Journal of Defense Studies and Resource Management ISSN: 2324-9315

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Review Article, J Def Stud Resour Manage Vol: 10 Issue: 3

Economic and Political Drivers behind the United States National Defense Expenditure

Bahram Adrangi* and Taren Bennett

Pamplin School of Business Administration, University of Portland, North Willamette Blvd, Portland, Oregon, Portland, Oregon, USA

*Corresponding Author:
Bahram Adrangi, Pamplin School of Business Administration, University of Portland, 5000 North Willamette Blvd, Portland, Oregon, Portland, Oregon, USA
E-mail: [email protected]

Received date: 26 April, 2022, Manuscript No. JDSRM-22-61866

Editor assigned date: 29 April, 2022, PreQC No. JDSRM-22-61866 (PQ)

Reviewed date: 13 May, 2022, QC No. JDSRM-22-61866

Revised date: 20 May, 2022, Manuscript No. JDSRM-22-61866 (R)

Published date: 27 May, 2022, DOI: 10.4172/ 2324-9315.1000148

Citation: Adrangi B (2022) Economic and Political Drivers behind the United States National Defense Expenditure. J Def Stud Resour Manage 10:3.

Abstract

This paper investigates the influence of economic and political factors that can affect the United States’ National Defense Expenditure (NDE). According to our data, the NDE depends on economic variables rather than social or political drivers. We show that both real gross domestic product and the unemployment rates have a positive correlation with the US NDE. However, the presidential party affiliation, or enjoying the majority or super majority in the Congress, does not have a consistent bearing on the NDE.

JEL Codes: H50, H53, H56

Keywords: US national defense spending; Politics of the US defense spending; Economics and US defense spending

Introduction

With the United States military presence around the world expanding, the size and scope of its operations are heavily scrutinized Peterson Foundation [1]. The US population is becoming more polarized in the debate around the role of the military throughout the world. Many believe that the most efficient means of national security is an occupational strategy where soldiers and other security assets are placed in a large number of countries across the world. As of 2017, the United States had 200,000 active duty service members deployed in at least 170 countries across the world Mc Carthy [2]. This strategy is meant to be proactive in deterring possible threats before they reach the US soil. Others argue that this overly aggressive stance causes more problems than it fixes due its large price tag and negatively affecting foreign relations.

Other researchers point to many long-term problems that are caused by conflagrations. Vine [3] discuss refugee problems related to the US activities abroad. They estimate number of a people displaced in the eight most violent wars the United States has waged since 2001 in Afghanistan, Iraq, Pakistan, Yemen, Somalia, the Philippines, Libya, and Syria. Using data from 2020 and 2021, they conservatively estimate that at least 38 million people have fled their homes.

Peltier [4] discusses the environmental and climatic problems that face the world and the US. Reallocating the scarce resources by reducing the military spending by $125 billion annually, as proposed by the sustainable defense task force, could potentially fund the green new deal. Economists from the IMF and elsewhere have estimated the environmental and climate change costs may be around 10% of US GDP by 2100.

Some think that the US needs to move back to a pre-cold war homeland strategy. This is where all or most national security assets and personnel work within the confines of our national borders. It’s argued that the US military is stretched too thinly across the world and has adopted a role of world police over protecting national interests instead. This strategy to shift prioritization of the US military from global to national interests makes the argument that the United States would experience the same level of security while cleaning up foreign relationships and drastically cutting the national defense expenditure.

The proposed 2020 Department of Defense (DOD) budget was the largest ever with a price tag of $750 billion. This is over 15% of the total US Federal spending, 50% of discretionary spending, and more than the next seven countries with the largest defense budgets combined (PGFF 2019). According to the PGPF, in 2021, the US NDE surpassed the total of defense spending budget of the next largest 11 defense spending budgets. Mintz [5] examines the association of the defense spending with business cycle, economic performance and the US military engagements.

With the advent of climate concerns and environmental threats to the globe, the NDE gains even a greater interest in the public debate. The diversion of scarce resources to NDE rather than Research and development focused on climate change and environmental challenges requires scrutiny. In addition, NDE may be a source of transfer of GDP from the tax paying public to the minority of the population that benefit from NDE as shareholders in defense related companies. In this paper, we aim to empirically examine the variables that are associated with the NDE. This research is important because the findings may point to the root causes of almost continuous military involvement of the US around the globe.

