Policy Rhetoric Can Have Economic Consequences: Presidential Rhetorical Liberalism and Economic Policy Uncertainty

Journal of Global Economics, Management and Business Research 4(2): 65-80, 2015

International Knowledge Press www.ikpress.org

POLICY RHETORIC CAN HAVE ECONOMIC CONSEQUENCES: PRESIDENTIAL RHETORICAL LIBERALISM AND ECONOMIC POLICY UNCERTAINTY CHRISTOPHER OLDS1* 1

University of Central Florida, 4297 Andromeda Loop N. Howard Phillips Hall, Rm. 302 Orlando, FL 32816, United States of America. AUTHOR’S CONTRIBUTION The sole author designed, analyzed and interpreted and prepared the manuscript.

Received: 14th May 2015 Accepted: 16th June 2015 Original Research Article Published: 2nd July 2015 __________________________________________________________________________________ ABSTRACT The response of economic variables to changes in the policy ideology expressed through presidential rhetoric is an area in need of extensive empirical exploration. An increase in liberal policy rhetoric can serve as an indicator of an executive branch that will, in general, attempt to increase government involvement in policy areas that have a significant bearing on the economy. Such policy rhetoric is a cue that should lower the extent of uncertainty about what the characteristics of economic policy will consist of. The recent development of a historical indicator of economic policy uncertainty in the United States allows researchers to assess whether the style employed in presidential rhetoric has any bearing on the extent of economic policy uncertainty that is present in the socio-political environment. The vector autoregression time series analysis performed in this project suggests presidential ideological tone Granger-causes economic policy uncertainty. The moving average representation analysis indicates an increase in presidential rhetorical liberalism results in a decrease in the economic policy uncertainty index for a period of time. Chief executives in the United States could be strategically expressing language supporting the expansion of federal government involvement in domestic affairs to mitigate the level of uncertainty about economic policy that is present in the socio-political environment. Keywords: Economic policy uncertainty; presidential rhetoric; macroeconomics and the presidency; policy liberalism; time series analysis; United States political economy. JEL Classification: E02, E60, C22.

1. INTRODUCTION The recent creation by Baker, Bloom, and Davis [1] of a measure of economic policy uncertainty has provided an opportunity to assess whether economic policy uncertainty is shaped in any way by the behavior of prominent political institutions. Whether

the ideological tone expressed by the president has any bearing on the level of economic policy uncertainty in the socio-political environment is an area of research that has yet to be explored in a significant way. The purpose of this project is to evaluate whether change in the ideological tone expressed by the President of the United States

_____________________________________________________________________________________________________ *Corresponding author: Email: [email protected];

Olds; JGEMBR, 4(2): 65-80, 2015

produces change in the level of economic policy uncertainty in the United States.

lower the level of economic policy uncertainty, which should lower the level of risk aversion that inhibits economic activity. Through empirical analyses of time series information about economic policy uncertainty, presidential ideological tone in rhetorical remarks to the public, and a collection of indicators that can possibly have an effect on both of these variables, the project offers an initial indication that the president’s ideological remarks regarding domestic policy have consequences on the level of economic policy uncertainty present in the country.

Such a study has value in that prior research on presidential ideological tone, in particular, the work of Wood [2], provides a very bleak indication that the policy preferences expressed by the president result in any consistent desired effect on the socio-political environment. The contribution of this study is to provide empirical evidence that the policy preferences expressed by the president can curtail the presence of economic policy uncertainty. It is not that the ideological tone expressed by the president does not matter at all; it is just that existing research has not made enough of an effort to look at the various other potential factors that presidential ideological rhetoric can influence. It is not necessarily the case that presidential policy rhetoric has no significant influence in altering variables in the socio-political environment. This project provides empirical evidence that presidential policy rhetoric can shape perceptions about the economy.

1.1 Literature Review 1.1.1 A brief background on economic policy uncertainty Economic policy uncertainty is a concept that represents an absence of clarity regarding the current and prospective qualities about monetary, spending, tax, and regulatory policies that all have implications on the economy. A rise in economic policy uncertainty increases risk-averse behaviors on the part of both firms and individuals that can eventually limit economic growth [1]. Such a proposal is derived from research on general economic uncertainty, as well as research with a focus more specifically on economic policy uncertainty.

Evaluating whether change in the ideological views expressed by the president in public has any role in changing the level of economic policy uncertainty that exists is important. The reason is that the empirical evidence will clarify whether the strategic usage of ideological policy cues by the president can direct change in the socio-political environment. Work from scholars like Edwards [3] and Canes-Wrone [4] create the perception that the most effective presidential administrations take advantage of opportunities in the socio-political environment. This has instilled the view amongst some that the most successful form of presidential leadership will be responsive political leadership; meaning that the president reacts to what is happening in the socio-political environment. While it is certainly the case that responsive political leadership from the president can be a very influential strategy in some circumstances or contexts, that does not mean the door should be closed on the possibility that the president can direct or guide the sociopolitical environment. Through the strategic usage of rhetoric that offers the cue that the president encourages economic activism by the federal government in most major domestic policy areas, the extent of uncertainty about the characteristics of economic policy in the country should decrease.

Research on general economic uncertainty tells us that periods of high general economic uncertainty are periods where firms will limit initiating investment projects [5]. For citizens, general economic uncertainty can lead to a reduction in consumption within individual households [6]. Individual citizens, at times of high general economic uncertainty, will be less inclined to purchase goods that reflect a level of comfort in consuming within the current marketplace (e.g. the purchase of cars, new housing, or appliances). Projects more specifically focused on economic policy uncertainty illustrate that periods of elevated uncertainty about government policy result in periods where stock prices decrease [7]. When an election approaches, and there is trepidation about what future economic policies will look like following the election, U.S. firms will lower foreign direct investment in foreign affiliates [8].

Since economic policy uncertainty increases risk aversion and lowers activities that stimulate economic growth [1], the president will increase the usage of language that suggests a general preference for increasing government involvement in domestic affairs. This rhetoric serves as a signal that the president has a policy agenda that will stimulate demand for goods and services. Such a signal will

The analyses of [1] show there are ramifications to economic policy uncertainty. When economic policy uncertainty increases, there is reduced investment, hiring, and industrial production in the business sector. Economic policy uncertainty appears to encourage risk-averse behaviors by actors in the economy. Whether or not the president can limit 66

Olds; JGEMBR, 4(2): 65-80, 2015

economic policy uncertainty through ideological rhetoric, such that risk-averse behaviors are less likely, is something that has yet to be explored empirically. If the president can direct change in economic policy uncertainty through ideological rhetoric, it suggests the president can play a potential role in the economy. 1.1.2 A brief background ideological policy rhetoric

on

will employ liberal policy rhetoric in major policy areas to give the impression the government will not allow an economic decline to occur. Many might perceive that the president goes to great lengths to minimize the appearance of acting like a partisan, given their perceived status as being the representative of the general will of the entire country [16]. Minimizing the appearance of partisanship from the president stems in part from the philosophy of some of the early prominent leaders in the development of the United States. George Washington, for instance, would offer views indicating that presidents should represent the entire country, instead of just working to advance the policies they personally prefer, which can help to generate partisan divisions within the country [2]. This means many of the Founding Fathers in the United States believed in representing the wishes of all members of the community, because that increases the likelihood the general will of the country is going to be represented.

presidential

Presidents are perceived as being a representative of all members of the country. Since the president and vice-president are the only officials elected by the whole country, the president can stake a claim as having singular authority as the nation’s primary leader [9]. When the president speaks, they speak as symbolic leader of the country, such that the public remarks of the president can be interpreted as being representative of the entire nation as whole [10]. The president is seen as the spokesperson of the country, since there is no other politician that has the potential to have such a consistent and close relationship with the general public [11].

