving problems I Spend h units of time with each problem that gets to them F So each manager can deal with 1/h problems I A manager of layer 2 tries to solve the problems workers could not solve F So problems that require knowledge larger tha

  6. Slide 6: Sketch of Theory: Cost Minimization

    Sketch of Theory in CRH (2012): Cost Minimization Consider a …rm that produces a quantity q. CL (q; w ) is the minimum cost of producing q with an organization with L layers, namely, ∑`=1 nL` wL` L CL (q; w ) = min fn L` ,zL` gLl=1 0 subject to q F (ZLL )nL1 , wL` = w̄ [czL` + 1] for all ` L, nL` =

  7. Slide 7: Sketch of Theory in CRH (2012)

    Sketch of Theory in CRH (2012) Average cost function AC(q) Hierarchy at  w() w() 0 2 4 6 8 10 Hierarchy at ' w(') w(') > w() C(q)/q w(') > w() w(') 0 2 4 6 8 10 Hierarchy at '' w('') w('') < w() w('') w31('') < w21() w('') 0 2 4 6 8 10 q() q(') Number of employees q('') Caliendo

  8. Slide 8: Implications of the Model

    Implications of the Model 1) Firms are hierarchical, nL1 ...nL` ... nLL for all L 2) Layers L, sales pq, and total labor demand ∑`=1 nL` , increase with α L 3) Given L, wL` and nL` increase with α at all ` 4) Given α, wL` decreases and nL` increases with an increase in L at all ` Caliendo, Monte, Ro

  9. Slide 9: Data description

    Data description Dataset collected by the French National Statistical Institute (INSEE) I We use the period from 2002 to 2007 F Before 2002 di¤erent occupational categories We match two sources from mandatory reports: I BRN: private …rms balance sheet data F 553,125 …rm-year observations in manufact

  10. Slide 10: Layers: occupational categories

    Layers: occupational categories PCS-ESE classi…cation codes that belong to manufacturing: 2 Firm owners receiving a wage F CEO or …rm directors 3 Senior sta¤ or top management positions F chief …nancial o¢ cers, head of HR, logistics, purchasing managers 4 Employees at the supervisor level F quality

  11. Slide 11: Firms with different number of layers are different

    Firms with di¤erent number of layers are di¤erent Value added distribution by number of layers Hours distribution by number of layers Raw data − thousands of 2005 euros Raw data .5 .5 .4 .4 .3 .3 Density Density .2 .2 .1 .1 0 0 1 10 100 1000 10000 100000 10 100 1000 10000 100000 1000000 Value added

  12. Slide 12: Firms with adjacent occupational categories

    Firms with adjacent occupational categories We select the sub-sample of …rms that satisfy the following criteria: I Layer 0 …rms are …rms with occupation codes 6 and 5 I Layer 1 …rms are …rms with occupation codes 6, 5 and 4 I Layer 2 …rms are …rms with occupation codes 6, 5, 4 and 3 I Layer 3 …rms

  13. Slide 13: Hours and wages are hierarchical

    Hours and wages are hierarchical Percentage of …rms that satisfy a hierarchy NL` = hours at layer ` of a …rm with L layers Unweighted # of layers NL` N L`+1 all ` NL1 NL2 NL2 NL3 NL3 NL4 2 85.6 85.6 - - 3 63.4 85.9 74.8 - 4 56.5 86.9 77.5 86.9 Unweighted # of layers wL`+1 wL` all ` wL2 wL1 wL3 wL2 w

  14. Slide 14: Variation in log wages

    Variation in log wages Mean share variation of wages explained by cross-layer variation Weighted by Firm-years Unweighted Hours VA All …rms 434,872 0.50 0.51 0.49 Firms with more than 0 layers 370,997 0.59 0.51 0.50 Firms with 0 layers 63,875 0.00 0.00 0.00 Firms with 1 layer 124,299 0.50 0.41 0.43

