Table of Contents

This report focuses on how Coloradans perceive the economic well-being of the middle class in their state and whether they believe significant changes in U.S. foreign policy could affect that well-being. It contextualizes those perceptions with relevant quantitative data. It has been written with U.S. foreign policy and national security professionals in mind.

Many U.S. foreign policy and national security professionals—who by necessity spend most of their time dealing with geopolitical and security developments abroad—worry that they may be disconnected from the economic realities that Americans confront. They may also be unaware of how current foreign policy debates in Washington, DC, and on the presidential campaign trail—related to issues such as trade policy, foreign aid, defense spending, and international approaches to climate change and energy—stack up with what Americans outside the political establishment actually experience and care about. It is difficult for them to weigh in on these debates, having, so far, largely deferred to domestic policy counterparts and economists to determine how the U.S. role abroad may or may not intersect with the economic well-being of the middle class at home. It is for these reasons, and for the benefit of these foreign policy and national security professionals, that Carnegie initiated research on “foreign policy for the middle class” and convened a bipartisan task force.

The task force members have served in senior policy roles, under Democratic and Republican administrations, at the departments of commerce, defense, state, and treasury; the Office of the U.S. Trade Representative; the National Intelligence Council; and the National Economic Council and the National Security Council at the White House. All the members believe that, at various points in their government service, they did not test their assumptions often enough about how the U.S. role abroad could be affecting the economic well-being of the middle class at home. They have, therefore, come together to provide strategic direction to this research, advise on where data gathered by the research team could be interesting or surprising to officials in the types of positions they once held, and ultimately offer detailed recommendations.

To inform these recommendations, Carnegie’s task force and research team is working with university researchers to gather data in three U.S. states in the nation’s heartland—Ohio, Colorado, and Nebraska. The task force deliberately chose to focus on the middle class in these three states, given the prevailing criticism that the nation’s top earners and executives in coastal cities have disproportionate access to and influence on those developing foreign policy in Washington, DC. Americans living outside of coastal cities rarely get a chance to interact with foreign policy professionals. This research aims to lift up their voices.

Who Is the Middle Class and How Is Foreign Policy Defined?

In the context of this project, “middle income” refers to households whose incomes range between two-thirds to double the median income, adjusted for household size and local cost of living. This is the widely employed definition from the independent Pew Research Center that allows for comparisons over time and across different metropolitan areas throughout the United States.1 Table 1 shows the middle-income range and median income in Colorado. “Middle class” or a “middle-class lifestyle” are more subjective terms that can refer to, among other things, income, wealth, educational status, potential for upward mobility, or social standing.2 Those interviewed in Colorado for this study tended to define a “middle-class lifestyle” as the ability to secure a job with adequate pay and benefits that enables them to meet their monthly expenses, tend to their families’ medical needs, buy a car, own a home, help their kids pursue decent postsecondary school education, take an annual vacation, save for retirement, and not be saddled with crippling debt.

“Foreign policy” serves as shorthand for the spectrum of foreign, defense, development, international economic, trade, and other internationally oriented policies perceived by those interviewed as most impactful to their economic well-being. Interviewees associated foreign policy with some issues that typically fall under the purview of domestic policy, such as foreign direct investment, immigration, and energy and climate change. Further information on the definition of terms, the rationale for the project, and relevant historical context can be found in the introductory chapter of the first report on Ohio.3

Why Colorado?

Like Ohio, Colorado’s political and economic diversity make it an excellent proxy for national debates. It has some of the most liberal counties in the country, such as Boulder and Denver, but also some of the most conservative, such as El Paso (where Colorado Springs is located) and Mesa (where Grand Junction is located). Other counties, especially in the suburban areas, proudly espouse centrist tendencies and are often political “toss-ups” in local-, state-, and national-level elections. Peace activists and environmental leaders in Colorado coexist alongside entire communities dependent on defense spending and extractive industries.

Colorado also offers a good counterpoint to Ohio. While Ohio’s per capita income relative to the rest of the nation has steadily declined since the mid-1950s, Colorado’s has climbed. Similarly, Ohio’s population growth has decreased, while Colorado’s has accelerated. Ohio trails the nation in residents with college degrees, whereas Colorado ranks near the top. Ohio is heavily dependent on goods manufacturing and international trade, while Colorado relies far more on professional services and domestic markets. See Table 1 for additional key state-level comparisons.

