Vladimir Putin on Channel One, 2000–2022

Figures & data

Figure 1. Corpus from Channel One.

Country labels are assigned with Newsmap (Watanabe, 2018).

Table 1. Sentiment analysis: classification results.

1.1. DV: Sentiment score
	(1)	(2)	(3)
Putin	0.343***	0.305***	0.230***
	(0.003)	(0.003)	(0.003)
Domestic news		0.197***	0.193***
		(0.002)	(0.002)
Year dummies			yes
Intercept	−0.066***	−0.169***	−0.442***
	(0.001)	(0.002)	(0.140)
R-sq	0.025	0.044	0.048
Observations	385,981	385,981	385,981
1.2. DV: Putin (binary)
	(4)	(5)	(6)
“Neutral” dummy	1.018***	0.941***
	(0.017)	(0.017)
“Positive” dummy	1.309***	1.154***	1.181***
	(0.018)	(0.018)	(0.018)
Domestic news		0.732***	0.546***
		(0.011)	(0.017)
Year dummies		yes	yes
Intercept	−2.946***	−0.044	−0.272
	(0.016)	(0.642)	(1.190)
Pseudo R-sq	0.024	0.067	0.089
Observations	385,981	385,981	183,653

(1), (2), and (3) represent the estimates of the OLS coefficients in the models where the dependent variable is the sentiment score estimated by Rubert-tiny. ***p < .001. (4), (5), and (6) provide the estimates of the logistic regression coefficients in the models where the dependent variable equals one if a news episode mentions Putin and zero otherwise. ***p < .001. (4), (5), and (6) use negative sentiment labels as the reference category. (6) is estimated based on a dataset that excludes news items labeled as “neutral.” The dummy variables for “Neutral” and “Positive” are based on the class labeling described in the section The Sentiment of News Reports.

Figure 2. Sentiment estimates (one-dimensional estimates of a sentiment score).

Figure 3. The daily share of news reports that mention Vladimir Putin.

Table 2. Daily references to Putin.

	(1)	(2)
Term 2	−0.036***	−0.030***
	(0.003)	(0.003)
Prime minister	−0.068***	−0.051***
	(0.003)	(0.003)
Term 3	−0.011***	0.010***
	(0.003)	(0.003)
Term 4	0.061***	0.068***
	(0.003)	(0.003)
Foreign episodes, daily		−0.001***
		(0.00)
Intercept	0.134***	0.146***
	(0.002)	(0.002)
R-sq	0.189	0.202
Observations	8,092	8,092

***p < .001. The table reports the estimates of the OLS coefficients in the models where the dependent variable is the share of news episodes that mention Vladimir Putin in the daily news flow from Channel One. The Augmented Dickey–Fuller test was conducted to assess the stationarity of the time series data (ADF = −5.506, p = 2.015). The test results indicate evidence suggesting that the time series is stationary.

Figure 4. The shares of all domestic and foreign news reports that mention Vladimir Putin.

Figure 5. The share of foreign news in all the stories that mention Vladimir Putin.

Figure 6. News reports about foreign affairs and foreign events that mention Vladimir Putin.

Watanabe, K. (2018). Newsmap: A semi-supervised approach to geographical news classification. Digital Journalism, 6(3), 294–309. https://doi.org/10.1080/21670811.2017.1293487

 Web of Science ®Google Scholar