Such a view is tied to Downs’ [17] work that suggests politicians like the president will move ideologically toward the median voter in order to increase the likelihood of garnering public support. Presidents should want to appeal to the ideological center. The reason is that many American citizens do not enjoy partisan politics, possess moderate ideological attitudes, and frown upon politicians that are only responsive to the individuals who voted for them [2]. A centrist model of representation proposes that presidents will express views that match the interests and priorities of most of the general nation as a whole, such that the president believes their primary constituency is the entire country.

The perceived authority of being the symbolic leader of the country is of particular importance as it relates to the economy. The state of the national economy is the main issue that is often used to gauge a president’s policy performance and political success [12]. When national economic performance changes, presidential approval ratings shift, with the direction of the shift contingent upon the direction of change in the economic climate within the United States [13]. The electoral success of both the president, as well as the political party the president belongs to, will be tied to the status of the American economy [14]. As a result, the president should be very mindful of the possible implications of their public remarks on the state of the national economy. Wood [15] has already proposed the president will employ optimism when speaking about the national economy to generate support for themselves and the political party they belong to.

Regardless of all these considerations, presidents might adopt a liberal policy tone if they believe that touting government involvement in domestic affairs will make economic actors less wary of participating in the national economy. The usage of liberal policy rhetoric could be an attempt at persuading economic actors that the government is not going to allow for an economic downturn to occur. While the president might have originally been expected to be a centrist that is representative of the nation overall, the president might have to use ideological rhetoric strategically to reduce any growth in economic policy uncertainty.

Speaking optimistically about the economy is not the only stylistic tool the president can use though when making public remarks about the economy. Using ideological rhetoric about major domestic policy areas can serve as a cue of the intended goals and plans of an administration. A president that speaks about more government involvement in domestic affairs is sending a policy signal that the government is going to strive for greater expenditures in most policy areas to make economic actors less risk-averse. The president is encouraging federal government activism in the national economy when repeatedly and consistently offering liberal policy rhetoric. Presidents

In an empirical analysis of presidential rhetoric and public opinion, Wood [2] finds that the president will usually express their own ideological policy views, regardless of what public sentiment is. He finds no systematic relationship between the public’s preferred 67

Olds; JGEMBR, 4(2): 65-80, 2015

level of liberal domestic policy, and the level of the president’s support for liberal domestic policy expressed in public remarks. This finding supports the perspective of Fenno [18] that one of the primary goals of politicians is to work to advance policies they personally prefer. Wood’s [2] analysis indicates that presidents will tend to offer partisan issue positions in an attempt to get the mass public to adopt the president’s preferred policy position. Interestingly, as the public’s preferred level of liberal domestic policy increases, presidential partisan rhetoric expresses a lower level of support for liberal policy views. This means the president is not responding to the public’s policy preferences.

considered all for naught. The president, in adopting an ideological tone, can serve another purpose. Rhetorical policy liberalism can quell sentiment of economic policy uncertainty.

1.2 Theory and Hypothesis Liberal ideological issue stances in United States politics usually take the form of positions in favor of a larger federal government, and a more expansive role from the federal government in domestic affairs [2]. These issue stances support more government involvement in areas that will ensure citizens can enjoy reasonable education, health, opportunity, and economic status. The reason for this could be the belief that freedom is only possible when all citizens are not plagued with disadvantages in these areas. For this reason, liberal ideological stances in American politics can be thought of as very similar to social liberalism. The expectation is that government will play a significant role in advocating for proper education, health, opportunity, and economic status for all citizens. When people are uncertain about the future direction of the economy, a consistent and pervasive usage of liberal policy remarks should instill the sense that the government will have significant involvement in the national economy.

Presidents, instead of acting as centrists, serve as natural partisans that have a preference for advancing partisan policy instruments [2]. It is possible that the expression of partisan views is done out of sheer necessity. Strategically, if adopting a tone about policy will make economic actors more willing to invest, hire, and produce goods, it is logical for the president to use a partisan tone that is in favor of increased government involvement in domestic affairs. Presidential rhetoric then adopts a partisan ideological tone, regardless of significant change in the ideological preferences of the public, or even in the face of change in presidential approval ratings. The ideology of the public though does not appear to shift directly as a result of changes in the president’s partisan ideological tone, which could lead to the conclusion that the president struggles to get the public to adopt the president’s perspective through rhetoric when policy views between the president and the public are divergent [19]. Despite the public not adopting the president’s policy views, this does not mean the expression of partisanship in public remarks is altogether useless.

When economic policy uncertainty is high, there is not a clear sense as to what the current and prospective makeup will be of policies that have important implications on the economy. Growth in economic policy uncertainty should make decisionmakers risk averse. The presence of risks make decision-makers unsure about the options available to them, as well as the potential outcomes that follow from the options that are available to them [20]. If economic policy uncertainty is high, economic actors will be unsure whether expending resources will derive a valuable return on the investment. As a result, economic actors are more tentative in engaging in economic activity. This presents potentially negative consequences for the president, given that a lagging economy might curb public approval ratings of presidential job performance [13]. The performance of the president is often framed in terms of the general economic health of the country [12]. At times when economic policy uncertainty is high, presidents will consider the usage of rhetorical techniques that might mitigate the level of uncertainty that exists.