The objective of this research is to investigate the role of political and economic factors in forming the NDE. The investigation is important because if economic prosperity is associated with rising NDE, then NDE is diverting scarce national resources from other programs like health, education, food, shelter, and mental health. Furthermore, the consequences of rising NDE often would mean being involved in conflicts around the world. Therefore, one may conclude that rising NDE indirectly enables the perpetuation of military conflicts. The most recent example of misallocation of resources through conflicts is the twenty-year military intervention in Afghanistan. Almost every military conflict since World War II may fall in this category. While it is outside the scope of this paper, we know that environmental impact of the military conflicts are also enormous. The Watson Institute of International and Public affairs reports that the environmental cost of war on terror executed by the US military around the globe included 1.2 billion tons of greenhouse gasses just in 2001. The same report also finds:

  • The US department of defense is the world’ssingle largest consumer of oil and as a result, one of the world’s top greenhouse gas emitters.
  • Destruction of military base garbage in burn pits and other military operations have exposed soldiers and civilians to dangerous levels of pollutants.
  • Deforestation in Afghanistan as a result of illegal logging, particularly by warlords, has destroyed wildlife habitat.
  • In Iraq, increases in cancer, birth defects, and other conditions have been associated with war-related environmental damage and toxins.

Political leaders and policy makers may be able to use the results of this study to better allocate national resources to programs that promote human development rather than destructive conflicts.

Regardless of individual positions on this debate, the NDE is something that interests many. It’s become a major political talking point. While it is normally referred to as “bringing troops home” or “expanding the Department of Defense (DOD) capabilities,” what’s really being debated is the expanding or shrinking of the NDE. Traditionally, members of the Democratic Party campaign on the position of the pre-Cold War era strategy and aim to minimize the expenses while the members of the Republican Party campaign on the strategy of expanding DOD capabilities. This is just an overarching generalization because each party has members with varying stances.

NDE may be determined by whether the United States is in a state of conflict. It’s reasonable to assume that either political party in power could raise or lower the NDE depending on the level of threat, real or perceived, that faces the nation. It could also be that NDE is a luxury driven by the GDP. Economic resources can expand the government budget and allow the expanding NDE. This paper empirically analyzes variables that may have a role in the size of NDE.

Review of the Relevant Literature

Many research papers discuss the relationship between the NDE and other variables such as US economy, political parties and international relations. Notable examples include Nordhaus et al. (2012), Heo (2010), Björn et al. (2016), Ostrom (1978), among others.

Nordhaus [6] examined how the threat of armed interstate conflict, along with other manifestations of increased conflict, can lead to an increase in NDE. This study was conducted across 165 countries from 1950–2000 and analyzed these countries as a collective. These countries include a plethora of different economic and political conditions as well as different levels of conflict ranging from actual conflicts to arms races. Their findings were that a one percentage point increase in the aggregate probability of a fatal militarized dispute leads to a three percent increase in a country's military expenditures.

Heo [7] discusses the role that defense spending has on economic growth within the United States. This study uses an augmented Solow [8] to look at the United States from 1954–2005. Their study seems to indicate that the US economy is not significantly affected by varying levels of defense spending.

Björn [9] look into a case study of Germany from 1951-2011. They use an ideological measure and compare it to the change in military expenditure for the country. This study highlights a changing perspective across the country’s history. Up until 1960, it was found that military expenditure would increase by 2.4% points when the ideological measure increased towards the right wing by one standard deviation. However, since the 1960s, it seems like the ideological spectrum disappeared since every political party “agreed on how to evaluate international risks and threats.” This seems to introduce an important point that long-time series studies can be heavily influenced by changing ideological stances.

Ostrom [10] looks deeper into the policy making that goes into the United States military expenditure. They analyze the impact of war, presidential party and public opinion on individual decisions and how these all affect the policy process. They show that, in the United States, these variables only have effects over some steps of the policy making process but not enough to make substantial impact on the end result. Ostrom [10] is a great example of an analysis of a variety of variables and their effects on the NDE, or at least the process of forming the NDE, within the United States. However, this study is quite outdated now and needs to be updated with the last 40 years of data and ideological movement.

