University of Washington
This essay reviews recent literature on American economic policymaking. It first asks, what light can contemporary research on economic policymaking shed on current circumstances? And secondly, where do policy process researchers go from here? This paper explores the extant literature under three broad themes. The first will deal with the extensive research into income inequality. The second will discuss the electoral implications of economic conditions. The third will analyze the institutional responses to economic developments. The final section will offer suggestions for future scholarship.
To gauge the current state of American economic policymaking, one could examine the inability of the United States government to effectively deal with the myriad problems that led the ratings agency Standard & Poor's to downgrade U.S. debt in August 2011 (StandardandPoor's,2011).1 That action — the first time in history that a ratings agency has downgraded U.S. sovereign debt — followed intense brinkmanship over raising the statutory debt limit, a procedural mechanism that does not even directly affect U.S. debt.2 While the debt limit has long been an opportunity for partisan posturing (see, for example, Kingdon, 1989, 182), serious doubts about a proposed increase rarely arise. The brinkmanship mirrored earlier conflicts between the Republican-controlled House of Representatives and President Barack Obama over the budget, and later in 2011, expiring tax cuts.
The gridlock and inertia that has resulted from this brinkmanship highlights the new territory that scholars and policymakers alike have entered in recent years. There has been a simultaneous failure in both ideas and institutions. The financial crisis that precipitated the recession beginning in 2007 has not only challenged the dominant ideas of economics and policy (Ormerod, 2010; Krugman, 2009), but political polarization appears to have reached levels unprecedented since the Civil War (McCarty, Poole and Rosenthal, 2006), which limits the ability of politicians to deal effectively with ongoing economic problems. These Yearbook articles are syntheses of recent research on theoretical and substantive policy issues. To that end, this essay first asks, what light can contemporary research on economic policymaking shed on current circumstances? And secondly, where do policy process researchers go from here?
One way to examine the literature and challenges facing the U.S. economy is to view relatively recent developments as the culmination of nearly 40 years of economic decline, starting with stagnating middle-class wages and growing income inequality. These phenomena have been linked to growing political polarization and gridlock (McCarty, Poole and Rosenthal, 2006) and reduced political efficacy (Bartels, 2008). Research also shows that ideological orientation influences individual-level economic assessments (Gerber and Huber, 2009, 2010). When taken together, this research presents the mix of politics and economics as a dangerous elixir that combats successful policymaking and undermines democratic accountability.
Another story presented in the literature presents the collapse of the U.S. economy as a failure of ideas and bureaucratic oversight. Starting with the 1998 passage of the Gramm-Leach-Bliley Act that deregulated the financial services industry, the American economy became increasingly "financialized," or driven more by banking than the production of tangible goods and services. This development caught regulators asleep at the switch (Khademian, 2009) and necessitated large-scale reform in the Dodd-Frank Act of 2010. But this attempt at reform may be doomed before it is even fully implemented due to the partisan nature of its enactment and its dearth of a sustained reform coalition (Carpenter, 2010). Moreover, many of the same policymakers who lacked the foresight to see the long-term implications of deregulation were intimately involved in crafting re-regulation. Many economists involved in the policymaking process, especially, have been forced to do much soul-searching in the face of the practical failure of their theoretical models (Krugman, 2009; House Committee on Oversight and Government Reform, 2008). But some of that introspection may be short-lived (Greenspan, 2011).
In this essay, I will review the extant recent literature by dividing it into three broad themes. The first will deal with the extensive research into income inequality: its causes, effects, and efforts to combat it. The second will discuss the electoral implications of economic conditions. Electoral politics directly contour the policy debate, since as Jones and Jenkins-Smith (2009) argue, public opinion can be used as both a resource for and constraint on policy advocates. The third section will analyze the institutional responses to economic developments. The final section will offer suggestions for future scholarship.