There are multiple ways in which the president can demonstrate rhetorical leadership through persuasion. While the ideological tone of presidential rhetoric and the ideological sentiment of public opinion do not appear to consistently influence each other, there is one other area where presidential rhetorical policy liberalism can serve a strategic purpose: lowering the level of economic policy uncertainty that exists. With the work of Baker, Bloom, and Davis [1], it is now possible to assess the role of ideological rhetoric on economic policy uncertainty. Instead of suggesting that the usage of ideological policy rhetoric is a futile effort, as the public does not appear to adopt the ideological policy views of the president, ideological presidential remarks should not necessarily be

If there is uncertainty about the economic policy in a country, major political actors of that country should be less willing to make public remarks that will exacerbate the level of uncertainty that exists regarding the economic policy of the country. In

68

Olds; JGEMBR, 4(2): 65-80, 2015

terms of presidential rhetorical strategy, presidents will express a liberal policy agenda in an attempt to lower economic policy uncertainty. A liberal policy agenda, which in general suggests an increase in government involvement in domestic affairs, indicates that the government will intervene in major policy areas that have a bearing on the national economy. Economic actors as a result will come to the view that the government will be very active in the economy, such that there is greater clarity as to what the framework of economic policy will be in the country.

liberalism, specific types of time series techniques are necessary. Vector autoregression and moving average representation are techniques that can assess whether change in one variable can help to predict change in another variable, while simultaneously allowing for the evaluation of a multidirectional relationship between the variables of interest. Vector autoregression and moving average representation have been employed previously when evaluating whether the president leads or reacts to the sociopolitical environment [2,12].

Pervasive rhetorical liberalism from the president should serve as an indication that the federal government will attempt to stimulate economic demand when private demand for goods and services is lacking. With a more refined sense of what the economic policy agenda of the government will be, the level of economic policy uncertainty should decline. Since the electoral success of a president and their party is based largely on the state of the domestic economy [14], presidents will use ideological remarks to persuade economic actors that a clear and consistent identity exists in regards to the economic policy of the country. As a result of this, economic actors should be less reluctant to engage in the national economy.

2.1 Vector Autoregression In vector autoregression (abbreviated in some of the scholarship as VAR) models, all of the variables within the specified system are treated as endogenous variables. This means each variable in the system is evaluated by past values of itself, and past values of the other variables in the system. The usage of prior values of variables helps to account for the inertial qualities each variable in the system might possess [23]. Each variable in the endogenous system is regressed on past values of itself, in addition to prior values of any other variable in the endogenous system. Vector autoregression procedures by the nature of the analytical technique itself accounts for history by incorporating lags for every variable that happens to be in the endogenous system.

With all of this in mind, the president’s rhetorical strategy serves as a means in which to alter the level of economic policy uncertainty in the country. The president does not want to heighten the level of economic policy uncertainty even more, as it can hinder growth in the economy. Worsening economic indicators will threaten the president’s chances for electoral success, for either themselves, or for their party. Given these considerations, the following research hypothesis is evaluated for this project.

In hypothesis testing, Granger [24] hypothesis tests are performed. Granger causality testing, through the usage of F-tests, evaluates whether the set of prior values of one variable collectively predict the values of an endogenous variable in the VAR system. The Granger tests evaluate the joint significance of each variable in each system equation via the F-tests. The statistical test is on the restriction that all the lags of a specific variable do not significantly enter into the regression of a particular variable in the endogenous system. The Granger causality test then speaks to whether the lags of one specific variable can together influence the current values of another variable in the endogenous system. When a block of coefficients is statistically significant, that suggests that a Granger causal relationship is present, such that prior values of a variable help to predict current values of a variable.

1.2.1 Research hypothesis An increase in presidential rhetorical liberalism predicts a significant decrease in economic policy uncertainty.

2. METHODOLOGY In analyzing variables that are dynamic in nature (fluctuate over time), the performance of time series analytical approaches are appropriate. There are many approaches to evaluating variables that are dynamic in nature [21,22].

In this project, there are two variables in the endogenous system: presidential rhetorical policy liberalism, and the historical economic policy uncertainty index. So vector autoregression assesses whether prior values of presidential rhetorical policy liberalism predict current values of the historical economic policy uncertainty index, and/or vice-versa. This is done without imposing a theoretical restriction regarding which of these two variables is a priori exogenous.

Given the possibility that change in presidential rhetorical policy liberalism can predict change in economic policy uncertainty, and also, the possibility that change in economic policy uncertainty can predict change in presidential rhetorical policy 69

Olds; JGEMBR, 4(2): 65-80, 2015

Concerns about the omission of variables within the endogenous system are lowered in part by knowing that the effects of any potentially omitted variable from the endogenous system are accounted for by prior values of the endogenous variables that happen to be within the system [12]. In addition, exogenous control variables are accounted for in the statistical analysis, as will be discussed shortly.

To have a more consistent interpretation of the magnitude of the change in one variable after another variable experiences a change (shock), every variable in the endogenous system is standardized to have a mean of zero, and a standard deviation of one. This is done to get a clearer idea about the amount of change a variable will exhibit when another variable in the endogenous system experiences a positive change. When using moving average representation for this project, it is possible to learn how economic policy uncertainty will react following a one standard deviation increase to presidential rhetorical policy liberalism. It is also possible to assess how presidential rhetorical policy liberalism will react following a one standard deviation increase to economic policy uncertainty.

Several Engle-Granger tests are performed at different lag lengths to determine whether the variables in the VAR system are cointegrated [25]. Cointegration means variables share a common trend across time. When cointegration is present, the time series techniques described above should not be used. Performing an Engle-Granger test applies a unit root test to regression residuals involving the variables within the endogenous system. Cointegration does not appear to be a concern based on the battery of EngleGranger tests performed at different lag lengths. These results, presented in Appendix A, mean standard vector autoregression and moving average representation techniques are appropriate to use for the purposes of this project.

2.3 Dataset 2.3.1 Measuring economic policy uncertainty Baker, Bloom, and Davis [1] offer a historical version of their economic policy uncertainty index to download at http://www.policyuncertainty.com/ us_historical.html. The research performed by [1] gives the indication that when the historical economic policy uncertainty index is elevated, these are periods wherein there is elevated sentiment suggestive of economic policy uncertainty. This means we should be comfortable in using it in attempts to assess whether change in presidential rhetorical policy liberalism influences, or is influenced by, change in economic policy uncertainty. Although the historical measure of the economic policy uncertainty index is established as a monthly indicator, for the sake of this project, the measure is converted to a quarterly average. The average score of the economic policy uncertainty index during each quarter of the year is calculated. The reason for doing this calculation and producing a quarterly measure of the economic policy uncertainty index is that the presidential rhetorical policy liberalism measure is a quarterly time series variable.

2.2 Moving Average Representation Vector autoregression techniques give an opportunity to learn about the potential causal direction between variables. VAR techniques though are not helpful for gaining clarity about the polarity (positive or negative direction) and magnitude of the relationship between variables. Given the measurement of multiple lags in the vector autoregression system, coefficient estimates do not provide much information due the presence of multicollinearity. A means in which to discover the polarity and magnitude of the relationship between dynamic variables is through moving average representation (abbreviated in some of the scholarship as MAR). When performing a moving average representation analysis, a simulated shock is given to a variable in the system, and then the trajectory of the response to the shock is observed over a period of time [2]. With the shock to a single variable in the system, you can review how other variables react following the change (shock) that is induced on one of the specific endogenous variables in the system. The usage of moving average representation can provide an indication of whether a variable will increase or decrease following a positive shift to another variable that is within the endogenous system studied. An MAR procedure also affords researchers to gain a sense as to the duration of the shift one variable will exhibit following the change induced on another variable.

