Table 3.

Description of model variables and proxies

Variable Type	Variable Name	Variable Generation	Indicator Description and Assignment Definition
Outcome Variables	Standardized sales	Yp	Quarterly sales of drugs of the same generic name in major cities after standardization
	Standardized sales volume	Yn	Quarterly sales volume of drugs of the same generic name in major cities after standardization
	Standardized reporting volume	Yn0	Standardized reporting volume of public announcements on the Sunshine Purchasing Network in pilot regions
Core explanatory variables	Control variables	Treatm	Treatm Pilot = 1; Non-pilot = 0
	Event variable	Postt	Post-intervention = 1; Pre-intervention = 0
Moderating variables	City Virtual Variables	m	M-1 Virtual variables to describe city FEs, M = 24
	Common Name Virtual Variables	i	I-1 Virtual variables to describe drug FEs, I = 20
	Time Virtual Variables	t	The study uses panel data from 2014–2018 (T = 5) with a cross-section of 24 cities and T-1 Virtual variables describing year FEs
	Quarter Virtual Variables	q	Q1 2018—Q4 2019 q = -4, -3, -2, -1, 1, 2, 3
Other explanatory variables	Number of outpatient consultations	Xtreat	Annual number of outpatient and emergency services visits to medical institutions in major cities (billions of visits)
	Per capita expenditure on medicines under medical insurance	Xoutcome	Per capita expenditure on medicines covered by basic medical insurance (yuan)
	Consumer price index for western medicines	XCPI	Retail price index of urban commodities in the western medicine category (previous year = 100)

(1) in order to avoid the influence of heteroscedasticity, the natural logarithm of the original data set was used; (2) the standardization of the outcome variable "volume" refers to the calculation of the minimum formulation unit (DDD) as the basic unit of multiplication; (3) the standardization of "price" is the volume multiplied by the spread coefficient to convert to the standardized price