  15. Slide 15: Representative hierarchies

    Representative hierarchies Hierarchy of a 1 layer firm Hierarchy of a 2 layers firm Average hourly wage Average hourly wage 30.89 27.17 18.15 0 10 20 30 40 50 0 10 20 30 40 50 Average hours (thousands) Average hours (thousands) Hierarchy of a 3 layers firm Hierarchy of a 4 layers firm Average hourly

  16. Slide 16: Representative hierarchies: normalized hours

    Representative hierarchies: normalized hours Average hourly wage Hierarchy of a 1 layer firm Hierarchy of a 2 layers firm Average hourly wage 30.89 27.17 18.15 0 5 10 15 20 25 0 5 10 15 20 25 Average hours normalized by the top layer Average hours normalized by the top layer Hierarchy of a 3 layers

  17. Slide 17: Layer transitions

    Layer transitions Distribution of # of layers at time t+1 given the # of layers at time t # of layers at t + 1 Exit 1 2 3 4 Total 1 15.3 67.5 15.2 1.9 0.2 100 # of layers 2 9.8 10.7 62.2 16.2 1.1 100 at t 3 7.7 1.2 13.1 67.6 10.5 100 4 6.2 0.2 2.0 20.5 71.3 100 Weighted by VA Caliendo, Monte, Rossi-

  18. Slide 18: Transitions across layers depend on value added

    Transitions across layers depend on value added Transitions of firms out of 1 layer Transitions of firms out of 2 layers .5 .5 .4 .4 Fraction of firms Fraction of firms .3 .3 .2 .2 .1 .1 0 0 1 10 100 1000 10000 100000 1 10 100 1000 10000 100000 Value added Value added to 2 lyrs to 3 lyrs to 4 lyrs t

  19. Slide 19: Transitions across layers depend on value added

    Transitions across layers depend on value added Transitions of firms out of 1 layer Transitions of firms out of 2 layers .5 .5 .4 .4 Fraction of firms Fraction of firms .3 .3 .2 .2 .1 .1 0 0 1 10 100 1000 10000 100000 1 10 100 1000 10000 100000 Value added (log scale) Value added (log scale) to 2 ly

  20. Slide 20: Trends before adding or dropping layers

    Trends before adding or dropping layers Firms with 1 layer before the transition Firms with 2 layers before the transition .8 1 .8 .6 dlog value added dlog value added .6 .4 .4 .2 .2 0 0 -2 -1 0 -2 -1 0 lag (0 = transition period) lag (0 = transition period) to 1 lyr to 2 lyrs to 3 lyrs to 4 lyrs to

  21. Slide 21: Firm level outcomes during transition

    Change in …rm level outcomes during transition Average behavior of …rms by change in the number of layers All Increase L No change in L Decrease L d lnhours -0.015*** 0.040*** -0.012*** -0.081*** - detrended - 0.055*** 0.003*** -0.066*** d ln ∑`=0 nL` L -0.011*** 1.362*** 0.012*** -1.404*** - detren

  22. Slide 22: Normalized hours change according to the theory

    Normalized hours change according to the theory Average log change in normalized hours for …rms that transition # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.537 0.018 0.00 10177 1 3 1 1.762 0.056 0.00 1263 1 4 1 2.266 0.212 0.00 97 2 1 1 ‐1.582 0.017 0.00 11106 2 3 1 0.716 0.012 0.

