This study employs the comparative method to systematically assess alternative explanations for PPB practices adopted in different countries (regions), and to examine effects of these practices on spending behaviors. The comparative method specifies the conditions under which one country can learn from another, helping researchers find new policy options in other countries and discover latent policy constraints and opportunities.
The selection of countries (regions) being compared is based on purposive sampling, with the United States as a highly developed country (GDP per capita of $44,197), Taiwan as a newly developed region (GDP per capita of $16,442), and China as the largest developing country (GDP per capita of $3,250).
where O is an observation of spending in a specific year, P represents the implementation of PBB, and the ellipses represent any possible observations of spending. This design allows multiple observations of spending, resulting in a hypothesis that spending variations after PBB indicate the possible impact of PBB. Multiple observations also allow a measurement of possible intermediate- and long-term impacts of PBB. In the assessment, government spending is measured by the spending growth rate and the spending level in relation to government revenues and GDP, defined as follows:
PBB practices should help control spending growth, so that a negative relationship between PBB and SGR is expected. DSL is a measure of the spending level in relation to revenues and GDP. A positive relationship between PBB and DSL is expected, indicating that PBB should increase the surplus or reduce the deficit.
This study subscribes to the normative economic theory of supply and demand to develop the empirical model for analysis, and it also adopts the prescriptive public administration theory in public spending (incrementalism) in selecting the control variables. Clearly, government spending is affected by numerous socioeconomic and organizational factors that should be controlled in the model detecting the impact of PBB. These factors can be classified as those that create societal demands for public spending, as well as those on the supply side (the government) that produce the capacities to meet the demand. Thus, spending (S) can be specified as a function of societal demands (D) for public programs and government’s capabilities to meet the demand (C): S = f (D, C)
Identification and inclusion of all of the supply and demand factors in a model is impossible and does not serve the exploratory nature of this study. After all, the main purpose of the analysis is to explore the relationship between PBB and budgetary outcomes by controlling socioeconomic and financial factors in a given political-economic context, rather than to construct a perfect model to explain expenditure levels or budget deficits. With the clarity and parsimoniousness of the model in mind, we use GDP as a proxy for societal demand factors and revenue growth as a proxy for supply factors.
This study also employs the prescriptive public administration theory in public spending, which holds that public expenditure is the result of previous spending behaviors (incrementalism). In considering past spending behaviors, multiple observations in expenditure growth in the past are included in the model. Therefore, a time-series statistical model is constructed for the United States, Taiwan, and Guangdong, respectively, for the period 1980–2006, which covers the time span of PBB, specified as:
Findings and Discussion
Table 2 shows regression results of the spending growth rate for the three countries (regions) in the analysis. Results using data for different time frames in parameter estimation of the models are similar, suggesting that the models are robust. The F tests of the models indicate that all of the models are statistically significant at the .1 level. Tests show that models do not suffer severe conditions of multicollinearity (all VIFs ≤ 2.195). The Durbin-Watson statistics and related tests indicate that the models may not suffer autocorrelation.
Table 2. Regression Results for the Spending Growth Rate
All regression coefficients for PBB are negative, showing the expected signs of direction for PBB impact. Also, the results indicate that PBB may have a significant impact on spending growth in Taiwan, where the regression coefficient is statistically significant at the .05 level. A significant relationship between PBB and spending growth in Taiwan suggests that PBB restrains spending growth. In the United States and Guangdong, however, the results do not provide evidence that PBB has a significant impact on the spending growth rate. These results indicate that the PBB impact varies, depending on purposes and implementation strategies. In Taiwan, as discussed earlier, a clear goal of spending control was developed for PBB and the corresponding strategies of spending control were implemented vigorously, while the U.S. adoption of PBB in spending control, though intended, is not fully enacted in practice. PBB in Guangdong was not purposely linked to spending control.
