Skip to main content

Table 9 Technological improvement rates and breakthrough innovations: different breakthrough measures

From: Digital technologies, technological improvement rates, and innovations “Made in Switzerland”

Dependent variables

FWD75

FWD90

FWD99

IMP75

IMP90

IMP99

Model

(1)

(2)

(3)

(4)

(5)

(6)

Digital

3.792***

2.991***

0.7378***

7.832**

5.470***

0.8458***

(0.9897)

(0.6914)

(0.1424)

(3.359)

(1.787)

(0.2454)

Non-digital: k

0.2260***

0.1480***

0.0247***

0.4444***

0.2831***

0.0331***

(0.0330)

(0.0221)

(0.0034)

(0.1177)

(0.0644)

(0.0081)

Digital: k

0.0387***

0.0228***

0.0035***

0.1482***

0.0755***

0.0095***

(0.0050)

(0.0043)

(0.0012)

(0.0114)

(0.0094)

(0.0027)

Control variables

inv

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

num_ctry

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

Claims

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

bwd_cits

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

Fixed effects

Industry

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

Year

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

Year-Industry

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

  1. All estimations include USPTO patents of the filing years 1990–2015 if at least one inventor has been from Canada, China, France, Germany, Japan, South Korea, Switzerland, the UK, or the USA. Similar to the main text, the first three specifications (1)–(3) use the number of forward citations as an indication of breakthrough innovations. The first model specification indicates a breakthrough innovation if the number of a patent’s five-year forward citations is among the top 25%, the second specification is similar to the used specifications in the main text, and the third specification if the number of a patent’s five-year forward citations is among the top one percent. The other three specifications (4)–(6) consider an importance measure based on patent text similarity as proposed by Kelly et al. (2021). For details, see the accompanying text. In particular, the specification of column (4) indicates a breakthrough innovation if the patent’s importance measure is among the top 25%, the specification of column (5) if it is among the top 10%, and the one of column (6) if it is among the top one percent. In order to show associations in percentage points, we have multiplied the FWD/IMP-dummies by 100. The industry fixed effects are at the level of the six industries included in the analysis, namely Computer/ICT, Machinery, Medical, Electrical, Chemical, and Pharma. The standard errors in parentheses are also clustered at this industry level. Significance levels for the coefficients are indicated as: \(p^*<0.1\); \(p^{**}<0.05\); \(p^{***}<0.01\)