Worth noting is that "economic policy" is amorphous. It could mean a great deal from budgeting and distributive policy to taxation. The effects of economic policy also have ramifications on a host of other policies, from social welfare and education to defense. Throughout this paper I employ the term "economic policy" broadly to mean large-scale policy solutions that seek to alter the macroeconomy. The interconnectedness of the global economy also means that debt crises around the world, and most recently in the European Union, affect domestic markets, trade opportunities, and national economic policy. In the United States, individual states face budget crises and rely on the Federal government during hard times, and state policy also influences the broader national discussion. Consistent with the Yearbook's mission, this paper focuses on the published literature on American economic policymaking, and I selected the themes of income inequality, political impact, and institutional response because of their dominant positions in that recent literature. As the last section will attest, there are plentiful avenues for future research on a host of subjects using multiple frameworks for analysis.
Income inequality in the United States has been a persistent problem that has only recently been thoroughly addressed as both an economic and political problem. As Figure 1 shows, those in the top one percent of all wage-earners have disproportionately benefitted from economic and policy conditions since the 1970s, while those in the bottom 90 percent have seen very small changes in their income. This disparity has led to a host of social problems, especially now that the economy has collapsed. As Hacker and Pierson (2010) describe, there are three leading causes for the increase in inequality. First, technological change has skewed income toward those better positioned to take advantage of the Internet and media (particularly those with at least a college education). Second, active policy manipulation of the upper tax brackets, lowering the effective marginal rate of taxation. And third, "policy drift" which leaves policy unchanged even when conditions change. For example, if technological change did indeed cause the spike in income inequality, policy failed to keep up and provide opportunities for more individuals.
Figure 1. Income levels of the top 1% of earners and the bottom 90% of earners, 1970-2008. Source: Saez and Piketty (2010).
In a series of articles analyzing poverty and income inequality, Besharov and Call (2009a,b) and Plotnick (2009) review what types of policies would be most beneficial given current circumstances. For example, fewer families in the United States are below the subsistence poverty line — the level at which basic needs such as food and shelter are met — but many suffer from a lack of income that makes them relatively impoverished. This relative poverty — the point at which a family can live "decently" — is described better by income inequality. Strategies to reduce this include the Earned Income Tax Credit, minimum wage, government-funded child care and the like. All of these are redistributive policies in that they involve taking income from top earners and giving it to low-income individuals and families. Besharov and Call (2009a) argue that these kinds of policies do not adequately address income inequality and that resources would be better directed to improvements in job training and human capital investment. They argue that direct income transfers, however, are more politically feasible, as they are simpler than an "active welfare state."
Plotnick does not disagree with Besharov and Call's conclusions but argues that absolute poverty statistics provide more insight into American economic conditions. Plotnick argues that children have been particularly hard-hit by a lack of poverty programs and that work-based support systems miss helping those closest to the poverty line. Plotnick argues that while universal anti-poverty programs are not politically feasible, there are changes that could be made to improve the conditions of the poor, including accurate background checks to ensure fairness in hiring and reduce statistical discrimination against African Americans in particular; automatic unemployment insurance extensions upon recessions; and better enforcement of child support mechanisms, to name a few.
While the policy solutions outlined above may help to bend the lower line in Figure 1 up with little change to the top line, any change to the social safety net will be up against harsh political pressures. And rightly, both Besharov and Call and Plotnick acknowledge that the politics of policy change will be difficult, especially as budgetary constraints increase with reduced revenue as a result of a declining tax base — a feature of most economic downturns. But the potential of recession-induced deficits are not the only political challenge to increasing the social safety net and reducing income inequality. Public support for these programs is not high, with the Americans tending to favor education spending instead (McCall and Kenworthy, 2009). Income inequality also tends to produce self-reinforcing tides of conservative mass opinion.
The obvious threat presented by income inequality to self-governance is that the wealthy will have undue influence over the government. This is the "unequal democracy" thesis presented by Bartels (2008). But this presupposes that the poor have divergent policy preferences from the wealthy, and indeed, the "economic model" of American politics promulgated since at least Downs (1957) would suggest that the poor would support policies that soak the rich with high marginal tax rates so that they could benefit from redistributive policies. But Kelly and Enns (2010) find that this divergence does not exist: greater income inequality reduces liberal/redistributive sentiment even among low income individuals.
