Information Technology and Administrative Efficiency in U.S. State Governments: A Stochastic Frontier Approach

Unlike in for-profit business, where the value of IT has been studied by a large body of literature, in public sector, whether IT makes contribution to government performance is unknown. Therefore, this paper examines the relationship between IT spending and government performance. However, due to the different nature of firm and government, the performance measurements widely used in prior literature based on production function become invalid in the context of government. So this paper specifically investigates whether IT investment is associated with cost efficiency of state governments.

To measure the dependent variable, cost efficiency, the paper employs a stochastic frontier analysis with cost function. This method derives a cost frontier, which describes a minimum level of inputs, given certain amounts of outputs. After that, the cost efficiency is regressed respectively on two IT spending measurements with a two-year lag in a fixed-effect regression model, along with multiple control variables. Furthermore, the paper tests three moderators on the efficiency returns to IT spending. Those three moderators represent contextual effects from economic aspect, demographic aspect, and political aspect.

The results indicate IT spending is positively and significantly associated with government cost efficiency. This main conclusion is consistent through robustness checks. The paper mentions that, numerically, one dollar IT spending will lead to 1.13 dollar cost saving, on average. In addition, three moderating effects are all verified, indicating that the efficiency return to IT spending will be impacted by some environmental factors.

The authors theorize that the national politics significantly affects IT investments in the federal government we discuss how the U.S. Congress influences federal IT investments.  They hypothesize that a federal agency’s capacity-building IT investments are associated with (i) legislative approval for the chief executive, (ii) government dividedness, and (iii) the agency’s ideological characteristic. They adopt a panel dataset from 135 federal agencies and bureaus in 2003-2016, the empirical analyses support all the hypotheses. It is demonstrated that the national politics has a significant impact on their IT investment profiles. A federal agency is more likely to make capacity-building IT investments when its chief executive is blessed with legislative approval, when the federal government is more united, and when it is ideologically more moderate They suggest in order to make more capacity-building IT investments, a federal agency needs more policy directives, greater political support and legitimacy, and sufficient resource endowment from Congress.

However, they did not analysis the effectiveness of IT investment or any other outcome variables of IT investment  in the study.

we discuss how the U.S. Congress influences federal IT investments.  They hypothesize that a federal agency’s capacity-building IT investments are associated with (i) legislative approval for the chief executive, (ii) government dividedness, and (iii) the agency’s ideological characteristic. They a panel dataset from 135 federal agencies and bureaus in 2003-2016, the empirical analyses support all the hypotheses. It is demonstrated that the national politics has a significant impact on their IT investment profiles. A federal agency is more likely to make capacity-building IT investments when its chief executive is blessed with legislative approval, when the federal government is more united, and when it is ideologically more moderate They suggest in order to make more capacity-building IT investments, a federal agency needs more policy directives, greater political support and legitimacy, and sufficient resource endowment from Congress. 

IT governance and business value in the public sector organizations –The role of elected representatives in IT governance and its impact on IT value in U.S. state governments

Pang (2014) examines the moderating effect of IT governance on the relationship between IT spending and performance, measured by cost efficiency, in U.S. state governments. Different from prior studies, Pang (2014) studies the impact of IT governance on performance in the public sector, which is the new setting in IT governance research, meanwhile Pang (2014) focuses on the principal side of IT governance effect of legislative controls on IT management and investments.

Pang (2014) presents the following hypotheses:

H1: All others being equal, the association between IT spending and cost efficiency is stronger in states whose legislature has an IT-related legislative committee than in ones without such a committee.

H2: All others being equal, the association between IT spending and cost efficiency is stronger in states whose CIO position is established by legislation than otherwise.

H3: All others being equal, the association between IT spending and cost efficiency is stronger in states whose legislature approves a state CIO nominee than otherwise.

Pang (2014) adopts two-stage estimation approach in this study. First, cost efficiency of each state-year observation estimated by a stochastic frontier model. Second, the estimated cost efficiency regressed on IT spending and governance measures and control variables.

There are several findings in this study: (1). The empirical analysis results show that IT spending in U.S. state governments has positive effects on cost efficiency in state governments. (2). Supervision of state legislatures on IT management and approval of a CIO appointee complement each other. The main contributions include: First, this research contributes to the literature on IT business value and IT governance. Second, a new finding presented to the IS literature that the involvement of principals in IT management helps organizations utilize IT resources more effectively. Third, this study contributes to the literature by recombining research from two distant literatures (IS and political sciences). Finally, this study has several important managerial implications for the public sector organizations.

Policy officers safeguard the public, but who protects our guardian? A working paper coauthored by Dr. Pang and Dr. Pavlou gives an interesting answer, information technology. This research investigated how IT could prevent violence against police officers. Specifically, the authors examined the relationship between IT use by the police and the number of police officers killed or assaulted in the line of duty.

Integrating the literature on IT-enabled organizational capabilities with the criminology, they theorized two mechanisms through which police IT use reduce violence against police officers by developing two key law enforcement capabilities, intelligence-led policing and community-oriented policing. The usage of three kinds of IT, analytics technologies, real-time response technologies and the internet, which facilitate such capability building, was hypothesized to make police officers safer.

To test the hypotheses, the authors utilized a large-scale dataset with 3921 police departments in 2 years combining police IT use, other operational information and annual crime statistics and public safety data. Random-effects models, along with negative binomial regressions and spatial auto-correlation models for validity check, provide a consistent and robust results that IT use for crime analysis, dispatch, and the internet is significantly associated with a decreased in the deaths of police officers. This effect is shown to be more pronounced in communities with a higher economic divide.