  23. Slide 23: Normalized hours change according to the theory

    Normalized hours change according to the theory Elasticity of nL` with VA for …rms that do not change L Reporting βL` from d ln nLit ` = α` + β` d ln VA + ε L L it it # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 2 1 0.042 0.012 0.00 64,536 3 1 0.039 0.009 0.00 91,253 3 2 0.013 0.010 0.20

  24. Slide 24: Wages change according to the theory

    Wages change according to the theory Average log change in wages for …rms that transition # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.129 0.005 0.00 10177 1 3 1 ‐0.332 0.020 0.00 1263 1 4 1 ‐0.678 0.117 0.00 97 2 1 1 0.167 0.005 0.00 11106 2 3 1 ‐0.050 0.002 0.00 16800 2 3 2 ‐0.2

  25. Slide 25: Wages change according to the theory

    Wages change according to the theory Elasticity of wL` with VA for …rms that do not change L ` = δ` + γ` d ln VA + ε Reporting γL` from d ln wLit L L it it # of layers in Layer γL` s.e. p-value obs the …rm (L) ` 1 1 0.077 0.007 0.00 45,045 2 1 0.100 0.006 0.00 64,536 2 2 0.118 0.006 0.00 64,536 3 1

  26. Slide 26: Representative hierarchies for one layer transitions

    Representative hierarchies for one layer transitions Firms with 1 layer After transition Firms with 2 layers After transition Average hourly wage Average hourly wage Average hourly wage Average hourly wage 39.8 30 32.9 25.5 23.2 20.4 17.6 16.7 0 1 2 3 4 5 0 1 2 3 4 5 0 5 10 15 0 5 10 15 Average hour

  27. Slide 27: Distribution of wages after minus before transition

    Distribution of wages after minus before transition Transition from 2 to 1 Transition from 3 to 2 Transition from 4 to 3 0.2 0.04 0.02 0.03 0.01 Log wage differences Log wage differences Log wage differences 0.15 0 0.02 -0.01 0.1 0.01 -0.02 0 -0.03 0.05 -0.01 -0.04 0 -0.02 -0.05 0 10 20 30 40 50 60

  28. Slide 28: Distribution of wages after minus before transition: common layers

    Distribution of wages after minus before transition Common layers Transition from 2 to 1 Transition from 3 to 2 Transition from 4 to 3 0.4 0.35 0.14 0.3 0.12 Log wage differences Log wage differences Log wage differences 0.3 0.25 0.1 0.2 0.08 0.2 0.15 0.06 0.1 0.04 0.1 0.05 0.02 0 0 0 0 10 20 30 40

  29. Slide 29: Distribution of wages after minus before

    Distribution of wages after minus before Conditioning on increase in VA > 0 and no transition Firms with 1 layer Firms with 2 layers 0.12 0.08 0.1 Log wage differences Log wage differences 0.06 0.08 0.06 0.04 0.04 0.02 0.02 0 0 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Percen

  30. Slide 30: How do firms change the average wage in a layer?

    How do …rms change the average wage in a layer? Extensive versus intensive margin Log diff. in hourly wage of hours entering the layer (after Log diff. in hourly wage (after minus before the transition) versus hours leaving the layer (before transition) for hours staying in the layer transition) # o

  31. Slide 31: How do firms change the average wage in a layer?

    How do …rms change the average wage in a layer? Education or experience to adjust knowledge and wages Elasticity of ‘knowledge’with VA for …rms that do not change L # of layers Layer Experience p-value Education p-value obs 1 1 0.0014 0.69 0.0015 0.03 45,009 2 1 -0.0101 0.01 0.0042 0.00 64,469 2 2 0

  32. Slide 32: How do firms change the average wage in a layer?

    How do …rms change the average wage in a layer? Education or experience to adjust knowledge and wages Average change in 'knowledge' for firms that change L # of layers Layer Experience p‐value Education p‐value obs Before After 1 2 1 ‐0.108 0.00 ‐0.004 0.00 10,171 1 3 1 ‐0.184 0.00 ‐0.003 0.29 1,261

  33. Slide 33: Conclusion

    Conclusion We use French data to study the organization of production I Organizing the data using layers of employees is meaningful and useful We document that: 1 Firms are hierarchical across layers in terms of employees and wages 2 The probability of adding a layer increases with value added F Fir