Nordhaus [6] and Björn [9] are great examples of variables that can affect a national defense expenditure. However, they do not look at the United States specifically. Theirs are examples of research that examines multi-nation and individual nation systems. Our paper extends their approach to the United States.

We draw on these studies to build a model that identifies and tests the role of economic and political variables and influences on the NDE. The main motivation for our study is to find out whether NDE is driven by economic affordability or political ideology.

Data

The data set we use for this paper came from a variety of sources. All economic variables were sourced from Federal Reserve Economic Data of St. Louis Fed (FRED). These variables include:

  • Real Gross National Product (RGNP)
  • M1 Money Supply (M1)
  • Real National Defense Expenditure (RNDE)
  • Net Domestic Investment (NDI)
  • Private Consumption Expenditure (PCE)
  • Un Employment Rate (UNEMPR)
  • Military conflicts (CONQ)

Our quarterly sample observations cover the second quarter of 1960 to the third quarter of 2019. In cases where the data were only available on annual basis, we convert the low frequency annual data to quarterly by fitting a polynomial of appropriate degree [11]. To avoid the issues of non-stationary, all variables are measured as percentage change. Unemployment rate is stationary. The lagged values of RGDP, PCE, M1, and NDI are included to accommodate possible serial correlation. Initially we include all the economic variables and eliminate those that were statistically insignificant. We briefly explain the possible association of each variable with NDE.

Real GDP measures the economic well-being and may be positively associated with the NDE. The higher levels of GDP may contribute to a larger budget for NDE. Similarly, domestic private investments and consumptions are proxies for economic affluence of a nation. The money supply (M1) is related to credit availability and interest rates. Higher levels of money supply are associated with lower interest rates. That could allow the government to borrow at low rates on interest for deficit spending. The US national debt in 2019 was at about the same level as the GDP. The more government ability to borrow, the higher possible NDE. The unemployment rate and the NDE may very well be positively associated. At higher civilian unemployment rate it would be easier to attract the unemployed labor force to the military. Thus, more recruits would mean a higher NDE [12].

Political variables are included in the analysis to examine if a particular political party gaining power has influence on the dependent variable. Political party influence can be separated into four distinct groups, Presidency (P), Majority (M), Super Democratic Majority (SDM), and Super Republican Majority (SRM). A political party is considered to be under the control of one of these groups for the year if they hold the office, or the majority vote of an office, for more days than any other party that year [13].

The first group is President (P); this is representative of whether a Republican or Democratic president held the office for that year. A dummy variable of 1 is used if the president is a Democrat and 0 if Republican.

Majority (M) represents which party held the controlling interest of two out of three of the following groups: House of Representatives, the Senate, and the presidency. A dummy variable of 1 is used to represent if the Democratic Party held two of the three controlling interests and 0 for Republican. Within our data window, 1960- present, only the Democratic and Republican parties have held the presidency or majority in any office, so we did not include any options for a third-party variable into our model [14,15].

Super Democratic Majority (SDM) represents cases when the Democratic Party held the presidency, a majority in the House of Representatives, and the Senate. A dummy variable of 1 is used to represent if the Democratic Party controlled all three groups and a 0 if they did not. A Super Republican majority (SRM) meant the same thing only the republican party held control over all three parties. The dummy variable for SRM is 1 if they held control of all three groups and 0 if they didn’t.

For our global conflict variable, we measured whether the US military was in a Conflict during a Particular Quarter (COPQ). Because the United States has been in an almost constant state of conflict since 1960, which is the starting date of our data, we had to be selective regarding what we consider a conflict year. Because of this we only counted years of major military engagements as conflict years [16,17]. We defined “major military engagements” as those deemed “wars” by the Department of Veteran Affairs (VA). This included the years of the Vietnam and gulf wars as well as the war on terror in Afghanistan from its start in 2001 to 2019. Start and stop dates to these conflicts are determined by the time that US troops participated in direct conflict only. Foreign government assistance after the pullout of US troops does not count as years of conflict. If conflict took place in any part of a year then the entire year is counted as a conflict year. Years the United States was engaged in a conflict by these criteria are designated with a dummy variable of 1, if not then they are designated with a 0.