The historical economic policy uncertainty index is developed through content analysis procedures. Baker, Bloom, and Davis [1] scan prominent newspapers for keywords that relate to the economy and policy uncertainty. They look for articles that incorporate the terms “uncertainty” or “uncertain,” as well as the terms “commerce,” “economy,” “economic,” “industry,” “business,” and “industrial.” For articles that mention these terms, there has to be a reference to one of the following terms to be considered a relevant story: “Federal Reserve,” “White House,” “Congress,” “legislation,” “deficit,” or “regulation.” This keyword-based search for 70

Olds; JGEMBR, 4(2): 65-80, 2015

articles is done to collect stories that all involve the categories of uncertainty, policy, and the economy.

about how large federal government should be, and how much federal government involvement there should be in domestic affairs.

The historical economic policy uncertainty index is constructed by scanning several high circulation news periodicals, with specific papers studied during various time periods, depending upon their availability electronically for content analysis procedures. The newspapers included are the Wall Street Journal, the New York Times, the Los Angeles Times, the Chicago Tribune, the Washington Post, the Boston Globe, the Miami Herald, the Dallas Morning Tribune, USA Today, and the San Francisco Chronicle. To account for the change in volume of news articles with each paper across time, [1] then divide the number of news articles that express economic policy uncertainty by the total number of news articles that have keywords about the economy and/or business. The time series for each paper is then normalized to standard deviation units, and the total sum of each paper’s time series is calculated to represent economic policy uncertainty. 2.3.2 Measuring liberalism

presidential

rhetorical

The research of [2] measures the ideological tone of presidential remarks in these nine issue areas operating under a similar premise, with the premise being that the ideological tone of presidential remarks in these nine issue areas will move together across time. When the president makes more liberal remarks than conservative remarks in these nine issue areas, this suggests the president generally supports a policy agenda where there tends to be more extensive federal government involvement in domestic affairs. To validate this, [2] electronically extracted sentences that used keywords involving any of the nine issue areas through the use of the PERL (Practical Extraction and Report Language) programming language. The extraction of sentences was performed with a file that was a record of the Public Papers of the Presidents, which comprises all of the public remarks made by each presidential administration. After the electronic extraction of sentences that mention any of the keywords in the nine issue areas, human-based coding procedures were performed where three people validated the ideological tone of presidential remarks. The general guideline was that liberal issue stances usually favor more expansive federal government involvement in domestic affairs, while conservative issue stances usually favor less expansive federal government involvement in domestic affairs. There were more specific coding guidelines for each issue area, and a summary of these guidelines is presented in Table 1.

policy

Presidential rhetorical issue liberalism is represented here using a measure devised by Wood [2] to gauge the influence of presidential ideological rhetoric on the ideology of the public. Presidential rhetoric was selected by [2] to measure ideology because public remarks from the president are a reflection of the president’s ideological position on specific issues. The president’s public remarks express an ideological outlook on things like budget proposals, legislation, and court decisions. Given that the president makes remarks on a daily basis in a permanent campaign of persuasion [2], it is possible to generate a fairly timerefined measure of presidential liberalism by performing an analysis of the ideological rhetorical tone of the president.

Once the human-based coding procedures were completed, the measure of presidential rhetorical policy liberalism was calculated by subtracting the number of sentences in a quarter expressing a conservative policy stance from the number of sentences in a quarter expressing a liberal policy stance. A high positive number on this measure during a specific quarter indicates the president frequently expressed liberal policy stances during that quarter. A high negative number on this measure during a specific quarter indicates the president frequently expressed conservative policy stances during that quarter.

Wood [2], being influenced by work from Stimson [22], suggests that it is possible to gauge a president’s ideological tone about domestic issues by focusing on public remarks made in nine issue areas. These nine issue areas are the following: health, crime, education, urban problems, welfare, the environment, military spending, race, and size of government. Stimson [22] developed a measure of mass public attitudes known as public policy mood based on survey information along these same nine issue areas. The measure calculates the interdependence of public opinion across the nine issue areas, as it is believed there is common movement across these nine issue areas observable over time. This common movement is attributed to general fluctuation in public sentiment

The period of presidential remarks evaluated by Wood’s [2] study ends at the first quarter of the year 2005. The current project does not attempt to measure presidential ideological tone beyond this time, due to the high labor and resource costs that would be involved in doing so, as [2] describes the coding procedure of presidential ideological tone as taking two full years. 71

Olds; JGEMBR, 4(2): 65-80, 2015

Table 1. Coding guidelines used for presidential rhetorical policy liberalism measure Issue area Health

Crime

Education

Urban problems

Welfare

The environment Military spending Race

Size of government

Coding guideline Liberal sentences encourage an expanded role from the federal government in offering health care, support reduced costs in health care, do not support the privatization of health care, and support higher government responsibility. Conservative sentences encourage a reduced federal government responsibility in the provision of health care. Liberal sentences support the rights of defendants, express opposition to capital punishment, encourage gun control, and support rehabilitation when trying to prevent future crime. Conservative sentences support capital punishment, express opposition to gun control rules, support more spending to tackle crime, and encourage a tough approach with criminals. Liberal sentences support higher spending from the federal government in terms of education, encourage an increase in federally sponsored education services, federal government intervention to ensure school integration, and support an overall expanded role for the federal government in education. Conservative sentences encourage a reduced level of federal government responsibility in the provision of education. Liberal sentences support federal government involvement in addressing the problems of unemployment, poverty, and ghettos in large, urban cities. Conservative sentences encourage a reduced role of the federal government in addressing these problems, but do support the usage of force to preserve law and order in large, urban cities. Liberal sentences support federal government involvement in helping those in economic need, encourage an increase in welfare spending, aiding those in poverty, and the promotion of economic equality in America. Conservative sentences encourage a reduced role of federal government involvement in terms of welfare spending and addressing those in need, and support individual responsibility in handling economic affairs. Liberal sentences support higher spending on the environment and an increase in efforts to protect the environment. Conservative sentences encourage less environmental regulation and less federal government spending to protect the environment. Liberal sentences support reduced military, national defense, and security expenditures. Conservative sentences encourage an increase in military, national defense, and security expenditures. Liberal sentences support federal government involvement in terms of the promotion of racial integration, voting rights, equal jobs, anti-discriminatory practices, equality in education, and the expansion of opportunities for traditionally underrepresented minorities. Conservative sentences encourage a reduction in the involvement of the federal government in terms of advancing these particular causes. Liberal sentences support policies that will increase the size of federal government, the level of federal government spending, the level of federal government taxation, and the extent of federal government regulation. Conservative sentences encourage a reduction in the size of federal government, the level of federal government spending, the level of federal government taxation, and the extent of federal government regulation. Note: Coding rules employed by Wood [2] in developing the rhetorical liberalism measure

Given the high level of time and amount of humanbased coding needed to extend Wood’s [2] measure, this current project will instead focus on the time period wherein the existing measures of historical economic policy uncertainty, presidential rhetorical policy liberalism, and exogenous control indicator information all overlap. As a result, the time period of focus ranges from the second quarter of 1948, to the first quarter of 2005. This is more than enough time to perform informative time series analyses exploring the question of whether an increase in presidential rhetorical policy liberalism lowers economic policy uncertainty.