  34. Slide 34: Agents working at different layers are different

    Agents working at di¤erent layers are di¤erent In terms of hours Percentiles Layer # Mean p5 p10 p25 p50 p75 p90 p95 0 48,532 1,443 2,028 5,372 13,189 32,944 81,226 163,721 1 22,309 588 1,280 1,984 4,056 10,953 31,381 66,195 2 22,669 593 1,240 1,820 3,640 8,575 27,493 60,806 3 2,310 338 711 1,773 1,

  35. Slide 35: Occupational categories

    Occupational categories Statistics on wage by occupation Average hourly wage by occupation in 2005 Euros CEO, Senior Blue Supervisors Clerks directors sta¤ collars Mean 81.39 47.83 26.58 19.01 20.70 p5 23.68 21.45 14.35 10.63 10.64 p10 28.60 25.01 16.21 11.79 11.82 p25 41.51 31.00 19.36 13.84 13.65

  36. Slide 36: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample, positive change in VA and normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.838 0.025 0.00 3568 1 3 1 2.394 0.083 0.00 437 1 4 1 3.023 0.272 0.00 47 2 1 1 ‐1.827 0.025 0.00 3787 2 3

  37. Slide 37: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on positive change in VA and normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.111 0.006 0.00 5331 1 3 1 ‐0.314 0.026 0.00 709 1 4 1 ‐0.594 0.127 0.00 65 2 1 1 0.168 0.007 0.00 6376 2 3 1 ‐0.017 0.003

  38. Slide 38: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample and positive change in VA # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.838 0.025 0.00 3568 1 3 1 2.394 0.083 0.00 437 1 4 1 3.023 0.272 0.00 47 2 1 1 ‐1.827 0.025 0.00 3787 2 3 1 1.252 0.023 0.00

  39. Slide 39: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample and positive change in total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.850 0.024 0.00 3819 1 3 1 2.545 0.076 0.00 467 1 4 1 3.242 0.300 0.00 35 2 1 1 ‐1.861 0.025 0.00 3888 2 3 1 1.314 0

  40. Slide 40: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample and positive change in normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.807 0.019 0.00 6675 1 3 1 2.398 0.061 0.00 780 1 4 1 2.650 0.206 0.00 72 2 1 1 ‐1.843 0.019 0.00 6886 2 3 1 1.

  41. Slide 41: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on positive change in VA and total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.510 0.030 0.00 3628 1 3 1 1.789 0.088 0.00 535 1 4 1 2.952 0.312 0.00 43 2 1 1 ‐1.512 0.027 0.00 4498 2 3 1 0.765 0.019 0.00 613

  42. Slide 42: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.807 0.019 0.00 6675 1 3 1 2.398 0.061 0.00 780 1 4 1 2.650 0.206 0.00 72 2 1 1 ‐1.843 0.019 0.00 6886 2 3 1 1.251 0.017 0.00 8260 2 3 2 1.092 0.017 0.

  43. Slide 43: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on positive change in VA # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.525 0.025 0.00 5331 1 3 1 1.745 0.075 0.00 709 1 4 1 2.654 0.263 0.00 65 2 1 1 ‐1.528 0.023 0.00 6376 2 3 1 0.720 0.016 0.00 9077 2 3 2 0.543 0.

  44. Slide 44: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on positive change in total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.478 0.024 0.00 5672 1 3 1 1.716 0.071 0.00 803 1 4 1 2.674 0.264 0.00 57 2 1 1 ‐1.513 0.022 0.00 6604 2 3 1 0.754 0.015 0.00 9140 2 3 2

  45. Slide 45: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on positive change in normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.537 0.018 0.00 10177 1 3 1 1.762 0.056 0.00 1263 1 4 1 2.266 0.212 0.00 97 2 1 1 ‐1.582 0.017 0.00 11106 2 3 1 1.106 0.012 0.00 13