Methodology

We estimate regressions starting with the largest number of variables and work our way down toward to statistically significant variables. In this process, if we suspect multicollinearity among explanatory variables, we eliminate some and retain only the ones that represent the group of multicollinear variables. Equation (1) is the initial model that captures the association of the percentage change in NDE with economic variables and their lags.

PCRNDE=f(RGDP, RGDP (t-1), RGDP (t-2), RGDP (t-3), RGDP (t-4), CONQ, CONQ (t-1),

CONQ (t-2), M1, M1 (t-1), M1 (t-2), NDI, NDI (t-1), NDI (t-2), PCE, PCE (t-1), PCE (t-2), UNEMPR, UNEMPR (t-1), UNEMPR (t-2)) (1)

Column 1 of Table 1 reports the estimation results for equation 1. Several coefficients are statistically insignificant even that 10%. This left us with only three relevant independent variables RGDP, CONQ and UNEMPR. The model consisting of the economic variables that had a statistically significant association with the dependent variable is represented by equation (2).

Major Economic Variables Relevant Economic Variables
Dependent Variable: Percentage Change in Real National Defense Expenditure
RGDP 0.0004 RGDP 0.0003
(-0.0003) (-0.0004)
RGDP (-1) 0.0008 b RGDP (-1) 0.0007 c
(-0.0003) (-0.0004)
RGDP (-2) 0.0007 b RGDP (-2) 0.0007 c
(-0.0003) (-0.0004)
RGDP (-3) 0.0004 RGDP (-3) 0.0004
(-0.0003) (-0.0003)
RGDP (-4) 0.0003 RGDP (-4) 0.0003
(-0.0003) (-0.0003)
PCE 0.0004 CONQ 0.0133 a
(-0.002) (-0.004)
PCE(-1) -0.0001 CONQ (-1) 0.0015
(-0.0025) (-0.0031)
PCE(-2) -0.0017 CONQ (-2) -0.0121 a
(0.0020) (-0.0043)
CONQ 0.0126 b UNEMPR 0.0056
(-0.0054) (-0.0041)
CONQ (-1) 0.001 UNEMPR (-1) 0.005
(-0.0076) (-0.0038)
CONQ (-2) -0.0122 b UNEMPR (-2) 0.0087 b
(-0.0053) (-0.0043)
M1 0.0003
(-0.0008)
M1 (-1) -0.0004
(-0.0009)
M1(-2) -0.0009
(-0.0009)
NDI 3.68E-06
(-8.56E-06)
NDI (-1) 3.64E-07
(-8.17E-06)
NDI (-2) -1.15E-06
(-0.000008)
UNEMPR 0.006
(-0.0041)
UNEMPR (-1) 0.0047
(-0.0044)
UNEMPR (-2) 0.0089 b
(-0.0041)
F 1.838 3.119b
R2 0.171 0.137
RESET 2.952 1.732

Table 1: Estimation results of impact of economic variables (1), Major economic variables (2) Relevant economic variables. Columns (1) Reports the estimation results of various economic variable’s impacts on the United States’ national defense expenditure. Columns (2) Reports the estimation results of the relevant economic variable’s impact on the United States’ National Defense Expenditure. LM statistic is for the Ramsey Test of Regression Specification Error Test (RESET).

a significant at 1% level, b Significant at 5% level, c Significant at 10% level.

PCRNDE=f(RGDP, RGDP (t-1), RGDP (t-2), RGDP (t-3), RGDP (t-4), CONQ, CONQ (t-1), CONQ (t-2), UNEMPR, UNEMPR (t-1), UNEMPR (t-2)) (2)

Adding the political variables to those in equation 2, we arrive at equations (3) through 5.