Kelly and Enns seek to understand the extent to which politicians respond to the policy demands of the wealthy over those of the poor or if politicians respond proportionately. With regard to their findings of similar preferences across both higher and lower income individuals, they can offer only speculation as to the relevant causal mechanisms. They suggest that media frames since the 1970s have driven public opinion toward "individualism" and against welfare policies, creating a negative link. When welfare policies are needed in hard economic times, these media frames change accordingly, but the current link is a damaging one.
Figure 2. Trust in government (1958-2008) and the Gini index of income inequality with a lowess fit line. Sources: U.S. Census and American National Election Survey.
There is, however, an alternative explanation not explored by Kelly and Enns. Income inequality, negative change in real disposable income, and declining trust in government are correlated (see Figures 2 and 3). And as Hetherington (2005) finds, trust in government is related to support for welfare programs. When the public no longer sees the government as a trustworthy ally, it no longer views its intervention positively. Accordingly, liberal programs since the New Deal and the Great Society have been undermined with mass political support (Hacker and Pierson, 2010). Income inequality could add another fold to this story if citizens distrust government because they see it perpetuating inequality through other policies (such as tax cuts or bank bailouts, e.g.). Citizens may take indiscriminate aim all government programs, as their distrust creates a cycle of inequality and reduced support for welfare programs.
There is, accordingly, a perverse feature to income inequality and negative views of welfare policies. As Figure 3 shows, trust in government rises when the economy is doing well. When real disposable income increases, citizens are more likely to trust government to do what is right. If Hetherington is correct, this means that support for welfare policies increases when they are least needed. During recessions, when income transfers such as unemployment insurance and other welfare programs are stretched, there lacks the political will to extend benefits and create government support for the hardest hit.
Figure 3. Trust in government (1964-2008) and year-over-year third quarter percent change in logged real disposable income per capita (chained 2005 dollars). The linear fit, with 95% confidence bands in gray, is statistically significant (p < 0.05). Sources: U.S. Department of Commerce, Bureau of Economic Analysis and American National Election Survey.
Income inequality highlights a central problem in economic policymaking. While government programs, such as capital gains tax cuts, have fostered inequality, government programs are likewise necessary to combat it. Recent research and analysis presented in this section show the political challenges to policies that seek to improve equality, even if there was consensus on which policy solutions would be the most effective to do so. Income inequality has fostered distrust in government and political polarization, which undermine support for social welfare programs and creates institutional gridlock. Furthermore, during recessions governments bring in less money due to a shrinking tax base, and the lack of trust (which declines further in poor economic times) limits what policy options politicians can pursue. Accordingly, there is less money and less political support for countercyclical policies. This creates a recipe for self-imposed austerity, pain and even greater income inequality. Highlighting these types of self-reinforcing cycles have been a common theme in recent research.
In April 2011, Gallup published a poll indicating that more than half of Americans still thought the United States was in a recession or depression. This finding ran counter to the National Bureau of Economic Research's conclusion that the so-called Great Recession ended in 2009, and it was also contrary to the economic research at the time that showed moderate growth. But it is also not difficult to understand, given that unemployment had largely stayed well above nine percent for nearly two years by that point. The Gallup poll underlines another important point about economic policy: there are economic problems and there are political problems. While the economic recession3 had ended, policymakers were still under political pressure to make economic policy to address the "recession" voters still perceived. There is a disjoint between policy instruments and elite political action, on the one hand, and mass electoral politics on the other hand.
In recent years, research on the electoral implications of economic policy has focused on the ways individuals perceive economic conditions. Gerber and Huber (2009) and Gerber and Huber (2010) find that partisan sentiment conditions economic perception and behavior, contrary to the well-known traditional model that argues vote choice is conditioned on economic perception. Gerber and Huber (2009) argue that individuals have a partisan identity as well as partisan beliefs. Consumption in Democratic-leaning counties goes up after a Democrat is elected to the presidency, regardless of observed economic conditions. This type of effect is not limited to presidential elections. Gerber and Huber (2010) find that a similar pattern emerges after congressional elections and statewide elections.
