DUI it yourself: Innovation and activities to promote learning by doing, using, and interacting within the firm

Figures & data

Table 1. Description of the dependent variable.

Variable name	Description	Mean	Std.	Min	Max
New-to-market product innovation	1 if the firm has introduced new or significantly improved goods or services new to the market for the period 2008–2010, 0 if not.	0.182	0.381	0	1

Table 2. Descriptive statistics and variable definitions of innovation activities.

Variable name	Description	Mean	Std.dev.	Min.	Max.
DUI internal	1 if the firm used brainstorming, interdisciplinary workgroups, job rotation, non-financial incentives and/or creativity training during the period 2008–2010, 0 if not.	0.41	0.49	0	1
STI internal	1 if the firm used internal expenditures on R&D during the period 2008 – 2010, 0 if not.	0.24	0.43	0	1
DUI external	1 if the firm has collaborated or interacted with suppliers, customers, competitors, and/or partners within the conglomerate to promote innovation during the period 2008–2010, 0 if not.	0.11	0.31	0	1
STI external	1 if the firm collaborated with laboratories, research institutes, universities, and/or consultants to promote innovation during the period 2008–2010, 0 if not.	0.10	0.30	0	1

Table 3. Regression results, model [1].

	New-to-market product innovation
New-to-market prod. innovationt-1	1.044***
	(0.136)
DUI internal	1.149***
	(0.102)
STI external	−0.111
	(0.197)
STI internal	1.887***
	(0.116)
DUI external	1.220***
	(0.173)
Firm size (log)	−0.282***
	(0.049)
Firm age (log)	−0.023
	(0.117)
Share of edu.emp. (log)	0.555
	(0.378)
One main plant	−0.151
	(0.175)
Export focus	−0.063
	(0.132)
Constant	−3.368***
	(0.817)
Observations	5,997
Pseudo R-squared	0.386

Robust standard errors clustered over firms in parentheses. All models include industry and regional fixed effects. *** p < 0.01, ** p < 0.05, * p < 0.1.

Table 4. Regression results, model [2] Dependent variable: New-to-market product innovation.

	(1)	(2)	(3)	(4)
New-to-market product innovation t-1	1.051***	1.030***	1.041***	1.040***
	(0.136)	(0.134)	(0.136)	(0.134)
DUI internal	1.267***	1.507***	1.199***	1.534***
	(0.107)	(0.132)	(0.105)	(0.132)
STI external	−0.070	−0.071	0.294	−0.373
	(0.192)	(0.191)	(0.321)	(0.418)
STI internal	1.860***	2.421***	1.875***	2.357***
	(0.116)	(0.172)	(0.116)	(0.186)
DUI external	1.821***	1.210***	1.223***	1.759***
	(0.275)	(0.168)	(0.172)	(0.364)
DUI ext. * DUI int.	−0.795***			−0.713*
	(0.287)			(0.407)
STI int. * DUI int.		−0.807***		−0.726***
		(0.194)		(0.212)
STI ext. * DUI int.			−0.504	0.406
			(0.325)	(0.466)
Controls	Included	Included	Included	Included
Observations	5,997	5,997	5,997	5,997
Pseudo R-squared	0.387	0.389	0.386	0.389

Table 5. Wald test for complementarity or substitution effects.

	Wald test Adj. P-value for H0:	Sign of the linear combination
DUI ext. * DUI int.	7.64***	<0
	0.005
STI int. * DUI int.	17.25***	<0
	0.000
STI ext. * DUI int.	2.40	<0
	0.121

Tests conducted on marginal effects. Adjusted p-value of inequality tests for Wald χ² statistics with 1 degree of freedom. Negative values indicate substitution effects, while positive values indicate complementarity.

Table 6. Regression results, consistent innovators and innovation-active firms.

	Previous innovators	Innovation-active
	(1)	(2)
DUI internal	1.464***	0.451***
	(0.218)	(0.150)
STI external	0.074	−0.225
	(0.472)	(0.349)
STI internal	1.734***	0.856***
	(0.259)	(0.181)
DUI external	1.471***	1.146***
	(0.423)	(0.287)
DUI ext.* DUI int.	−0.205	0.312
	(0.524)	(0.394)
STI int.* DUI int.	−0.351	0.277
	(0.295)	(0.211)
STI ext.* DUI int.	−0.441	−0.279
	(0.471)	(0.330)
Observations	2,252	2,874
Pseudo R-squared	0.325	0.221

Robust standard errors clustered over firms in parentheses. All models include controls, industry and regional fixed effects. *** p < 0.01, ** p < 0.05, * p < 0.1.

Table 7. Regression results, new-to-firm innovation.

	New-to-firm product innovation		New-to-firm process innovation
	(1)	(2)	(3)	(4)
New-to-firm innov. t-1	1.105***	1.101***	0.891***	0.883***
	(0.113)	(0.112)	(0.107)	(0.105)
DUI internal	1.241***	1.490***	1.082***	1.430***
	(0.094)	(0.114)	(0.097)	(0.120)
STI external	−0.259	−0.575	−0.050	−0.125
	(0.215)	(0.427)	(0.174)	(0.384)
STI internal	1.918***	2.229***	1.228***	1.739***
	(0.111)	(0.175)	(0.114)	(0.196)
DUI external	1.378***	2.054***	1.172***	1.605***
	(0.187)	(0.346)	(0.156)	(0.324)
DUI ext. * DUI int.		−0.944**		−0.558
		(0.401)		(0.363)
STI int. * DUI int.		−0.499***		−0.754***
		(0.203)		(0.216)
STI ext. * DUI int.		0.524		0.161
		(0.488)		(0.425)
Controls	Included	Included	Included	Included
Observations	6,024	6,024	6,004	6,004
Pseudo R-squared	0.408	0.410	0.254	0.259

Robust standard errors clustered over firms in parentheses. All models include industry and regional fixed effects. *** p < 0.01, ** p < 0.05, * p < 0.1.