  46. Slide 46: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Conditioning on selected sample, positive change in VA and total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.868 0.030 0.00 2494 1 3 1 2.595 0.093 0.00 328 1 4 1 3.534 0.373 0.00 26 2 1 1 ‐1.862 0.030 0.00 2600 2 3 1 1.3

  47. Slide 47: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample, positive change in VA and normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.039 0.006 0.00 3568 1 3 1 ‐0.157 0.029 0.00 437 1 4 1 ‐0.293 0.129 0.03 47 2 1 1 0.093 0.007 0.00 3787 2 3 1

  48. Slide 48: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on positive change in VA and normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.111 0.006 0.00 5331 1 3 1 ‐0.314 0.026 0.00 709 1 4 1 ‐0.594 0.127 0.00 65 2 1 1 0.168 0.007 0.00 6376 2 3 1 ‐0.017 0.003 0.0

  49. Slide 49: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample and positive change in VA # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.039 0.006 0.00 3568 1 3 1 ‐0.157 0.029 0.00 437 1 4 1 ‐0.293 0.129 0.03 47 2 1 1 0.093 0.007 0.00 3787 2 3 1 0.004 0.004 0.28

  50. Slide 50: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample and positive change in total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.160 0.006 0.00 3819 1 3 1 ‐0.305 0.031 0.00 467 1 4 1 ‐0.829 0.215 0.00 35 2 1 1 0.200 0.007 0.00 3888 2 3 1 ‐0.086 0

  51. Slide 51: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample and positive change in normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.066 0.005 0.00 6675 1 3 1 ‐0.170 0.022 0.00 780 1 4 1 ‐0.372 0.119 0.00 72 2 1 1 0.102 0.005 0.00 6886 2 3 1 ‐0.

  52. Slide 52: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on a positive change in VA and total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.194 0.008 0.00 3628 1 3 1 ‐0.417 0.032 0.00 535 1 4 1 ‐0.872 0.170 0.00 43 2 1 1 0.270 0.008 0.00 4498 2 3 1 ‐0.064 0.004 0.00 6

  53. Slide 53: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.066 0.005 0.00 6675 1 3 1 ‐0.170 0.022 0.00 780 1 4 1 ‐0.372 0.119 0.00 72 2 1 1 0.102 0.005 0.00 6886 2 3 1 ‐0.026 0.003 0.00 8260 2 3 2 ‐0.075 0.004 0

  54. Slide 54: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on a positive change in VA # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.111 0.006 0.00 5331 1 3 1 ‐0.314 0.026 0.00 709 1 4 1 ‐0.594 0.127 0.00 65 2 1 1 0.168 0.007 0.00 6376 2 3 1 ‐0.021 0.003 0.00 9077 2 3 2 ‐0.219

  55. Slide 55: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on a positive change in total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.254 0.006 0.00 5672 1 3 1 ‐0.517 0.028 0.00 803 1 4 1 ‐1.120 0.169 0.00 57 2 1 1 0.287 0.007 0.00 6604 2 3 1 ‐0.113 0.004 0.00 9140 2 3

  56. Slide 56: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on a positive change in normalized hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.129 0.005 0.00 10177 1 3 1 ‐0.332 0.020 0.00 1263 1 4 1 ‐0.678 0.117 0.00 97 2 1 1 0.167 0.005 0.00 11106 2 3 1 ‐0.046 0.003 0.00

  57. Slide 57: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on selected sample, positive change in VA and total hours # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.102 0.007 0.00 2494 1 3 1 ‐0.229 0.035 0.00 328 1 4 1 ‐0.567 0.215 0.01 26 2 1 1 0.180 0.009 0.00 2600 2 3 1 ‐0.0

  58. Slide 58: Change in average wages for firms that do not transition

    Change in average wages for …rms that do not transition Conditioning on selected sample ` = δ` + γ` d ln VA + ε Reporting γL` from d ln wLit L L it it # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 0 0 0.067 0.009 0.00 39,914 1 0 0.106 0.009 0.00 42,071 1 1 0.118 0.009 0.00 42,071 2 0 0.145

  59. Slide 59: Change in normalized hours for firms that do not transition

    Change in normalized hours for …rms that do not transition Conditioning on selected sample Reporting βL` from d ln nLit ` = α` + β` d ln VA + ε L L it it # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 1 0 0.027 0.014 0.06 42,071 2 0 0.036 0.009 0.00 71,424 2 1 0.013 0.011 0.26 71,424 3 0 0.

