Appendix: Chapter 4 Coding of Chinese Public Compensation

This section reports my method for coding actual compensation rates from the Shandong county line-item budgets, as analyzed in Chapter 4. Existing studies employ government budgets that are published in statistical yearbooks and government websites, which only list budgets by broad spending categories (e.g., education, health, construction). The category of “administration costs” reported in these budgets in fact only partially captures total administrative costs. The line-item budget data that I use has the unique advantage of listing government spending by items for each county in Shandong province. Table A4.1 shows a sample of the data structure (unit = 10,000 Yuan). From the line-item budgets, we can tell how much each county spent on formal salaries, bonuses, allowances, travel, etc.

This data source provides a valuable opportunity to estimate actual compensation levels, including both formal salary and fringe benefits and pay. Measuring formal salary (jiben gongzi) is straightforward, as there is a single, clearly defined column for this item. There isn’t a single column, however, for “fringe compensation,” as it is spread out over multiple line items within the reported budgets. To accurately code fringe compensation, I drew on technical budget manuals that explained the composition of each line item and interviews with more than 40 budgetary officials.

To construct fringe compensation, I included two main line items in the budgets: “personnel spending” and “administrative spending.” Personnel spending refers to government spending on financial remuneration for government employees beyond formal salary, such as bonuses, subsidies, and allowances. Administrative spending, on the other hand, captures the provision of in-kind (or non-monetized) staff benefits. For example, a wealthy tax bureau can build a lavish office or disburse gift baskets to staff members. Other well-documented perks include “entertainment,” “training,” and “conferences and meetings,” which commonly translate into free wining and dining and subsidized vacations for government employees. For a conservative estimate, I excluded administrative spending items that are not likely to contribute to staff welfare (e.g., purchase of specialized equipment and materials). Table A4.2 lists the budgetary line items I included to estimate fringe compensation in the Shandong county governments. Descriptive statistics are presented in Table A4.3.