PCRNDE=f(RGDP, RGDP (t-1), RGDP (t-2), RGDP (t-3), RGDP (t-4), CONQ, CONQ (t-1), CONQ (t-2), UNEMPR, UNEMPR (t-1), UNEMPR (t-2), P, M, SDM) (3)

PCRNDE=f(RGDP, RGDP (t-1), RGDP (t-2), RGDP (t-3), RGDP (t-4), CONQ, CONQ (t-1), CONQ (t-2), UNEMPR, UNEMPR (t-1), UNEMPR (t-2), P , M, SRM) (4)

PCRNDE=f(RGDP, RGDP (t-1), RGDP (t-2), RGDP (t-3), RGDP (t-4), UNEMPR, UNEMPR (t-1), UNEMPR (t-2)) (5)

In the next section we report the estimation methodology and findings of equations (1) through (5).

Empirical Results

All regressions are estimated by the Newey-West Heteroscedastic Autocorrelation- Consistent (HAC) methodology to allow for serial correlation in the time series variables. Table 1 Column (1) indicates that our initial model has a number of statistically insignificant economic variables. Table 1 Column (2) shows only the economic variables that fall within the10% significance range. Moving from Table 1 Column (1) to Column (2) we see that even though we drastically reduce the number of variables in our model, the R2 only decreased by 0.034. Fairly unchanged R2 indicates that eliminating some variables did not reduce the explanatory power of the model. We conclude that statistically insignificant variables were possibly collinear.

With the elimination of the statistically insignificant variables, the F statistic raises from 1.838 in table one Column (1) to 3.120 in Table 1 Column (2), which is statistically significant at the 5% level. Because of this, we move forward with the variables in Table 1 Column (2) instead of those in the original model.

Table 2, Column (1) presents the estimation results with the inclusion of the party dummy variables (P, M, and SDM) to the Table 1 Column (2). The added political variables of P, M, and SDM are all significant at the 1% level. However, there is inconsistency with their coefficients. The coefficient for a democratic president is -0.010, which indicates that there is a negative relationship between democratic presidents and NDE. This is further supported by the democratic majority coefficient of 0.012. Meanwhile, the coefficient of the super democratic majority, 0.0187, seems to indicate that the relationship with NDE is positive. The interpretation is that when the democrats do not have super majority, they are forced to unite behind the party president. This results in a reduction in the NDE. However, if the democratic party enjoys a super majority, internal party disagreements on the NDE break out. This is plausible, as the party does not feel the pressure of the republican opposition and is likely to be less homogenous on their vote.

Democratic Party Variables Republican Party Variables
Dependent Variable: Percentage Change in Real National Defense Expenditure
RGDP 9.74E-05 RGDP 0.0004
(-0.0003) (-0.0004)
RGDP (-1) 0.0003 RGDP (-1) 0.0008 c
(-0.0004) (-0.0004)
RGDP (-2) 0.0003 RGDP (-2) 0.0008 b
(-0.0003) (-0.0004)
RGDP (-3) 0.0001 RGDP (-3) 0.0005
(-0.0003) (-0.0003)
RGDP (-4) 0.0002 RGDP (-4) 0.0004
(-0.0002) (-0.0003)
CONQ 0.0110 a CONQ 0.0123 a
(-0.0024) (-0.0036)
CONQ (-1) 0.0021 CONQ (-1) 0.0016
(-0.0025) (-0.003)
CONQ (-2) -0.0087 a CONQ (-2) -0.01207 a
(0.0031) (-0.004)
UNEMPR 0.0025 UNEMPR 0.0065 c
(-0.0036) (-0.0037)
UNEMPR (-1) 0.0025 UNEMPR (-1) 0.0060 c
(-0.0031) (-0.0036)
UNEMPR (-2) 0.0048 UNEMPR (-2) 0.0099 b
(-0.0035) (-0.004)
P -0.0096 a P 0.0008
(-0.0023) (-0.0032)
M -0.0124 a M -0.0065 b
(0.0027) (-0.0029)
SDM 0.0187 a SRM 0.0014
(-0.0038) (-0.0033)
F 9.982a F 4.112a
R2 0.396 R2 0.212
LM 18.273a LM 2.088

Table 2: Estimation results of impact of political variables (1), Democratic Party Variables (2), Republican Party Variables. Columns (1) Reports the estimation results of relevant economic and political (Democratic Party) variable’s impacts on the United States’ National Defense Expenditure. Columns (2) Reports the estimation results of the relevant economic and political (Republican Party) variable’s impact on the United States’ National Defense Expenditure. LM statistic is for the Ramsey test of regression specification error test (RESET).

a Significant at 1% level, b Significant at 5% level, c Significant at 10% level.