The Gerber and Huber studies present further difficulties for scholars contending that retrospective economic assessments predict vote choice. If partisanship affects economic assessments and vote choice is heavily correlated with economic assessments, then partisanship is still driving vote choice. Lewis-Beck, Nadeau and Elias (2008) argue that the traditional account is still more accurate. Like Gerber and Huber, Lewis-Beck, Nadeau and Elias use panel data and attempt to "exogenize" party identification. When the causality is taken into fuller account, Lewis-Beck, Nadeau and Elias argue, economic assessments still matter greatly to vote choice. According to the authors, this finding not only saves the extant research on economic voting, but it also means citizens can still hold elected officials accountable for the state of the economy.
The extent of this accountability, however, is questionable. Renno and Gramacho (2010) find that less sophisticated voters in Brazil and Chile are more likely than sophisticated voters to be indiscriminate with their opprobrium for the economy. They tend to want to "blame everybody," regardless of institutional features and constraints. More sophisticated voters seem to pinpoint their anger on the president. The "blame everybody" approach seems to be generalizable to the United States as well, given the 2010 midterm election results. Voters seemed more than willing to vote Democrats out of office despite low approval ratings for Republicans.4
The reverberations of these electoral studies can be felt on policy outcomes. Elections, obviously, have policy import, and how the public interprets economic conditions influences their policy demands. The research in this section suggests that the public's economic assessments are driven by partisan ideology, and economic assessments drive voter's electoral decisions. Moreover, voters seem remarkably willing to blame everyone regardless of institutional constraints. In such an environment, there are multiple avenues for political gamesmanship. The strategies used by political actors to garner electoral favor would alter veto points for legislative passage, which would at least moderate policies and at worst create insurmountable gridlock. So if policy can be created it is likely to be far short of any technocratic ideal.
Policy responses to the Great Recession are ongoing, despite the end of the nominal recession. Congress, the President, and the assorted bureaucracies of economic oversight (such as the Fed, the Treasury Department, and many independent agencies) continue to write rules to regulate financial markets, generate job growth, address the needs of the long-term unemployed, as well as deal with European sovereign debt crises and the national debt, which as of this writing stands at more than $14.5 trillion. Moreover, much of the data on which original decisions were based continue to be revised by statistical agencies within the government.5 Accordingly, there is little by way of solid policy analysis to ground any assessment of institutional actors. In other words, the jury is still out on the policymaking behind and effects of the American Reinvestment and Recovery Act (the "stimulus" bill) in 2009 and several other programs.
The research that has made it to print thus far has described the political dynamics of the legislation that has flowed from the start of the recession in 2007 (see Table 1), particularly the Dodd-Frank Act re-regulating the financial industry in 2010. There have been no fewer than 10 legislative responses to the economic downturn that began in December 2007, not including such legislation as healthcare reform that may have longer-term or indirect effects on economic productivity. Given the high levels of polarization and predictions of gridlock, the amount of legislation may seem surprising, but it was how the legislation managed to pass that warrants greater analysis.
Legislators in Washington began addressing an apparent economic downturn well before the banking crisis in September 2008 but still months after the start of the recession.
Table 1. The Great Recession and Its Legislation. Recession dating follows the National Bureau of Economic Research.
A Democratic Congress and a Republican President managed to pass a tax-cut stimulus in spring 2008. When the full extent of the banking crisis became apparent, the same divided government passed the provisions of the Troubled Asset Relief Program that gave the Department of the Treasury and Federal Reserve Board of Governors unparalleled power to rescue failing investment banks and mitigate the systemic risk all banks faced. But this legislation came only after Treasury Secretary Hank Paulson and Fed Chair Ben Bernanke made congressional leaders a take-it-or-leave-it offer and invoked the specter of another Great Depression during a tense weekend meeting (Sorkin, 2009, 442-43). After the inauguration of Barack Obama, a period of unified government saw the passage of the remaining seven pieces of legislation meant to ease the suffering of homeowners, bring stability to the banking industry, and prevent future crises from occurring.