Therefore, Colorado—a state with similar political and economic debates to Ohio but with a different economic outlook and base—helps shed new light on the perceived and measurable economic effects of U.S. foreign policy on the middle class.


It is not possible to credibly quantify or model the economic impact of the sum total of all U.S. foreign policy activity on a specific income group within a single state. There are far too many variables involved, including the wars the U.S. wages or prevents, the stability for the global economy it provides, the commerce it enables, the trade in goods and services it conducts, the foreign investment it makes or receives, the aid it delivers, and the friendly relations it forges. Therefore, this report focuses on perceptions of the economic impact of foreign policy on the middle class and the ways in which those perceptions stack up against hard data. To gauge such perceptions, Carnegie’s research team largely conducted focus groups and one-on-one interviews, where it was possible to engage in conversation and seek clarification on complicated issues.

The research team, including Salman Ahmed and Allison Gelman at the Carnegie Endowment for International Peace and Richard Wobbekind and Brian Lewandowski at the Leeds School of Business at the University of Colorado Boulder (CU Boulder), engaged more than 125 individuals between March and June 2019. The research took place in ten of Colorado’s sixty-four counties, including in each of its three main regions—the Front Range, the Western Slope, and the Eastern Plains—to capture its socioeconomic and political diversity.

Front Range

  • Boulder (city and county): A liberal, affluent university town (CU Boulder) that is home to several federal labs and prospering tech, engineering, and defense sectors.
  • Denver (city and county): A liberal city with a thriving economy anchored by state government, universities, information technology (IT) services, financial services, tourism, and the Denver International Airport.
  • Arapahoe and Douglas: Two affluent, highly educated, and politically moderate suburban counties in the Denver area that used to reliably vote Republican but are now considered “swing areas”; they host the headquarters of several Fortune 500 companies, such as Arrow Electronics, Dish Network, and Liberty Interactive.
  • El Paso (Colorado Springs): The state’s most populous county after Denver, with an economy heavily oriented around several military installations, including the U.S. Air Force Academy, and the defense industry; El Paso is well-known for its conservatism and active evangelical community.
  • Pueblo (city and county): An ethnically and politically diverse manufacturing town on the southern end of the Front Range that arguably bears as much resemblance to the industrial Midwest as it does to other parts of Colorado; Pueblo voted for Barack Obama (Democrat) in the 2008 and 2012 presidential elections and for Donald Trump (Republican) in 2016.
  • Weld (Greeley): A conservative-leaning county on the northeastern edge of the Front Range that is one of the nation’s leading oil and gas producers and that has a strong farming community.  

Western Slope

  • La Plata (Durango): A traditionally conservative ranching county that is home to the fast-growing, liberal city of Durango, which relies heavily on tourism and outdoor recreation.
  • Mesa (Grand Junction): A conservative county whose fortunes have been heavily tied to mining and oil and gas extraction but is now diversifying its economy and leveraging the presence of Colorado Mesa University and its position as a major commercial and transportation regional hub.

Eastern Plains

  • Otero (La Junta, Rocky Ford, and Fowler): A conservative-leaning, but politically mixed, rural county with a large Hispanic population; it relies heavily on agriculture and related manufacturing and has among the state’s lowest levels of growth and per capita income.

More places on the Front Range were studied because the region accounts for the vast majority of the state’s population and economic activity.

In all of the case study areas, CU Boulder leveraged its own contacts and worked with local economic development organizations and community leaders to identify focus group participants who could speak authoritatively on the local economy and middle-class fortunes. Many participants characterized themselves as middle income. The groups as a whole included state and local government officials, economic developers, teachers, first responders, healthcare professionals, small business and farm owners, and middle managers in large firms. Participants in the upper-income bracket spoke as employers of, and policy-setters for, middle-income workers.

See Appendix A for a list of the individuals engaged, and Figure 1 for more statistics on each selected county and for the voting patterns by county.