Table 2, Column (1) also shows the same situation with the conflict variable (CONQ). While the CONQ and CONQ (t-2) are statistically significance at the 1% level, CONQ and CONQ(t-1) show a positive association with NDE, the coefficient of CONQ(t-2) is negative. We interpret this finding to mean that current budgets are mainly associated with the ongoing conflicts and influence the NDE levels.

Table 2, Column (2) presents the estimation of the equation (1) with the Republican party variables included. Both the RGDP (t-1) and RGDP (t-2) are statistically significant at the 10% level, and both have a positive relationship to PCNDE. Once again, both the CONQ and CONQ(t-1) are positively associated with the NDE. CONQ (t-2) show a significance level of 1%, but CONQ(t-2) has a negative coefficient of -0.01207. UNEMPR and its lagged values are statistically significant and have a positive relationship with PCRNDE. This observation is plausible and shows that high unemployment rate may be associated with the rising NDE as more pool of recruits becomes available.

The only political variable that is statistically significant in Table 2, Column (2), is the majority variable. It’s significant at the 5% level, and it has a negative coefficient of -0.0065. The other political variables, the presidency and super Republican majority, both have positive coefficients but are statistically insignificant. The upshot of these findings is that the politics of the ruling party may not be a critical factor in determining the NDE. The economic variables and being in military conflict may be the main factors associated with the NDE.

The F statistics for both equation estimates in Table 2 are statistically significant. However, the low R2 and the RESET test indicate the estimated models have some room to improve. Based on the findings of the RESET test, we added the squares of the explanatory variables to capture possible nonlinearity in the association of the NDE and the political and economic variables. This experiment did not qualitatively improve the model estimates.

Table 3 shows our final model, which includes only variables that showed consistency in coefficient signs and statistical significance. This table shows that the lagged real GDP and the unemployment rate are positively and significantly associated with the NDE. Overall, the empirical findings confirm that while political party preferences may be playing a role in the US national defense spending, the health of economy is consistently associated with the national defense spending. The interpretation of our findings is that NDE is driven by economic consideration and is independent of political parties. It may be deducted that ultimately the US GDP enables it to boost its defense spending and perhaps drives the US military involvement around the globe.

Significant Economic Variables
Dependent Variable: Percentage Change in Real National Defense Expenditure
RGDP 0.0004
(-0.0004)
RGDP (-1) 0.0006
(-0.0004)
RGDP (-2) 0.0006 c
(-0.0004)
RGDP (-3) 0.0004
(-0.0003)
RGDP (-4) 0.0003
(-0.0003)
UNEMPR 0.0063
(-0.0039)
UNEMPR (-1) 0.0052
(-0.0038)
UNEMPR (-2) 0.0074 c
(-0.0041)
F 2.231537b
R2 0.075373
LM 8.310 b

Table 3: Estimation results of impact of significant economic variables (1).

Notes: Columns (1) Reports the estimation results of significant economic variable’s impacts on the United States’ national defense expenditure. LM statistic is for the Ramsey test of Regression Specification Error Test (RESET).

a Significant at 1% level, b Significant at 5% level, c Significant at 10% level.

Including nonlinear quadratic for didn’t improve the model.

Our findings corroborate the findings of Ninicic and Cuasack [18]. They show that for the period of 1948-1976 for the US, the main factors that drive the military spending dynamics were economic considerations. Specifically, the stabilizing effects of such government expenditures on the aggregate demand were significant. Furthermore, they find that the perceived positive economic impact of defense spending translated into political and electoral advantage.

Summary and Conclusion

This paper aims to empirically investigate the US national defense spending and its association with the political parties and the US economy. We examine several regression models that include party role and domestic economic variables as well as the levels of geopolitical conflicts.