But unified government did not mean easy or swift passage of that legislation, particularly with regard to reregulation. Carpenter (2010) and McCarty et al. (2010) both document the passage and prospects for Dodd-Frank. For Carpenter (2010) the Dodd-Frank legislation represented threats to established bureaucratic regimes, and strategies of "partisan intransigence" through exploited veto points. These summed to "institutional strangulation" that undermines consumer reform especially. McCarty et al. (2010) compare the 2010 reforms to the creation of the Resolution Trust Corporation in the wake of the Savings and Loan scandal of the late 1980s. In the intervening time, McCarty et al. argue, a newly-developed "regulatory-financial" complex creates a cycle of lax enforcement by regulators, little oversight by congressional committees, increasing complexity in the financial world, and growing income inequality and polarization. These forces team up to inhibit a secure marketplace, in the first instance, and timely responses in the event of a crisis.
This literature squarely blames partisan polarization for watered-down policies created to hurdle previously non-existant veto points and compromises that led to greater delegation to the bureaucracy, which then led to turf battles. This literature also blames an orthodox "free market" ideology that looks at any form of government regulation or taxation as inherently wrong. These issues, and ideology in particular, reduced the capacity of the standing bureaucracies to foresee or forestall the slow-motion collapse of the housing market and the systemic risk in the banking industry, since Bush Administration appointees to the relevant regulatory agencies did not prioritize oversight of the subprime mortgage industry (Khademian, 2009). As Khademian (2009) notes, the irony of the Bush Administration is that while it putatively wanted to reduce bureaucracy's reach into the private sector, its legacy was a massive delegation of control to the Fed and Treasury to rescue the banking industry and the reorganization of financial regulators.
Taylor (2009) and Congleton (2009) argue, by contrast, that the government's reaction to the financial crisis extended the impact of the banking problems. Congleton (2009) argues that talk of a "crisis" hindered the ability of policymakers to respond with appropriate deliberation and analysis. Taylor (2009) argues that government programs subsidizing homeownership helped create the mortgage bubble in the first place along with the Fed's loose monetary policy – without the housing boom, there would be no housing bust. Second, relevant regulators misdiagnosed the problems as the market imploded. And third, the government did not establish consistent criteria in rescuing banks — saving Merrill Lynch, for example, while letting Lehman Brothers fail. All of these actions created uncertainty in the markets and made the situation worse.
Finally, the crisis and the response to it reignited a debate in economics over Keynesian fiscal stimuli. In a paper written before the stimulus measures of 2008-2009 were passed, Magud (2008) finds that government spending in the event of a negative economic shock is stimulative conditional on the state of the government prior to the shock. The more fiscally fragile the government is before the shock, the more contractionary (rather than expansionary) fiscal stimulus is likely to be. Magud (2008) argues that information failures create asymmetric business cycles that may result in recession when complete information would avert one. This piece creates a dilemma for the purely ideological on both ends of the spectrum in the American case. Would America's high level of indebtedness make stimulative spending contractionary? Or would America's large economy offset risk of default and make stimulative spending expansionary? There are downside risks associated with each position that partisans are unlikely to address.
In their analysis of more than eight centuries of financial crises, Reinhart and Rogoff (2009, 289) conclude that "banking crises tend to be protracted affairs." This seems to be case regardless of the policy responses employed by the relevant governments. The research presented in this essay would seem to collaborate Reinhart and Rogoff's conclusion. Recently, rising income inequality, declining trust in government, and greater partisanship and ideological polarization have conspired to create a "regulatory-financial complex" that strangulate policymaking institutions and create gridlock when veto points — constitutionally imposed or otherwise — cannot be surmounted. These forces lead voters to judge the economy with their political ideology firmly in mind, but also retrospectively assess the policymakers. And the circle becomes self-reinforcing as the public's views turn more conservative and less supportive of government intervention to solve economic problems.