Research Bias

The majority of interviewees and focus group participants held positions of responsibility and/or served as leaders within their respective communities. There was insufficient representation from those in the lower-middle class, minority groups, and millennials, because they do not occupy as many public leadership positions relative to their overall numbers. The research team attempted to correct for this bias by supplementing the interviews with data drawn from other studies on Colorado’s middle class, such as from the Bell Policy Center, which gives attention to underrepresented groups. Additionally, to stay focused on lifting up local voices not heard in Washington, DC, formal interviews were not conducted with some national leaders who may be best placed to connect the dots between the U.S. role abroad and its economic impact at home (for example, top executives in multinational corporations, representatives of national-level business and trade associations, and members of Congress).

To help deal with political biases, the research team deliberately conducted interviews and focus groups in areas that voted heavily for Trump or Hillary Clinton in the last presidential election, as well as in places more evenly split. The authors indicate where there appeared to be a correlation between the views being expressed and the politics of the interview locations. And they note where some individuals voluntarily self-identified themselves as conservative or liberal. Notwithstanding, because focus group participants were not required to indicate their political affiliations, no definitive judgments were made in this study about the link between political preferences and opinions on foreign policy. And the research team generally steered the conversations back toward what people could comment on based on their actual experiences, rather than simply echoing what they had heard from national-level political figures and media personalities.

The questions posed during the interviews and focus groups, as well as the supplemental research, were consciously biased toward the economic effects of foreign policy. Some foreign policy concerns, such as the threat of terrorism, would probably have been expressed more frequently had the discussions focused on noneconomic factors. Likewise, the views expressed on issues like immigration may have differed if the conversations focused on cultural issues rather than economic considerations. Nevertheless, some noneconomic factors, such as values, identity, and political affiliation did come up occasionally and, as recent research suggests, likely played a role in the views expressed by those interviewed.5

Finally, although the Colorado study is significantly informed by the participants’ views, the qualitative research was just a starting point. Based on emerging themes, the team conducted additional research and collected quantitative data. The analysis and thoughts in the concluding chapter also draw on the personal experiences of the task force members.


1 Rakesh Kochhar, Pew Research Center, “The American Middle Class Is Stable in Size, but Losing Ground Financially to Upper-Income Families,” September 6, 2018,

2 For a more detailed account of how the definition of “middle class” can vary, see Richard V. Reeves and Katherine Guyot, Brookings Institution, “There Are Many Definitions of ‘Middle Class’—Here’s Ours,” September 4, 2018,

3 Salman Ahmed, et al., “U.S. Foreign Policy for the Middle Class: Perspectives From Ohio.”

4 Sources for Table 1:

Middle-Income Range: Kochhar, “The American Middle Class Is Stable in Size, but Losing Ground Financially to Upper-Income Families.”

Population and Population Growth: U.S. Census, "Quick Facts: United States, Colorado, Ohio,”,OH,US/PST045218, accessed May 29, 2019.

Population With a Bachelor’s Degree or Higher: U.S. Census, “Educational Attainment: 2017 American Community Survey 1-Year Estimates,”, accessed May 29, 2019.

GDP: Bureau of Economic Analysis, “Annual Gross Domestic Product by State: Millions of Current Dollars,” last revised May 1, 2019,, Accessed May 29, 2019.

Median Value of Owner-Occupied Housing: U.S. Census, “Median Value (Dollars) Owner-Occupied Housing Units 2017 American Community Survey 1-Year Estimates, ”, accessed May 29, 2019.

Poverty Rate: Center for American Progress, Talk Poverty, “Poverty Data—Overall Poverty 2018”, accessed April 10, 2019.

Unemployment: Bureau of Labor Statistics, “Regional and State Unemployment—2018 Annual Average,” February 28, 2019,, accessed May 29, 2019.