  60. Slide 60: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on a positive change in VA and normalized hours - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.083 0.004 0.00 5331 1 3 1 ‐0.159 0.014 0.00 709 1 4 1 ‐0.368 0.088 0.00 65 2 1 1 0.087 0.003 0.00 6376 2 3 1 ‐0

  61. Slide 61: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on positive change in VA and total hours - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.110 0.005 0.00 3628 1 3 1 ‐0.199 0.017 0.00 535 1 4 1 ‐0.529 0.116 0.00 43 2 1 1 0.111 0.004 0.00 4498 2 3 1 ‐0.043 0.

  62. Slide 62: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on positive change in VA - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.083 0.004 0.00 5331 1 3 1 ‐0.159 0.014 0.00 709 1 4 1 ‐0.368 0.088 0.00 65 2 1 1 0.087 0.003 0.00 6376 2 3 1 ‐0.027 0.002 0.00 9077 2

  63. Slide 63: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on positive change in total hours - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.121 0.004 0.00 5672 1 3 1 ‐0.222 0.014 0.00 803 1 4 1 ‐0.488 0.097 0.00 57 2 1 1 0.119 0.004 0.00 6604 2 3 1 ‐0.048 0.002 0.0

  64. Slide 64: Change in average wages for firms that change layers

    Change in average wages for …rms that change layers Conditioning on positive change in normalized hours - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.080 0.003 0.00 10177 1 3 1 ‐0.155 0.010 0.00 1263 1 4 1 ‐0.302 0.068 0.00 97 2 1 1 0.086 0.003 0.00 11106 2 3 1 ‐0.022 0

  65. Slide 65: Wages change according to the theory

    Wages change according to the theory Average log change in wages for …rms that transition - DADS data # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.080 0.003 0.00 10177 1 3 1 ‐0.155 0.010 0.00 1263 1 4 1 ‐0.302 0.068 0.00 97 2 1 1 0.086 0.003 0.00 11106 2 3 1 ‐0.026 0.002 0.00 1680

  66. Slide 66: Wages change according to the theory

    Wages change according to the theory Firms that do not transition - DADS data ` = δ` + γ` d ln VA + ε Reporting γL` from d ln wLit L L it it # of layers in Layer γL` s.e. p-value obs the …rm (L) ` 0 0 0.000 0.003 0.94 45,606 1 0 0.001 0.002 0.63 65,114 1 1 0.019 0.003 0.00 65,114 2 0 -0.005 0.002 0.

  67. Slide 67: Firms with adjacent occupational categories

    Firms with adjacent occupational categories We select the sub-sample of …rms that satisfy the following criteria: I Layer 0 …rms are …rms with occupation codes 6 and 5 I Layer 1 …rms are …rms with occupation codes 6, 5 and 4 I Layer 2 …rms are …rms with occupation codes 6, 5, 4 and 3 I Layer 3 …rms

  68. Slide 68: Firms with adjacent occupational categories

    Firms with adjacent occupational categories We select the sub-sample of …rms that satisfy the following criteria: I Layer 0 …rms are …rms with occupation codes 6 and 5 I Layer 1 …rms are …rms with occupation codes 6, 5 and 4 I Layer 2 …rms are …rms with occupation codes 6, 5, 4 and 3 I Layer 3 …rms

