dg ecfin ecfin– 12/10/2009 slide 1 effect of the changeover to the mandatory reply on the response...

15
DG ECFIN ECFIN– 12/10/2009 ECFIN– 12/10/2009 Slide Slide 1 Effect of the changeover to the Effect of the changeover to the mandatory reply on the response rate mandatory reply on the response rate for the French surveys for the French surveys José Bardaji, Olivier Biau, José Bardaji, Olivier Biau, Vincent Bonnefoy, Jean-Baptiste Pinault Vincent Bonnefoy, Jean-Baptiste Pinault Fourth Joint EU OECD Workshop on Business and Consumer Surveys Brussels, 12 -13 October 2009 European Commission Directorate General Economic and Financial Affairs

Upload: cory-andrews

Post on 14-Jan-2016

214 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

ECFIN– 12/10/2009ECFIN– 12/10/2009Slide Slide 11

Effect of the changeover to the Effect of the changeover to the mandatory reply on the response rate mandatory reply on the response rate

for the French surveysfor the French surveys

José Bardaji, Olivier Biau,José Bardaji, Olivier Biau, Vincent Bonnefoy, Jean-Baptiste PinaultVincent Bonnefoy, Jean-Baptiste Pinault

Fourth Joint EU OECD Workshop on Business and Consumer Surveys

Brussels, 12 -13 October 2009

European CommissionDirectorate General Economic and Financial Affairs

Page 2: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 22

IntroductionIntroduction

• In January 2004, two surveys became mandatory:– Monthly « Activity » – Quarterly « Investment »

• Biannual « Cash » survey remained voluntary

• Did the response rate improve ?

• Did the new respondents have specific behaviour?

Page 3: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 33

AgendaAgenda

• Short presentation of the 3 surveys

• Descriptive analyse of the impact of the obligation– On the response rates (short and medium term)– On the behaviour of new respondents

• Confirmation based on microeconometric study

Page 4: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 44

3 French Industry Surveys3 French Industry Surveys

• Monthly: « Activity » survey (11 per year)• Quarterly: « Investment » (3 per year)

Both included in the JH EU Programme of BCS

• Biannual: « Cash » survey (2 per year) Not included in the JH EU Programme of BCS

• Same survey sample (n=4.000 firms), stratified sampling with exhaustive stratum

• Qualitative questions: ↑ = ↓

Page 5: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 55

Common methodology Common methodology to analyse the response rateto analyse the response rate

• To eliminate seasonality in response rate, analyse on several dates before and after 2004

• Short term (2003-2004): 6 « Activity », 2 « Investment » and 2  «Cash »

• Medium term and econometric analysis (2001-2006): 66 « Activity », 18 « Investment » and 12 « Cash »

• To eliminate the periodically updates of the sample, study based on balanced samples

• Response: at least one in the questionnaire

Page 6: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 66

Analyse of the response rates Analyse of the response rates short term: 2003-2004short term: 2003-2004

 

• Significant increase of response rate for « Activity » (+2 points) and « Investment » (+3 points)

• Not for « Cash »

Activity Investment Cash

Before obligation 78,9 % 71,4 % 76,4 %

After obligation 81,2 % 74,3 % 75,3 %

Difference +2,3 pp +2,9 pp -1,1 pp

t-Test (difference=0) Rejected (p-value = 0,0002)

Rejected (p-value = 0,0002)

Rejected (p-value = 0,1834)

Page 7: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 77

Analyse of the response rates Analyse of the response rates medium term: 2001-2006medium term: 2001-2006

 

• Strong significant increase of response rate for « Activity » (+5 points) and « Investment » (+9 points)

• But also for « Cash » (+1.5 point)

Activity Investment Cash

Before obligation 84,4% 76,0% 82,2%

After obligation 89,7% 84,9% 83,7%

Difference +5,3 pp +8,8 pp +1,5 pp

t-Test (difference=0) Rejected (p-value < 0,0001)

Rejected (p-value < 0,0001)

Rejected (p-value = 0,0075)

Page 8: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 88

Analyse of the response rates Analyse of the response rates medium term: 2001-2006medium term: 2001-2006

 Cumulative fonction: P(number of responses ≤ n)

n max =33

• For « Activity », 60% of firms respond at least 30 times before obligation and 70% after.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 3 6 9 12 15 18 21 24 27 30 33

Before

After

Page 9: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 99

Behaviour of new respondentsBehaviour of new respondents

• Focus on « Activity » - 3 months before and 3 months after the changeover to mandatory reply

• Difference in difference estimator DD = (Y1 –Y0)Treatment group – (Y1-Y0)Control group

Y1: output AFTER obligation (e.g. balance, %stable) Y0: output BEFORE obligation

• Same fixed effects in both groups: same trend in economic outlook

Page 10: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1010

Behaviour of new respondentsBehaviour of new respondents

• Control group: firms that respond to the 6 surveys (no effect of the change over)

• Treatment groups: firms whose number of responses changes (eg. Better)

Before After

none one two three

none Impossible Nothing after Nothing after Nothing after

one Nothing before Same Worst Worst

two Nothing before Better Same Worst

three Nothing before Better Better Control

Page 11: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1111

Behaviour of new respondentsBehaviour of new respondents

• Y = %stable in question Q5 (expectation)

• No specific behavior of the new respondents.