There remain questions for future researchers. How the macropolitical climate affects policy outcomes, and how advocates and interest groups mobilize is ripe territory for new research. Additionally, the precise dynamics of institutional relationships need to be further explored. How Congress, for example, interprets Fed action, and how the President and Executive Branch try to get Congress on board with policy solutions in a timely way is not well developed in an integrative manner. And of the many impediments to successful policymaking listed above, we know little about how these institutions focus attention on economic problems and use information to craft effective policy solutions. Krause and Corder (2007) find, for example, that macroeconomic forecasts on which policymakers base their decisions vary depending on the political responsiveness and reputational impacts of the forecaster.Of the many theories of the policy process available, I suggest that the Great Recession lends itself best as a critical case for information processing theories (see Workman, Jones and Jochim, 2009; Baumgartner et al., 2009). When the BEA's GDP estimates are more than six percentage points off, as it was in 2008 for example, how is policy affected? The data on which policymakers act is crucial to understanding the policy solutions they choose. The BEA's inability to provide accurate data is akin to policy information failure during the Iraq War, an example used in Workman, Jones and Jochim (2009). Moreover, if the information is wrong and the policy inadequate, how can policymakers adapt given institutional constraints? Getting all the institutions of governance "on board" for economic policy action is a particularly tall order, even when recessions or other economic problems demand it. Jones and Strahan (1985) argue that during crises the president and the executive branch have an incentive to coordinate concerted action while the decentralized Congress has incentives to splinter responsibility across committees and subcommittees and "spread the wealth" so to speak. Yet because the responsibility for economic policy is diffuse, understanding how information flows between relevant institutions is important in understanding the policy process.
But these information flows are not devoid of politics: institutional dominance and reputation are at stake. The separate branches of government and the highly-specialized bureaucracy are interested in maintaining their independent spheres of influence (Hall and Jones, 1997; Khademian, 1992). Secondly, there is partisan politics, especially at times of divided government (Alesina and Rosenthal, 1995). But independent agencies may not be beyond partisan politics as well, as the Federal Reserve shows a greater propensity to help Republican presidents ceteris paribus (Galbraith, Giovannoni and Russo, 2007).
What is particularly interesting about information processing is how the information signals are conditioned first by what the institution prioritizes and then, secondly, by politics. May, Workman and Jones (2008) find that how organizations pay attention to relevant issues conditions to what they pay attention. Given differing institutional political incentives (the president generally receives more blame for the state of the economy than Congress, (see, for example, Lewis-Beck and Stegmaier, 2007; Alesina and Rosenthal, 1995)), one would expect the separate institutions not only to behave differently when processing economic problems as Jones and Strahan (1985) would anticipate, but pay attention to different issues as well.
This disjointed processing raises questions about how "policy windows" open, what causes them to open, and when they open to address economic problems. For example, one is left to wonder if an opportunity for policy change cannot appear without simultaneous processing of similar concerns across institutions. Extant research also leaves unanswered whether specialized bureaucracies must be on board with administration plans for Congress to act, or whether opinion leaders in Congress in concert with the administration can overcome bureaucratic objections.
Studying economic policymaking from a position of information processing could be divided into three analytical units. First, there are political incentives for political actors. The president, for example, takes more electoral blame for the economy so may be more responsive to economic information. Moreover, the informational infrastructure is located in the Executive Branch. Second, there is reputation for bureaucratic actors, which could be subdivided into credibility and expertise. Reputation is a function of whether agency can speak the truth, especially to the powerful — an agency's credibility — as suggested by Khademian (2009). And reputation is also a function of expertise — whether the agency has a track record of success. Third, an analyst must recognize if the various institutions under study can credibly commit to a given course of action. For example, under what circumstances could President Obama have credibly threatened to veto any debt limit extension during the debt ceiling "crisis" in 2011? There would seem to be too much political pressure to believe a veto was possible, giving Congress the upper hand in negotiations.
Recent events have given researchers cause to reassess their models in light of new information and search for novel ways to examine policymaking anew. Economic inequality, Keynesian stimulus, economic voting, and political polarization have all played a role in the latest set of economic problems and responses to them. In this paper, I have suggested that trust in government and information processing comprised of political incentives, reputation and credible commitment be given greater examination in the future.
Barry Pump. Graduate fellow, Center for American Politics and Public Policy, Department of Political Science, University of Washington, firstname.lastname@example.org.
The author would like to thank Peter J. May, the editors, and the reviewers for their constructive suggestions.