The paper initiatively builds a link between IT capabilities in organizations with criminology, contributing to the nascent literature on business value of IT in the public sector and broader societal impact of IT. Most importantly, the finding of this paper could be arguably generalized to the significant role of IT in safety of other occupations under unpredictable dangers.

The authors examined whether IT improves administrative efficiency in U.S. state governments. This empirical study contributes to the IS literature by expanding the scope of IT value research to the public sector.

To be specific, U.S. state governments spent approximately 5% of the tax revenues in IT in 2004. On the other hand, for-profit firms used 3.6% of the sale revenue on IT in 2004. This fact enhances the necessity for policy makers to keep their eyes on the IT spending in public organizations. Two important factors motivated the authors to study: 1) the public sector is different from the private business context, and 2) the approach to measure the value effect of IT investments in the private sector may not be appropriate in the government context. In particular, for-profit context uses production function framework, but the public sector context is more suitable to use cost function.

For the analysis, the authors estimated the impact of IT investments on administrative efficiency with a two-stage estimation approach based on a multi-product translog cost function. In the first stage, they estimated cost inefficiency with a stochastic frontier model, and in the second stage, the estimated cost inefficiency was regressed on IT intensity and other exogenous factors by reflecting on the fact that the government outputs are exogenously given. As government agencies are expected to deliver necessary public service within a budget collected from tax, efficiency benefits of IT is more applicable in the public sector context than adopting the effect of IT on costs of goods sold. The applicable area of cost efficiency is not limited to human resources but opens to services, which is affected by automating organizational process by focusing on the automate and informate roles of IT.

Pang and Pavlou (2016) studies the role of IT in preventing violence against police officers. Specifically, it examines the relationship between IT use by the police and number of police officers killed and assaulted.

The study theorizes that IT use by police helps to develop two kinds of capabilities—intelligence-led policing and community-oriented policing. First, the intelligence-led policing capabilities can help police to identify and solve crime, reduce the chance of violent encounters, and enable control crimes more proactively. Second, a cooperative relationship with the community can effectively control crime, and ensure police officers’ safety. Thus, both the intelligence-led policing and community-oriented policing leads to a reduction in violence against police officers. Then the paper proposes that the use of analytics technology, real-time response technology, and the Internet helps police to develop these IT capabilities.

Empirically, this study used panel data set collected from multi sources (such as LEMAS and UCR). The estimated model uses each functionality of the technology as a variable. The paper estimates the model with random-effects and uses negative binomial regression and spatial autocorrelation models for robustness checks. Also, the paper examines the moderating effects of income inequality and racial disparity. The results show that IT use for crime analysis, dispatch, and the Internet is associated with lower killings of police officers. At the same time, dispatch and in-filed report are associated with fewer assaults of police officers.

This paper provides innovative ideas to combine IT capabilities in organizations with criminology. It empirically demonstrates how different technologies are associated with the safety of the police officers. Also, it expands the current IS literature of IT business value in the public sector.

Politics and Information Technology Investments in the U.S. Federal Government in 2003-2016

Motivation：

The U.S. federal government spends a considerable amount of tax revenues in IT every year. Hence, taxpayers, who pay for these IT expenditures, would ask whether these investments are being well made. However, information systems (IS) researchers to date have paid little attention to what influences IT investments and management in the government sector (reference anonymized for blind review), an issue that would be significant to policy makers and the public in general as well as private-sector IT industries.

Research Questions:

This study investigates how the national politics affects IT investment profiles in U.S. federal agencies. They hypothesize that a federal agency’s capacity-building IT investments are associated with (i) legislative approval for the chief executive, (ii) government dividedness, and (iii) the agency’s ideological characteristic.

Main Findings:

With a panel dataset from 135 federal agencies and bureaus in 2003-2016, the empirical analyses demonstrate that the national politics has a significant impact on their IT investment profiles. They find that a federal agency is more likely to make capacity-building IT investments when its chief executive is blessed with legislative approval, when the federal government is more united, and when it is ideologically more moderate. In sum, their study supports the central proposition that in order to invest more in major capacity-building IT development, the federal agencies need to secure compelling policy mandates and political legitimacy from Congress for implementation of strategic policy initiatives..

Contributions:

They contribute to the IS literature by demonstrating that budget allocation decisions between IT development and maintenance in governments are affected by political environments. They offer several policy prescriptions in federal IT management for policymakers and practitioners in the public sector.

Pang, Tafti, and Krishnan (2015) conjoin multiple perspectives of the Information system, Public economics, and Political science literatures to examine the relationship between IT and governmental size. More specifically, the authors attempt to reveal the impact of IT budgets of a state CIO on state government size.

Theoretically, we can expect both positive and negative relationships. On the one hand, as the Hypothesis 1A argues, “the size of CIO IT budgets in a state is associated with lower state government spending.” This argument can be supported in that the introduction and advancement of IT systems may 1) increase the productivity of administrative processes, 2) ameliorate information asymmetry between legislatures and governmental agencies with lower monitoring costs, and 3) decrease the costs of transaction and coordination by more privatization of public services. Collectively, these benefits facilitate the reduction of government size.

On the other hand, the competing Hypothesis 1B contends that “the size of CIO IT budgets in a state is associated with larger state government spending.” Its main rationales are, 1) the IT infrastructure and enterprise systems may create strategic initiatives to serve public areas which used to be underserved; 2) the enhanced capabilities to monitor and coordinate administrative processes can lead to push further the boundary of administration, even incorporating some roles of federal, local, and private players.

Based on system GMM model with a five-year unbalanced panel data of 190 observations from 44 states, the empirical test supports the Hypothesis 1A, implying that more IT investment by state CIOs are associated with lower state governmental expenditures. This main finding suggests that the increase in the IT budgets by state CIOs may mitigate “bureaucrats’ interests to maximize their expenditures.”