Table 3 shows the results of regression on the deficit or surplus level. The models are robust after repeated tests using data from different time frames to estimate the models’ parameters. The F tests indicate that models for the United States and Taiwan are statistically significant at the .001 level, while the model for Guangdong is statistically significant at .107. Tests show that the models do not suffer severe multicollinearity (all VIFs ≤ 3.004). The Durbin- Watson statistics and related tests indicate that the models may not suffer autocorrelation.
Table 3. Regression Results for the Deficit or Surplus Level
Results for the deficit or surplus level are equally diverse as those for the spending growth rate. PBB may reduce the deficit or surplus level in the United States (p < .001), but such impact disappears in Taiwan and Guangdong. Interestingly, the results show a positive relationship between PBB and the deficit or surplus level in Guangdong. This suggests that the implementation of PBB propels, not reduces, the spending level in relation to revenue and GDP, or that PBB reduces the revenues collected in relation to expenditures. Either way, PBB appears to increase the budget deficit in Guangdong.
The findings seem to provide evidence that PBB has gradually become an essential part of administrative practice in the U.S. federal government, indicating that the performance-based reform strategy may affect the deficit level. A recent study shows that a relatively large number of federal managers use performance information in making resource allocation decisions, and they have adopted performance-based practices in management (Newcomer 2007). Although these practices have little impact on curtailing expenditure growth, they may help federal managers make funding decisions more cautiously within revenue constraints.
In Taiwan, the administration’s request for a comprehensive PBB reform has placed a very stringent restriction on expenditure growth, providing incentive for agencies to restrain discretionary spending, which explains the negative relationship between PBB and the expenditure growth rate. However, this finding should be read with caution, because some supplemental and special spending items (e.g., some items for disaster response and national defense) are not included in the expenditure calculation and actual expenditure levels can be higher.
PBB has been in place for a short time in Guangdong, and it has been experimented with in only a few selected agencies, which may explain why it has no impact on expenditure growth. The experimental agencies, often large in size and operation (Niu, Ho, and Ma 2006), may take advantage of PBB to advocate their achievements and to request larger budgets. This may explain why PBB may increase the deficit or reduce the surplus level.
The foregoing findings indicate that the PBB impact varies across countries and regions that have different PBB goals and implementation strategies. Legislative leadership support in form of a formal legislation, as in the United States and Taiwan, appears to be critical in PBB implementation. Passing a law and specifying how agencies should follow it provides public officials with a legal basis and responsibility for implementation. PBB legislation reinforces political urgency and underscores the importance of the reform, and it also provides legitimacy for requests and approvals of financial support for PBB, which is critical in PBB implementation and performance-based funding.
Strong executive leadership is critical to substantiate and sustain PBB impact. Executive leadership in PBB reflects in enforcing the law and carrying out administrative responsibilities in PBB implementation. Top executives, who have strong political connections with legislators and knowledge of management and daily operations, are catalysts in obtaining resource support and acquiring the necessary technical capabilities for PBB (e.g., hiring capable performance analysts and developing a good performance information system).
Strong intent and effort to link performance results with resource allocation decision making, as shown in Taiwan, also appears important for the PBB impact on spending. The literature indicates that performance measurement is widely used in budget preparation and examination, but less so in funding decisions by the legislature for a variety of technical and political reasons, including concerns about the validity of the measurement system and fear that performance funding is a threat to the political power structure. The threat of these obstacles should be recognized and eliminated in order for PBB to influence spending.
Implementation length may also play a role in realizing the PBB impact. As the U.S. and Taiwan experiences show, PBB is an exhaustive assessment process demanding intensified effort in time and paperwork to define performance, develop data collection systems, conduct analysis, and evaluate the results. Accounting and audit systems should provide the information needed for assessment. Also important is the creation of positive incentives to encourage compliance with and support for PBB. Clearly, the development and improvement of these necessary elements in PBB take time, and the maturation of a PBB system may indicate the development of an organizational culture that is more accustomed to performance funding.