  69. Slide 69: Layer transitions

    Layer transitions Distribution of # of layers at time t+1 given the # of layers at time t Weighted by VA # of layers at t + 1 Exit 1 2 3 4 Total 1 11.3 65.3 19.5 3.3 0.6 100 # of layers 2 7.1 6.6 62.7 21.5 2.1 100 at t 3 5.8 0.2 2.4 72.6 19.0 100 4 7.7 0.0 0.2 13.4 78.8 100 Back Caliendo, Monte, Ros

  70. Slide 70: Fraction of firms that transition to an adjacent layer

    Fraction of …rms that transition to an adjacent layer What is the fraction of …rms that transition up or down to an adjacent layer? I Conditioning of …rms with adjacent layers # of Transition layers Up Down 1 75.5 - 2 82.3 91.5 3 100 60.6 4 - 75.9 back Caliendo, Monte, Rossi-Hansberg The Anatomy ()

  71. Slide 71: Firm level outcomes during transition

    Change in …rm level outcomes during transition Average behavior of …rms by change in the number of layers All Increase L No change in L Decrease L dlnhours -0.014*** 0.046*** -0.011*** -0.084*** - detrended - 0.060*** 0.003*** -0.070*** dln ∑`=0 nL` L -0.011*** 1.342*** 0.012*** -1.385*** - detrende

  72. Slide 72: Firm level outcomes during transition

    Change in …rm level outcomes during transition Average behavior of …rms with adjacent layers by change in the number of layers All Increase L No change in L Decrease L dlnhours -0.009*** 0.038*** -0.008*** -0.059*** - detrended - 0.048*** 0.001 -0.050*** dln ∑`=0 nL` L -0.011*** 1.788*** 0.008*** -1

  73. Slide 73: Normalized hours change according to the theory

    Normalized hours change according to the theory Firms that change export status and do not change L, selected sample Reporting βL` from d ln nLit ` = α` + β` d ln VA + ε L L it it # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 1 0 -0.045 0.044 0.31 4,550 2 0 0.009 0.026 0.73 8,031 2 1 -0.01

  74. Slide 74: Data description

    Data description By number of layers in the …rm, DADS data # of Average Median layers Firm-years VA Hours wage 0 81,909 205 7,946 10.18 1 126,069 403 16,450 12.08 2 161,449 2,821 85,674 14.22 3 87,211 8,879 227,070 15.71 Value added in 000s of 2005 euros. Back Caliendo, Monte, Rossi-Hansberg The Ana

  75. Slide 75: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Average 2 years ahead log change # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.338 0.023 0.00 4152 1 3 1 1.603 0.090 0.00 344 1 4 1 3.067 0.355 0.00 24 2 1 1 ‐1.677 0.026 0.00 4916 2 3 1 0.608 0.015 0.00 7081 2 3 2 0.473 0.015 0

  76. Slide 76: Change in normalized hours for firms that change layers

    Change in normalized hours for …rms that change layers Average 3 years ahead log change # of layers Layer Change s.e. p‐value obs Before After 1 2 1 1.384 0.033 0.00 1891 1 3 1 1.558 0.144 0.00 132 1 4 1 3.264 0.550 0.00 10 2 1 1 ‐1.674 0.037 0.00 2378 2 3 1 0.639 0.021 0.00 3440 2 3 2 0.509 0.020 0

  77. Slide 77: Change in wages for firms that change layers

    Change in wages for …rms that change layers Average 2 years ahead log change # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.155 0.007 0.00 4152 1 3 1 ‐0.489 0.045 0.00 344 1 4 1 ‐1.033 0.244 0.00 24 2 1 1 0.138 0.007 0.00 4916 2 3 1 ‐0.055 0.004 0.00 7081 2 3 2 ‐0.279 0.006 0.00 708

  78. Slide 78: Change in wages for firms that change layers

    Change in wages for …rms that change layers Average 3 years ahead log change # of layers Layer Change s.e. p‐value obs Before After 1 2 1 ‐0.144 0.010 0.00 1891 1 3 1 ‐0.550 0.075 0.00 132 1 4 1 ‐0.598 0.207 0.02 10 2 1 1 0.125 0.010 0.00 2378 2 3 1 ‐0.043 0.005 0.00 3440 2 3 2 ‐0.268 0.008 0.00 344

