Table 1 DID regression of a topic’s comparative post-prizewinning growth on six measures.

From: Scientific prizes and the extraordinary growth of scientific topics

Growth measures

(1)

(2)

(3)

(4)

(5)

(6)

Productivity

Citations

Impact of topic’s lead scientists

#Incumbents

#Entrants

#Disciplinay stars working on the topic

Prizewinning (\({\beta }_{1}\))

0.005

−0.002

0.002

0.003

−0.005

0.004

   S.E.

(0.0387)

(0.0442)

(0.0259)

(0.0408)

(0.0413)

(0.0362)

   p-value

(0.8979)

(0.965)

(0.9295)

(0.9444)

(0.9042)

(0.9176)

Post (\({\beta }_{2}\))

0.609***

1.240***

1.047***

0.973***

0.750***

0.667***

   S.E.

(0.0115)

(0.0151)

(0.00970)

(0.0135)

(0.0117)

(0.0119)

   p-value

(0.0000)

(0.0000)

(0.0000)

(0.0000)

(0.0000)

(0.0000)

Prizewinning * Post (\({\beta }_{3}\))

0.213***

0.169***

0.188***

0.271***

0.206***

0.254***

   S.E.

(0.0134)

(0.0172)

(0.0128)

(0.0157)

(0.0138)

(0.0137)

   p-value

(8.207e-57)

(8.735e-23)

(3.447e-48)

(2.928e-66)

(2.126e-50)

(9.208e-76)

Fixed effect controls:

   Discipline

Yes

Yes

Yes

Yes

Yes

Yes

   Year

Yes

Yes

Yes

Yes

Yes

Yes

Const

2.926***

4.502***

4.237***

1.136***

3.217***

1.358***

   S.E.

(0.123)

(0.136)

(0.0988)

(0.135)

(0.131)

(0.117)

   p-value

(2.860e-124)

(3.571e-236)

(0)

(3.475e-17)

(7.454e-131)

(7.269e-31)

N

1,453,914

1,453,914

1,453,914

1,453,914

1,453,914

1,453,914

R-sq

0.154

0.301

0.392

0.290

0.209

0.274

  1. Prizewinning topics are associated with extraordinary growth. Prizewinningi is a dummy variable quantifying whether the topic \(i\) is a prizewinning topic or a non-prizewinning topic from the matched groups. \({{{{{{{\mathrm{Post}}}}}}}}_{t}\) is a dummy variable measuring whether time \(t\) is before or after the prizewinning event. If the topic belongs to the matched group, the prizewinning year of the related prizewinning topic will be used as the reference point. Regression is specified with fixed effects for discipline and prize year and robust standard errors. All estimated \({\beta }_{1}\)s have p-values > 0.05, indicating that prizewinning and matched topics have no differences before the prizewinning event. \({\beta }_{3}\)s are significant for all six categories (all p-values < 0.001 even after adjustments were made for multiple comparisons), demonstrating that prizewinning topics grow unexpectedly larger than matched topics after the year the prize is conferred. OLS models are used to perform the DID analysis.
  2. Standard errors in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.
Back to article page