Before After Difference Double difference (1) (2) (2)-(1)

Control Group Control (1 731 firms) 55 56 2 Nothing before (106) 57 nd nd Nothing after (188) 38 nd nd Worst (235) 55 55 0 -2 Same (124) 56 58 2 0 %

stab

le Q

5

Traitment Groups

Better (238) 53 55 2 0

Page 12: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1212

Microeconometric studyMicroeconometric study

• Statistical unit: firm

• Balanced samples 2001-2006

• Merge with BRN’s database (fiscal declaration and annual survey)

• Size of samples : 1 500 to 2 000

• Modelling the probability of non response: yit=1 ou 0

• Obligation: 1 if t ≥ 2004

• 3 models:– Without fixed effects (Logit)– With fixed effects (Random effects Probit and Conditional Logit)

Page 13: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1313

3 models3 models

Logit:

yit* = Xit.β + Xi.γ + εit

Random Effects Probit:

yit* = Xit.β + Xi.γ + ci + uit

Conditional Logit:

yit* = Xit.β + ci + vit

Page 14: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1414

Random Effects ProbitRandom Effects Probit

yit* = Xit.β + Xi.γ + ci + uit Enquête

Activité Enquête

I nvestissement Enquête

Trésorerie(1) (N = 95 491) (N = 40 919) (N = 17 600)

β σβ Signif. β σβ Signif. β σβ Signif. Constante -1,41 (0,08) *** -0,95 (0,08) *** -1,47 (0,13) *** Délais moyens de règlement client 0,35 (0,09) *** 0,05 (0,10) 0,38 (0,19) ** Salaire moyen -0,23 (0,13) * -0,22 (0,15) 0,00 (0,20) Part des salaires dans la production -0,41 (0,14) *** -0,27 (0,16) * -0,51 (0,28) * Taux d’endettement 0,08 (0,05) * 0,08 (0,06) 0,34 (0,13) *** Taux de marge -0,03 (0,01) *** -0,04 (0,01) *** -0,04 (0,01) *** Nomenclature économique de synthèse

Industrie agroalimentaire 0,14 (0,10) 0,11 (0,10) 0,15 (0,14) Industrie de biens de consommation 0,12 (0,07) * 0,19 (0,07) *** 0,24 (0,10) ** Industrie automobile 0,24 (0,15) 0,24 (0,15) 0,03 (0,25) Industrie de biens d’équipement -0,03 (0,07) -0,01 (0,06) 0,05 (0,10) Industrie de biens intermédiaires Référence référence référence

Énergie 0,33 (0,53) 0,45 (0,27) * 1,37 (0,51) *** Chiffre d’affaires

moins de 15 millions d’€ -0,11 (0,04) *** -0,12 (0,05) ** -0,23 (0,09) *** entre 15M€ et 150M€ Référence référence référence

150 M€ et plus -0,02 (0,05) -0,07 (0,07) -0,04 (0,10) Effectif

moins de 100 salariés -0,01 (0,05) -0,10 (0,05) * -0,25 (0,09) *** entre 100 et 500 salariés Référence référence référence

500 salariés et plus 0,36 (0,05) *** 0,39 (0,06) *** 0,49 (0,09) *** Obligation -0,34 (0,01) *** -0,43 (0,02) *** -0,09 (0,03) *** σc 0,99 (0,03) *** 0,99 (0,02) *** 1,19 (0,04) ***

Page 15: DG ECFIN ECFIN– 12/10/2009 Slide 1 Effect of the changeover to the mandatory reply on the response rate for the French surveys José Bardaji, Olivier Biau,

DG ECFIN

Olivier BiauOlivier BiauEuropean Commission, Economic and Financial AffairsEuropean Commission, Economic and Financial Affairs

ECFIN – 12/10/2009ECFIN – 12/10/2009Slide Slide 1515

ConclusionsConclusions

• Significant improvement of the response rate for « Activity » and « Investment »

• Stronger effect at medium term

• Whatever the model, the obligation increased the probability of response

• Response rate for optional « Cash » survey also improved, albeit to a smaller extent