  79. Slide 79: Change in normalized hours for firms that do not transition

    Change in normalized hours for …rms that do not transition 2 years ahead Reporting βL` from ln nLit ` ` ` ` +2 /nLit = αL + βL ln (VAit +2 /VAit ) + εit # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 2 1 0.063 0.015 0.00 34,685 3 1 0.067 0.010 0.00 53,515 3 2 0.025 0.012 0.03 53,515 4 1 0.1

  80. Slide 80: Change in normalized hours for firms that do not transition

    Change in normalized hours for …rms that do not transition 3 years ahead Reporting βL` from ln nLit ` ` ` ` +3 /nLit = αL + βL ln (VAit +3 /VAit ) + εit # of layers in Layer βL` s.e. p-value obs the …rm (L) ` 2 1 0.112 0.019 0.00 18,068 3 1 0.081 0.013 0.00 30,224 3 2 0.046 0.015 0.00 30,224 4 1 0.2

  81. Slide 81: Wages change according to the theory

    Wages change according to the theory 2 years ahead ` Reporting γL` from ln wLit ` ` ` +2 /wLit = δL + γL ln (VAit +2 /VAit ) + εit # of layers in Layer γL` s.e. p-value obs the …rm (L) ` 1 1 0.046 0.009 0.00 26,170 2 1 0.067 0.007 0.00 34,685 2 2 0.101 0.007 0.00 34,685 3 1 0.099 0.006 0.00 53,515 3

  82. Slide 82: Wages change according to the theory

    Wages change according to the theory 3 years ahead ` Reporting γL` from ln wLit ` ` ` +3 /wLit = δL + γL ln (VAit +3 /VAit ) + εit # of layers in Layer γL` s.e. p-value obs the …rm (L) ` 1 1 0.021 0.010 0.04 26,170 2 1 0.059 0.008 0.00 34,685 2 2 0.102 0.009 0.00 34,685 3 1 0.081 0.006 0.00 53,515 3

  83. Slide 83: How do firms change the average wage in a layer?

    How do …rms change the average wage in a layer? Extensive versus intensive margin Log diff. in hourly wage of new hours entering the layer Log diff. in hourly wage of hours leaving the layer versus versus hours staying in the layer (after transition) hours who stayed in the layer (before the transit

  84. Slide 84: Firms with different number of layers are different

    Firms with di¤erent number of layers are di¤erent Value added distribution by number of layers Hours distribution by number of layers After removing year and industry FE − thousands of 2005 euros After removing industry and year FE .5 .5 .4 .4 .3 .3 Density Density .2 .2 .1 .1 0 0 1 10 100 1000 1000

  85. Slide 85: Sources of changes in average wage during a transition

    Sources of changes in average wage during a transition 0 w̄L`0 itL+1 /w̄Lit wLL0 it +1 /w̄Lit From/to 2 3 4 From/to 2 3 4 1 0.963 0.865 0.733 1 1.507 1.501 1.602 (10,167 ) (1,262 ) (96 ) (10,166 ) (1,263 ) (97 ) 2 0.926 0.876 2 2.040 2.021 (16,783 ) (1,128 ) (16,783 ) (1,129 ) 3 0.958 3 4.385 (14,09

  86. Slide 86: Log wage differences

    Distribution of wages after minus before Conditioning on decrease in VA < 0 and no transition Firms with 1 layer Firms with 2 layers 0.01 0.02 0 0.01 Log wage differences Log wage differences -0.01 0 -0.02 -0.01 -0.03 -0.02 -0.04 -0.03 -0.05 -0.04 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60

Banque Nationale De Belgique — Investor Presentation (March 2014) on 6ix