2.3.3 Exogenous controls used The two variables in the endogenous system, presidential rhetorical policy liberalism and economic policy uncertainty, are not the only variables accounted for in the project. Multiple exogenous control variables are included, because there are events or indicators that can have a significant effect on either the current level of economic policy uncertainty, or the current level of presidential rhetorical policy liberalism. To account for the effect of partisan control of government, a dichotomous dummy variable indicator 72

Olds; JGEMBR, 4(2): 65-80, 2015

is included (with a score of “1” indicating the presence of something, and a score of “0” indicating the absence of something) measuring whether the Democratic Party controls both the executive branch and legislative branches of federal government (as indicated by the Democratic Party having a clear majority of seats in both chambers of Congress). This is done because Democratic Party control of both the executive and legislative branches could encourage a president to express more liberal policy remarks without fear of significant backlash from the legislature. It is also possible that Democratic Party control of government could serve as an indicator of the type of policy agenda the country will take, which has the potential to play a role in the level of economic policy uncertainty.

real personal consumption expenditures is also accounted for. The information for these two indicators comes from the U.S. Bureau of Economic Analysis. Quarters that are periods of a recession are coded as a “1,” and all of the other quarters are coded as a “0” in the analysis. Recession periods are indicated by information from the National Bureau of Economic Research. As an attempt to account for the change in policy uncertainty or change in ideological tone that result from actual major economic initiatives passed, a variable measuring significant economic initiatives passed through Congress that the president proposed and promoted is included. The reason the variable is included is due to the possibility that economic policy uncertainty levels change as a result of the potential confusion regarding the consequences that the passage of these economic initiatives will have on the makeup of the nation’s economic policy. The president might alter their ideological tone when initiatives that the administration promotes make it through the policymaking process in Congress. The list of major economic initiatives during the time period studied, information collected in [15], is presented in Appendix B. The variable is coded as a “1” in those quarters where a significant economic initiative proposed by the president was actually passed and formally signed into law by the president. The variable is coded as a “0” for those quarters where there was no major economic initiative proposed by the president that happened to be approved.

Since [1,8] indicate that uncertainty about election outcomes can heighten economic policy uncertainty, a variable representing the occurrence of campaigns and elections is incorporated in the project. The variable is coded as a “1” for the quarter immediately before and the quarter during either or both the occurrence of congressional or presidential elections. Presidents might also temper the expression of ideological rhetoric during election periods in attempts to appeal to moderate voters, and/or those who identify as politically independent of either major party. The occurrence of wars or military conflicts can potentially influence either economic policy uncertainty or presidential rhetorical liberalism. As a result, the analysis measures the quarters that mark the duration of formal war or military conflict participation by the United States. A unique measure is included for each separate war or military conflict. The quarters that designate the “periods of war” as classified by the Congressional Research Service are scored as a “1,” and all of the other quarters are coded as a “0” for each formal war or military conflict analyzed. The wars and military conflicts measured in this analysis are the following: Korea, Vietnam, Lebanon, Grenada, the first Iraq War, and the second Iraq War.

Lastly, an indicator of major ideological policy enactments is included in the empirical assessment. This measure is the calculated difference between one measure that tracks the total extent of major liberal policy enactments that occurred in a quarter, and a second measure that tracks the total extent of major conservative policy enactments that occurred in a quarter. For both liberal and conservative indicators, each unique policy enactment is coded as a “1,” with exceptionally important policy enactments given a score of “2.” These guidelines are those used by [21] and Bølstad [26], based on the classification by ideology of congressional enactments compiled by Mayhew [27].

To account for actual economic conditions, several indicators are incorporated in the analysis. To measure inflation, the quarterly average percent change in the Consumer Price Index for all urban consumers is represented. To represent unemployment, the quarterly percent change in the civilian unemployment rate from the prior quarter is measured. Both of these measures are calculated from information provided by the U.S. Bureau of Labor Statistics. To capture Gross Domestic Product, the quarterly percent change in the GDP from the prior quarter is included. The quarterly percent change in

After incorporating all of these exogenous controls, it is possible to perform a thorough examination of how the dynamics of presidential rhetorical liberalism and economic policy uncertainty relate to each other. Visual dynamics of the historical economic policy uncertainty index and presidential rhetorical liberalism for the time period where information is available for both measures is presented in Fig. 1.

73

Olds; JGEMBR, 4(2): 65-80, 2015

Fig. 1. Economic policy uncertainty and presidential rhetorical liberalism dynamics Source: Wood [2] for presidential rhetorical liberalism dataset, and Baker, Bloom, and Davis [1] for historical economic policy uncertainty dataset

Some interesting periods visible within this time frame bear mentioning. Between 1964 and 1967, it appears that economic policy uncertainty levels are low, while presidential rhetorical policy liberalism is high. The number of presidential liberal policy sentences is never close to being below zero during this period. This period stands in contrast to two other periods, from 1981 to 1983, and 2001 to 2005. During these two periods, it appears economic policy uncertainty is high, while presidential rhetorical policy liberalism is low. During these periods, we see the number of liberal policy sentences reach into negative values, which indicates there are more conservative policy sentences being expressed by the president than liberal policy sentences.

period studied, when presidential rhetorical policy liberalism increases, economic policy uncertainty decreases for a discernible period of time.

3.1 Vector Autoregression Analysis Findings The vector autoregression analysis suggests that prior levels in presidential rhetorical policy liberalism predict current values of the economic policy uncertainty index. In other words, presidential rhetorical policy liberalism in public remarks regarding major domestic policy areas Granger-causes economic policy uncertainty (P=.04). Economic policy uncertainty levels are influenced by prior levels of presidential rhetorical liberalism. It should also be noted that economic policy uncertainty is inertial, such that prior values of economic policy uncertainty can predict current values of economic policy uncertainty (P=.00).

While contrasting the dynamics of these two variables visually is of value, we do not gain a sense as to whether change in presidential rhetorical policy liberalism predicts change in economic policy uncertainty, and/or vice-versa. In order to gain a sense of this, time series analyses are needed that contrast the dynamics of the economic policy uncertainty and presidential rhetorical policy liberalism endogenous system, while simultaneously accounting for potential alternative factors outside of the endogenous system that can potentially predict changes in either or both of these variables.