Top Nongovernment Employers: Colorado Department of Labor and Employment,; ReferenceUSA,; D&B Hoovers,; Ohio Development Services Agency, “Ohio Major Employers—Section 1,” May 2018,; and Jennifer Calfas, Money, “The 6 Biggest Employers in the U.S. Right Now,” April 27, 2017,

Manufacturing Employment: Bureau of Labor Statistics, “Quarterly Census of Employment and Wages, 2018 Annual Averages,” data extracted September 24, 2019,

Defense Spending: U.S. Department of Defense, Office of Economic Adjustment, “Defense Spending by State, Fiscal Year 2017, Revised Version March 2019,” March 2019,

Goods Imports: U.S. Department of Commerce, International Trade Administration, “State-by-State Imports From a Selected Market: 2018 NAICS Total All Merchandise Imports From World, ”, accessed August 2019. Calculation for the percent of GDP uses nominal 2018 GDP (current dollars): Bureau of Economic Analysis, “Regional Data: Gross Domestic Product (GDP) by State: All Industry Total (Millions of Current Dollars),” May 1, 2019,, accessed August 2019.

Goods Exports: U.S. Census Bureau, USA Trade Online, data extracted May 10, 2019,; calculation for the percent of GDP uses nominal 2018 GDP (current dollars): Bureau of Economic Analysis, “Regional Data: Gross Domestic Product (GDP) by State: All Industry Total (Millions of Current Dollars),” May 1, 2019,, accessed May 16, 2019.

Service Exports: Nick Marchio and Joseph Parilla, “Export Monitor 2018,” Brookings Institution, April 30, 2018,

Party Affiliations: GovTrack,, accessed April 2019; United States House of Representatives History, Art, and Archives, “Party Divisions of the House of Representatives, 1789 to Present,”, accessed September 5, 2019; 270towin, “Colorado” and “Ohio,”, accessed September 5, 2019.

5 Stephen Hawkins, Daniel Yudkin, Míriam Juan-Torres, and Tim Dixon, Hidden Tribes: A Study of America’s Polarized Landscape (New York: More in Common, 2018), According to polling from the nonpartisan Pew Research Center about the 2016 presidential campaign, 40 percent of those who voted for Trump said their main source of news was Fox; 39 percent who voted for Clinton said their main source of news was CNN, MSNBC, NPR, or the New York Times. See Michael Barthel, “Trump, Clinton Voters Divided in Their Main Source for Election News,” Pew Research Center, January 18, 2017,

6 Sources for Figure 1:

Population, Population Growth, Hispanic or Latino, Foreign Born, Bachelor’s Degree or Higher, Median Household Income, Owner-Occupied Housing Median Value: U.S. Census, “Quick Facts: Boulder, Denver, Arapahoe, Douglas, Weld, El Paso, Pueblo, Otero, La Plata, and Mesa Counties, Colorado,” data extracted June 13, 2019,,denvercountycolorado,arapahoecountycolorado,douglascountycolorado,jeffersoncountycolorado/PST045218.

Unemployment: Bureau of Labor Statistics, “Labor Force Data by County, 2017 Averages,” February 28, 2019, data extracted April 11, 2019,; and Bureau of Labor Statistics, “Local Area Unemployment Statistics: Colorado,” extracted April 19, 2019,

Workforce (Natural Resources and Mining, Manufacturing, Professional and Business Services): Bureau of Labor Statistics, “Quarterly Census of Employment and Wages, 2018 Annual Averages,” data extracted June 17, 2019,

Industry Share of GDP (Natural Resources and Mining, Manufacturing, Professional and Business Services): IMPLAN, 2017 data, data extracted May 19, 2019,

Defense-related employment: Paul Rochette, Shannon Anderson, and Tom Binnings, “The Economic Impact of Department of Defense, Veterans and Military Retirees, and the Department of Veterans Affairs Activities in Colorado,” Summit Economics, May 2018,; Total Employment: Bureau of Economic Analysis, “Local Area Personal Income and Employment, ”

Defense-related employment is calculated on a county level according to following criteria. Direct employment involves direct Department of Defense and Veterans Affairs expenditures, including employment from military installations, Department of Defense (DOD) contracts and assistance awards, the National Guard and Reserves, Tricare, and Veteran Administration operations, programs, and contracts. Indirect impacts come from an entity’s expenditures in the economy, such as support services and facility maintenance. Induced impacts come from expenditures of earnings, including expenditures by veteran and military retirees with pensions. Employment numbers also include trade flow effects: jobs created in one county due to DOD or Veteran Administration expenditures in another. The total calculated employment is then taken as a share of total 2016 jobs (both full-time and part-time jobs, including wage and salary jobs, sole proprietorships, and individual general partnerships).