Association for Public Policy Analysis & Management

What makes some federal agencies in the U.S. digitally advanced and others lagging? Does politics matter to information technology (IT) investments in the federal government? Increasingly, IT has become an indispensable means to implement any major contemporary public policy initiatives, as evident in the cases of the Affordable Care Act or the Wall-Street reform. Yet, we do not have a full understanding of how the federal government makes IT investments, which are expected to provide necessary organizational capabilities for policy implementation and execution. We also witness that some agencies actively invest in new, capacity-building IT systems, while others continue to spend in maintaining decades-old legacy systems that are inflexible and expensive to maintain. This study investigates how the national politics affects IT investment profiles in U.S. federal agencies.

Drawing upon various theories from political sciences, public administration, and information systems disciplines, we propose a theory of IT bureaucracy, in which we posit that without sufficient policy input and strong mandates from Congress, federal agencies would not voluntarily invest in risky IT development for new organizational capabilities. This is because new IT development is inherently risky and prone to project delays, technical failures, and cost overrun, as vividly illustrated by in the case of Healthcare.gov. This deters risk-averse bureaucrats with concerns for career prospects from proactively invest in new IT development. Based on this theoretical model, we hypothesize that a federal agency makes more capacity-building IT investments and spend less in IT maintenance (i) when it performs homogenous functions, (ii) when its head is appointed with legislative approval, (iii) when the federal government is less divided, and (iv) when the agency is neither too conservative nor too liberal.

With a panel dataset from 133 federal agencies and bureaus in 2003-2015, our empirical analyses support our hypotheses and produce several intriguing findings. For instance, when both the Senate and the House of Representatives are controlled by the ruling party, federal agencies are predicted to invest approximately 4.19%-point more in new IT development and modernization than when the opposition party holds the majority in both chambers. We also find that when the head of a federal bureau is not confirmed by the Senate for more than one year or does not require a Senate confirmation, the bureau spends 5.91%-point less in new IT development. It is predicted that federal agencies that are slightly conservative make more capacity-building IT investments than other more conservative and more liberal ones.

We contribute to the literature in two fronts. We study IT investments profiles in the U.S. federal government, which consumes more than $75 billion of tax revenues for IT annually yet has received scant attention by prior literature. We also examine what affects budget allocation decisions between IT development and maintenance. While IT managers in both the public and the private sectors struggle to maintain a right balance between the two with a limited amount of IT budgets, to the best of our knowledge, few prior studies so far have studied this important issue.