While economic policy uncertainty is responsive to prior values of presidential rhetorical policy liberalism, presidential rhetorical policy liberalism does not appear to be responsive to prior levels of economic policy uncertainty. Economic policy uncertainty, at least during the time frame analyzed here, does not Granger-cause presidential rhetorical policy liberalism (P=.95). There is no indication in the vector autoregression analysis that presidential rhetorical policy liberalism is shaped in any meaningful way by prior change in economic policy uncertainty. It is clear though that presidential rhetorical policy liberalism is inertial, as prior values of rhetorical liberalism Granger-cause current values of rhetorical liberalism (P=.00). The full results are demonstrated in Table 2.

3. RESULTS AND DISCUSSION The findings of the time series analyses confirm the research hypothesis of the project, that change in the level of presidential rhetorical policy liberalism expressed predicts change in the level of economic policy uncertainty. Using information from the time 74

Olds; JGEMBR, 4(2): 65-80, 2015

Table 2. Granger tests for economic policy uncertainty and presidential policy liberalism Independent variable

Dependent variable

Presidential rhetorical policy liberalism ===>

Presidential rhetorical policy liberalism

Economic policy uncertainty index Exogenous controls Democratic party control of government (+, P= .01*) Elections (P= .42) Korea (P= .33) Vietnam (P= .84) Lebanon (P= .44) Grenada (P= .94) First Iraq War (P= .64) Second Iraq War (P= .33) Recession (P= .41) Approved presidential economic initiatives (P= .27) Inflation (P= .48) Gross Domestic Product (P= .99) Unemployment (P= .74) Real personal consumption expenditures (P= .60) Enacted policy liberalism (P= .89) Presidential rhetorical policy liberalism ===> Economic policy uncertainty index ===>

Economic policy uncertainty index

P-value [F-Statistic] .00 [79.52] .95 [0.18]

.04 [2.56] .00 [127.02]

Exogenous controls Democratic party control of government (P= .42) Elections (P= .73) Korea (P= .68) Vietnam (P= .57) Lebanon (P= .46) Grenada (P= .69) First Iraq War (P= .57) Second Iraq War (+, P= .004**) Recession (P= .62) Approved presidential economic initiatives (P= .70) Inflation (P= .14) Gross domestic product (P= .84) Unemployment (P= .52) Real personal consumption expenditures (P= .08) Enacted policy liberalism (P= .09) Note: The arrows indicate Granger-causality from the block of coefficients for the independent variable to the dependent variable based on 0.05 significance levels. The p-values are from F-tests for the null hypothesis of no Granger-causality. The system includes a deterministic constant. The results of the exogenous controls are based on t-test results; with * representing a P < .05, and ** representing a P < .01. A “+” represents a positive significant relationship, and a “-” represents a negative significant relationship. Each of the independent variables in the endogenous system includes four quarterly lags to control for the inertia of the variables. Lag length selected by Bayesian information criterion (BIC). As a result, VAR estimation with lags is performed with information between Quarter Two of 1949 and Quarter One of 2005

In terms of the exogenous controls, it is striking that very few of the indicators have a significant effect on either economic policy uncertainty, or presidential rhetorical policy liberalism. The only exogenous control that significantly predicts the economic policy uncertainty index is the formal “period of war” in the

second Iraq conflict in 2003 (P=.004). A positive relationship is present, which parallels the view of [1] that significant military conflicts have the potential to induce an increase in economic policy uncertainty. The relatively low level of tangible information about the ramifications of the second major conflict in Iraq 75

Olds; JGEMBR, 4(2): 65-80, 2015

on the framework of economic policy in the country produced an increase in economic policy uncertainty. None of the other exogenous controls when assessing economic policy uncertainty demonstrated a significant effect (when electing to establish P < .05 as the threshold) though, including other military conflicts. If one were to use P < .10 as the threshold for significance in exogenous controls, then the effects of real personal consumption expenditures (P = .08), as well as enacted policy liberalism (P = .09), would also be variables that can potentially be considered as those that can help to shape economic policy uncertainty levels.

presidential rhetorical policy liberalism, there is a significant decrease in economic policy uncertainty that is apparent for a period. In order to observe the dynamics that demonstrate this, review the plots presented in Fig. 2. Fig. 2 displays the response of variables in the endogenous system when there is a simulated shock imposed on each individual variable within the system. The number of quarters following a positive shock to a variable is shown along the horizontal axis of each plot. The positive or negative standard deviation shift from the standardized mean after a variable in the system experiences a positive shock is displayed along the vertical axis of each plot. The variable that is given a simulated shock is the same within each column of Fig. 2.

Only one of the exogenous controls analyzed significantly predicts presidential rhetorical policy liberalism. The sole statistically significant exogenous variable in this case is the Democratic Party control of government indicator (P = .01), and a positive relationship is present. It is intuitive that Democratic administrations will feel more comfortable expressing liberal policy rhetoric about major domestic policy areas that have a bearing on the economy when fellow Democrats control both chambers of the legislature. What is striking is that none of the other indicators are significant, including the indicators of major economic conditions. The work of [2] suggests variables describing the economic context, such as unemployment, GDP, and inflation, are significant predictors of presidential rhetorical policy liberalism. The major difference between [2] and this study is that the current study has a direct measure of economic policy uncertainty, the economic policy uncertainty index variable developed by [1]. Future research should continue to explore whether the effects of more traditional economic indicators on presidential rhetorical policy liberalism disappear when also accounting for the concept of economic policy uncertainty.

Within the first column of Fig. 2, the variable that receives the one standard deviation positive shock is presidential rhetorical policy liberalism. In row one of column one, the response of presidential rhetorical policy liberalism to the positive shock in itself is displayed. In terms of the current project, our focus is on the plot presented in row two of column one in Fig. 2. This displays the response over several quarters in the level of economic policy uncertainty following a positive shock to rhetorical policy liberalism from the president. Based on the dynamics presented in the plot, a one standard deviation positive increase in presidential rhetorical policy liberalism produces a .05 standard deviation contemporaneous decrease in economic policy uncertainty. This means at the point of the positive increase in presidential rhetorical policy liberalism, economic policy uncertainty immediately decreases below the standardized mean. This decrease is statistically significant, as demonstrated by the dashed lines representing the 95% confidence interval being bounded away from the standardized mean of zero.

To summarize, the Granger tests of the endogenous system of variables suggest that the level of economic policy uncertainty is responsive to presidential ideological rhetoric, and not the other way around. Still, the analysis so far does not give a clear picture as to the magnitude of, and the polarity of, change in economic policy uncertainty in the face of an increase in presidential rhetorical policy liberalism. In order to get at this aspect, a review is required of the moving average representation findings.