Our findings, based on the quarterly sample data spanning 1960 through 2019, suggest that the only statistically significant variables that consistently act upon the national defense spending of the US are the economic variables real GDP and the unemployment rate. Specifically, these two variables seem to be the most prominent and statically significant at the 10% level. The overall regression’s F statistic also indicated that the model is statistically significant at the 5%level based on the F test.

We eliminate all other economic and political variables from the model because of their inconsistent association with the national defense spending. Most economic variables, except for the real GDP and the unemployment rate, were statistically insignificant in all models.

The variable accounting for the involvement of the US in global military conflicts (CONQ) was statistically significant in most models. However, since the coefficients were commonly both negative and positive for different lagged value, this variable may be not specific enough to account for the association of the defense spending and conflicts. Furthermore, the United States has been in conflict for most of the sample period that our model takes place, from 1960 to the present. Since 2001, our data recorded conflict in every year, so it is plausible that the variable CONQ has no role in NDE determination. We were aware of this possibility in our data collection phase, however, tested the association of CONQ with NDE empirically.

The final estimation results show that political variables for both the democratic and republican parties were eliminated. The democratic variables of P, M, and SDM (Super Democratic Majority) were all statistically significant at the 1% level. However, because the President (P) and Majority (M) switched signs form positive based on democrats having super majority, we found their association with the defense spending erratic and deleted them from the final model.

This inconsistency within the Democratic Party may have a few explanations. The party runs on the platform of lowering the NDE, but when the Democratic Party has super majority with no other political party scapegoat, then they prioritize national defense more heavily. It is also possible that they are consistently fighting for larger expenditure across the board but can only pass these bold proposals when they have a super majority. These are only speculative and are just a couple of many different explanations for this inconsistency. Regardless of the reasoning, the data suggest that the parties are not consistent with their NDE policies across all walks of political power, and we deleted them out of the final model.

The Republican party’s effect on the NDE seemed to be much more randomized. The presidency and super majority had a positive relationship to the NDE, but these results were insignificant. The only variable that was statistically significant was majority and that suggested a negative relationship to the dependent variable. The data suggest a wide variety of stances with respect to the NDE across varying levels of political influence or no hard position; because of these inconsistencies, we also ruled the Republican Party variables out of the final model.

The only variables that remained after our eliminations were major economic variables which seems to suggest that economic conditions are the driving force behind the NDE. The data suggest that an increase in the real GDP and the unemployment rate both contribute to an increase in the national defense expenditure.

We have two theories as to why this might be. The first theory is that an increase in GDP leads to a larger taxable income for the US federal government to tap into. The larger tax income could lead to a bigger budget across the board which just affects the NDE to the same extent as any other government program. The second theory is that the NDE is being used as a tool to lower unemployment rates. As unemployment rates rise, the US federal government could increase the NDE to create more incentive programs, such as signing bonuses, to help with recruiting purposes. As more people are enticed into the military the unemployment rate would fall. The larger NDE could also go towards defense manufacturing or other contracting services which would encourage further hiring of personnel into these industries to help supply the increased demand for products or services.

The main findings of this study are that, first; defense spending is driven by a complex interaction of domestic and international political as well as economic considerations. Secondly, economic infrastructure contributes to national defense spending and potentially engages the US in many global conflicts that are not sustainable on any grounds, including politically, environmentally and in terms of human suffering. Finally, our findings show that wealthy nations are more likely to afford waging disastrous wars around the globe. President Eisenhower warned the nation about the perils of the military-industrial complex and the pressures to engage in conflicts in his last presidential speech. Conflicts redirect and drain the nation’s resources from constructive endeavors toward destructive ones. Therefore, imposing fiscal discipline by the US congress on all branches of the government would be a crucial step in the right direction. The US national debt currently exceeds the US GDP. Given that the US military spending is 3.5% of the GDP, fiscal responsibility will have serious impact on curtailing the national debt and reducing the odds of the US involvement in international military conflicts. Domestically, wealthy nations, including the US, can enact amendments to the processes that allow wider direct input by the populace beforeinvolvingnationsinglobalconflicts.

Internationally, strengthening organizations such as the UN Security Council and the International Court of Justice are critical for reducing conflicts.

References

Track Your Manuscript