The first quarter after the immediate decrease to economic policy uncertainty, the impulse response of economic policy uncertainty is below the standardized mean, but the 95% confidence interval is not totally bounded away from zero. This means we cannot say in the first quarter there is significant evidence after the shock to presidential rhetorical policy liberalism that economic policy uncertainty decreases. Nonetheless, there is significant evidence that an increase in rhetorical liberalism decreases economic policy uncertainty in quarters two and three.

3.2 Moving Average Representation Analysis Findings

There is a statistically significant decrease in economic policy uncertainty following quarter one and before the start of quarter four following a positive shock to presidential rhetorical policy

When reviewing the moving average representation results, there is again support for the research hypothesis of the project. When there is an increase in 76

Olds; JGEMBR, JGEMBR 4(2): 65-80, 2015

liberalism. The 95% confidence interval is clearly bounded away from zero during this period. In the second quarter after the shock to presidential rhetorical policy liberalism,, economic policy uncertainty is about .08 standard deviations below the standardized mean of zero. This again suggests a decline in economic policy uncertainty occurs following an increase in presidential rhetorical policy liberalism. The decline persists into the third quarter. During this time frame then, economic policy uncertainty is around .05 and .08 standard deviations below the standardized mean. For essentially a major portion of the first year following a positive increase to presidential rhetorical policy liberalism, economic policy uncertainty is reduced.

year in which the positive jolt to presidential rhetorical policy liberalism occurs. Regardless, at the point of the positive in increase to presidential rhetorical policy liberalism, as well as following quarter one, up to the period between quarters three and four following the positive shock, economic policy uncertainty declines. Such a result is congruent with the research hypothesis sis of the project, which predicted a negative relationship between presidential rhetorical policy liberalism and economic policy uncertainty. It is worth noting that in row one and column two of Fig. 2, the response to presidential rhetorical policy liberalism ralism following a positive increase in economic policy uncertainty is displayed. At no point is the 95% confidence interval bounded away from the standardized mean of zero. This means there is no statistically significant evidence that an increase in economic omic policy uncertainty will bring about an increase in presidential rhetorical policy liberalism. The chief executive of the United States is not influenced by economic policy uncertainty levels in any meaningful or observable way. Instead, the president can appear to use rhetoric to direct the level of economic policy uncertainty in the socio socio-political environment.

For the entire second year following the shock to presidential rhetorical policy liberalism, the response of economic policy uncertainty is a bit less clear, since the dynamics are not bounded away from the standardized mean for any real period of time to distinguish a clear direction and magnitude of change. From the fourth quarter onwards, the 95% confidence interval is never clearly bounded away from zero. This means any effect an increase rease in presidential rhetorical policy liberalism has on economic policy uncertainty will most likely be confined to the same

Fig. 2. MAR impulse responses to uncertainty and presidential liberalism system Note: Dashed lines represent 95% confidence intervals. The numbers along the horizontal axis of each plot represent the specific quarter following the positive shock. The numbers along the vertical axis of each plot represent the size of the standard deviation tion shift from the standardized mean of zero. “Pres. Liberalism” represents presidential rhetorical policy liberalism, while “Econ. Uncertaint.” represents the economic policy uncertainty index. For a more intuitive interpretation of change across variables es in the system, each variable in the system is standardized. Standardization means each variable is rescaled to have a mean of zero and a standard deviation of one. This helps establish a uniform interpretation when comparing change across the presidential al rhetorical policy liberalism and the economic policy uncertaint uncertainty index variables

77

Olds; JGEMBR, 4(2): 65-80, 2015

In heightening the usage of policy rhetoric with a liberal tone, the president is strategically sending a clear message that there should be increased government involvement in major areas of domestic affairs that can have major ramifications on the economy. Encouraging government activism clarifies what the framework of economic policy will likely be, such that there is less uncertainty about what economic policy in the country will likely consist of. The reason is that the chief executive is making clear that the intention is to stimulate economic demand, such that individuals and firms should have less trepidation about engaging in the economic marketplace.

of these other respective countries. Fortunately, Baker, Bloom, and Davis [1] have already devised economic policy uncertainty index measures for other countries, such as Russia, Canada, and India. Although it will take significant resources and effort, scholars should code the ideological rhetoric of the leaders of these countries, and then contrast this information with the dynamics of the economic policy uncertainty index of those respective countries. Research from Laver and Garry [28] has already developed a policy position dictionary that is appropriate for use with the policy positions of elites in the United Kingdom. For this reason, a logical starting place in attempting to perform a replication analysis in another context is the United Kingdom. Developing unique policy position dictionaries for other countries will be a taxing, but ultimately fruitful research effort. Work like this will provide a clearer sense of the consequences of ideological tone from political leaders on economic policy uncertainty levels. For now, at least in the United States context, there is initial evidence that the ideological rhetorical strategy employed by the president has an influence on economic policy uncertainty.

4. CONCLUSION The purpose of this project was to make some headway in learning about the consequences of presidential ideological tone in public remarks on economic policy uncertainty. The focus of the project was to contrast the dynamics of the ideological tone of domestic policy remarks expressed by the President of the United States with the dynamics of the historical economic policy uncertainty index in the United States. During the time period studied, the empirical analyses performed indicate an increase in presidential rhetorical policy liberalism results in a decline in the level of economic policy uncertainty that lasts for several months.

The initial evidence of this project that presidential rhetoric influences historical economic policy uncertainty indicates that the president can engage in direct leadership of an important aspect of the economy. It is important to contrast this finding with other work [29] examining contemporary economic policy uncertainty that shows the president is responsive to economic policy uncertainty levels, opting to simplify the linguistic and substantive complexity of public remarks in the face of rising economic policy uncertainty. An important research agenda well into the future will involve assessing under what set of circumstances presidential rhetoric can direct change to economic policy uncertainty, and what set of circumstances presidential rhetoric is responsive to change in economic policy uncertainty.

More presidential rhetorical policy liberalism is a signal of a policy agenda to expand federal government involvement in domestic affairs. A proposed expansion of federal government involvement in domestic affairs can help to clarify questions as to what will happen regarding economic policy in the country. Rhetorical policy liberalism can potentially lower the sense that the economic climate is murky. The reason is that heightened rhetorical policy liberalism serves as a signal that the government is going to be actively involved in most major domestic policy areas, such that it is clear the government is going to make an effort to stimulate economic demand. As a result, when economic policy uncertainty is high, we should expect the president to heighten the usage of rhetorical policy liberalism, such that the extent of economic policy uncertainty does not increase. The result is intended to be that economic actors will be more willing to engage in the economic marketplace, and abstain from risk-averse behaviors.

COMPETING INTERESTS Author has declared that no competing interests exist.

REFERENCES 1.

2.

In order to validate these findings, researchers should check the economic policy uncertainty index of other countries, and contrast this information to the level of rhetorical policy liberalism expressed by the leaders

3.

78

Baker SR, Bloom N, Davis SJ. Measuring economic policy uncertainty. Monograph series, policyuncertainy.com. 2013;1-55. Wood BD. The myth of presidential representation, Cambridge University Press, New York, USA; 2009. Edwards III GC. The strategic president: Persuasion and opportunity in presidential

Olds; JGEMBR, 4(2): 65-80, 2015

4.

5.

6.

7.

8.

9. 10.

11. 12.

13.

14.

15.

16.

leadership, Princeton University Press, New Jersey, USA; 2009. Canes-Wrone B. The president’s legislative influence from public appeals. American Journal of Political Science. 2001;45:(2)313329. Bernanke BS. Irreversibility, uncertainty and cyclical investment. Quarterly Journal of Economics. 1983;98(1):85-106. Zhang Y, Wan GH. Liquidity constraint, uncertainty and household consumption in China. Applied Economics. 2004;36:(19)22212229. Pastor L, Veronesi P. Uncertainty about government policy and stock prices. The Journal of Finance. 2012;67(4):1219-1264. Julio B, Yook Y. Policy uncertainty, irreversibility and cross-border flows of capital. Social Science Research Network. 2014;1-48. Available: http://ssrn.com/abstract=2024612 (April 2014) Patterson TE. The American democracy, McGraw-Hill, New York, USA; 1990. O’Connor K, Sabato LJ. American government: Roots and reform, Macmillan, New York, USA; 1993. Cohen J, Nice D. The presidency, McGrawHill, New York, USA; 2003. Eshbaugh-Soha M, Peake JS. Breaking through the noise: Presidential leadership, public opinion and the news media, Stanford University Press, California, USA; 2011. MacKuen MB, Erikson RS, Stimson JA. Peasants or bankers? The American electorate and the U.S. economy. American Political Science Review. 1992;86(3):597-611. Erikson RS. Economic conditions and the presidential vote. American Political Science Review. 1989;83(2):567-573. Wood BD. The politics of economic leadership, Princeton University Press, New Jersey, USA; 2007. Cronin TE, Genovese M. Paradoxes of the American presidency, Oxford University Press, New York, USA; 2004.

17. 18. 19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

79

Downs A. An economic theory of democracy, Harper and Row, New York, USA; 1957. Fenno RF. Congressmen in committees, Little, Brown, Massachusetts, USA; 1973. Hill KQ. The policy agendas of the president and the mass public: A research validation and extension. American Journal of Political Science. 1998;42(4):1328-1334. Fischhoff B, Kadvany J. Risk: A very short introduction, Oxford University Press, USA; 2011. Erikson RS, MacKuen MB, Stimson J. The macro polity, Cambridge University Press, Cambridge; 2002. Stimson JA. Public opinion in America: Moods, cycles, and swings, Westview Press, Colorado, USA; 1999. Sims CA. Macroeconomics and reality. Econometrica. 1980;48(1):1-48. Granger CWJ. Investigating causal relations by econometric models and cross-spectral models’, Econometrica. 1969;37(3):424-438. Engle RF III, Granger CWJ. Co-integration and error correction: Representation, estimation and testing. Econometrica. 1987;55(2):251– 276. Bølstad J. Thermostatic voting: Presidential elections in light of new policy data. PS: Political Science & Politics. 2012;45:(1)44-50. Mayhew D. Datasets and materials: Divided we govern; 2013. Available at: [Accessed 5 September 2014]. Laver M, Garry J. Estimating policy positions from political texts. American Journal of Political Science. 2000;44(3):619-634. Olds C. Analyzing the economic policy uncertainty index and presidential rhetorical simplicity in the United States. The Empirical Econometrics and Quantitative Economics Letters. 2014;3(2):15-32.

Olds; JGEMBR, 4(2): 65-80, 2015

APPENDIX Appendix A. Engle-Granger cointegration test results using different lags Number of lags Test statistic Critical value .01 Critical value .05 Critical value .10 Observations

3 -2.36141 -3.94751 -3.36440 -3.06440 225

4 -2.64038 -3.94773 -3.36452 -3.06448 224

5 -3.00233 -3.94794 -3.36464 -3.06456 223

6 -2.69728 -3.94816 -3.36476 -3.06465 222

Note: Null hypothesis is no cointegration (residual has a unit root). Based on results above, at no point can null hypothesis be rejected. No cointegration exists, so standard vector autoregression and moving average representation techniques are appropriate

Appendix B. Major presidential economic initiatives passed through congress Economic initiative name Manpower Development and Training Act of 1962 Revenue Act of 1962 Revenue Act of 1964 Economic Opportunity Act of 1964 Excise Tax Reduction Act of 1965 Revenue and Expenditure Control Act of 1968 Employment Security Amendments of 1970 Emergency Employment Act of 1971 Revenue Act of 1971 Comprehensive Employment and Training Act of 1973 Tax Reduction Act of 1975 Unemployment Compensation Amendments of 1976 Tax Reduction and Simplification Act of 1977 National Energy Act of 1978 Crude Oil Windfall Profits Tax Act of 1980 Economic Recovery Tax Act of 1981 Tax Equity and Fiscal Responsibility Act of 1982 Job Training Partnership Act of 1982 Deficit Reduction Act of 1984 Balanced Budget and Emergency Deficit Control Act of 1985 Balanced Budget and Emergency Deficit Control Reaffirmation Act of 1987 Omnibus Budget Reconciliation Act of 1990 Omnibus Budget Reconciliation Act of 1993 Balanced Budget Act of 1997 Economic Growth and Tax Reform Reconciliation Act of 2001 Job Creation and Worker Assistance Act of 2002 Jobs and Growth Tax Relief Reconciliation Act of 2003 Working Families Tax Relief Act of 2004

Time period approved Quarter One, 1962 Quarter Four, 1962 Quarter One, 1964 Quarter Three, 1964 Quarter Two, 1965 Quarter Two, 1968 Quarter Three, 1970 Quarter Three, 1971 Quarter Four, 1971 Quarter Four, 1973 Quarter One, 1975 Quarter Four, 1976 Quarter Two, 1977 Quarter Four, 1978 Quarter Two, 1980 Quarter Three, 1981 Quarter Three, 1982 Quarter Four, 1982 Quarter Three, 1984 Quarter Four, 1985 Quarter Three, 1987 Quarter Four, 1990 Quarter Three, 1993 Quarter Three, 1997 Quarter Two, 2001 Quarter One, 2002 Quarter Two, 2003 Quarter Four, 2004

Note: List of initiatives derived from Wood [15]. This list comprises the presidential economic initiatives variable, with the variable being scored as a “1” in a specific quarter in a specific year where an economic initiative advocated by the president was formally approved and signed into law. The variable is scored as a “0” for those quarters where there was no presidential economic initiative that was formally approved

__________________________________________________________________________________________ © Copyright International Knowledge Press. All rights reserved.

80