Abstract
Legalizing marijuana at the state level establishes via fiat a new industry. In practice, legal marijuana constitutes a large-scale test of consumption-driven economic development policies, which seek to shape industry and job growth via changes in local spending. Drawing on the critical case of Colorado, I assess the economic development outcomes and future potential of legal marijuana, and evaluate the broader applicability of consumption-driven development processes heretofore confined to the arts and isolated local cases. While most states should be able to replicate Colorado’s import-substitution gains, the most enduring economic benefits to legalization require place amenities that legalization alone cannot provide.
Introduction
Marijuana legalization represents both a change in social norms and the development of a new consumer industry. Each of the four US states currently allowing recreational marijuana businesses reports strong growth in monthly sales, production, and business services. In politics and discourse, legalization is strongly associated with Colorado, which became the first state to sell the drug recreationally in January 2014. Colorado’s subsequent surges in tax revenue, population, and economic expansion anchor economic arguments for legalization—arguments that led voters in California, Massachusetts, Maryland, and Nevada to approve legalization in the November 2016 elections. Making sense of these sweeping legal changes means coming to terms, both conceptually and empirically, with a newly minted industry that confounds both standard economic development measures and the long-standing belief that exports, rather than shifts in consumption, drive regional economies.
Legal marijuana businesses contribute directly to the central economic development goals of employment growth, business expansion, and economic diversification. Yet they do so through the unconventional mechanism of shifting consumer spending. Once treated as an impossibility by regional economic theories that prioritize material exports as the foundation of growth, the idea of consumption-driven development now finds traction with local authorities seeking a measure of control over economies otherwise constrained by the threat of capital mobility (Markusen and Schrock 2009). Marijuana legalization effectively constitutes a large-scale test of the potential to generate economic growth and diversification by shifting local consumer spending from imported consumer goods to locally produced goods and services. In the short term, mapping the pathways through which redirected consumer spending contributes to economic growth can provide a guide to evaluating the potential of the emerging, multibillion-dollar marijuana industry. Perhaps more important, the overnight emergence of a legal marijuana industry in Colorado provides a large-scale empirical test of consumption-driven development ideas for which large-volume data remain elusive: It provides the possibility to observe in real time the development of new products, services, and business specializations through the interactions of consumers and entrepreneurs.
To date, assessments of the economic outcomes to shifts in local consumption suffer from the absence of spatial data on household consumption patterns, a condition that has necessarily limited research to small-scale case studies (Markusen 2013). As a sudden, highly regulated shift in consumer spending, marijuana legalization presents a systematic test of the processes identified in these case studies—a way to examine at scale the diffuse processes that constitute consumption-driven economic development. Accordingly, this article undertakes an in-depth study of Colorado’s emerging marijuana industry to both assess the probable impacts of marijuana industries elsewhere and to evaluate the postulated mechanisms through which consumption changes lead to economic growth and diversification. As the first US state to permit marijuana retail sales, Colorado represents what Yin (2008) terms a critical case: An object of study with a unique and unequaled ability to bring together the research phenomena of interest. Because Colorado’s marijuana retailers and producers have been in business longer than their peers elsewhere, and because the state’s marijuana business service providers enjoy a first-mover advantage in the emergent national market, Colorado’s marijuana industry provides more evidence than others on the ways in which marijuana production and sale mature.
My analysis of legalization in Colorado offers two related but distinct sets of findings. For the marijuana industry, it indicates that other states will likely be able to reproduce Colorado’s short-term gains, as licensed marijuana sales effectively amount to legally mandated import substitution. However, Colorado’s development of advanced marijuana services and its appeal to lifestyle migrants results from the state’s first-mover status and from its long-term inculcation of do-it-yourself culture. These gains are more durable and significant in development terms and far more difficult for other states to develop. Seen this way, the mechanics of development in Colorado suggest that consumption-driven development shares the same goals, development processes, and susceptibility to place-luck as conventional development. In other words, it is likely to favor places that already enjoy a favorable position in subnational economic competition.
From Illicit Substance to Big Business: Evaluating the $5-Billion Cannabis Industry
Full retail legalization of marijuana sales in Colorado, Washington, Oregon, and Alaska arrived not as an abrupt shock but rather as the escalation of decriminalization procedures that have gradually changed the organization and legal status of marijuana production over the past twenty years. Liberalization began with the approval of medical marijuana sales in California in 1996, a move that halted the long-term trend toward heightened criminalization. Gradual movement preceded the current spate of outright legalization: Twenty-three US states now have medical use laws, and three more have decriminalized possession.
As a precursor to outright legalization, medical marijuana laws provided legitimate commercial channels through which entrepreneurs could gain business experience, develop proprietary cannabis strains, and establish the retail practices and product diversification that now characterize legal marijuana sales. Medical use also seeded the partial consolidation of medical marijuana into conglomerates owning multiple facilities, and the development of industry-specific legal and professional services. With few exceptions, the cannabis businesses with the largest market capitalizations and fastest rates of growth are owned and operated by veterans of the medical marijuana industry (IBISWorld 2015a, 2015b).
The simplest argument for legalization concerns tax revenue. Marijuana legalization creates at the stroke of a pen businesses subject to a range of new taxes, such as a 25 percent excise tax on producers in Washington and a 15 percent retail tax in Colorado. The variance in state-level taxation procedures signals the importance of state law in shaping the industry’s structure and trajectory. Following the US Department of Justice’s “Cole” memo, which prohibited the movement of marijuana across state lines, Colorado, Washington, Alaska, and Oregon all mandate seed-to-sale tracking systems to ensure that plants remain in-state from seed to smoke (US Department of Justice 2013). However, other regulations integral to industry structure and operation vary substantially by state. For example, Oregon and Colorado allow individual-level cultivation, which Washington law prohibits. Colorado law devolves retail licensing authority to municipalities, which have no power to deny licenses in the Pacific Northwest (Vermont Legislature 2016). Highly variable from state to state, these and other technical measures shape the industry’s growth, consolidation, and competitive standards. The industry’s primary economic development mechanisms, however, are less sensitive to place.
Import Substitution, Industry Diversification, and Human Capital Attraction: How Consumption Industries Support Economic Development
Consumption-driven economic development follows a number of intuitive pathways that resist easy measure. Embodied in policies that prioritize human capital recruitment and the potential contributions of highly skilled individuals to new products and industries, a consumption-focused approach values consumer changes both as a means of creating distinctive places with appeal to professionals and for the instrumental role that shifts in spending can play in seeding new products and industries (Markusen 2004; Mathur 1999). Hypothesized processes of consumption-driven development unite growing interest in culture and the arts (Markusen and Schrock 2006), innovation (Beauregard 2014), urban agriculture (Vitiello and Wolf-Powers 2014), and local artisanal industries (Heying 2010). To date, tests of consumption-driven development have suffered from poor data availability and the micro character of the processes at play: The best evidence necessarily comes from intensive, process-based research exploring the ways consumer spending shapes growth in particular places. This limitation has precluded the broader testing of hypotheses.
Individually, studies of consumption-driven development treat the process as an alternative to the industrial development familiar to economic development practitioners. Yet as Markusen and Schrock (2009) argue in an influential article, consumption-driven and export-driven development share many basic mechanisms. Incorporating the language, analytical units, and mechanisms of classic economic development, they identified four specific pathways through which changes in local spending can add to the overall base of economic activity and employment: (1) by shifting spending toward goods produced locally, (2) by shifting spending toward goods with higher labor content, (3) by supporting the growth of local industries that will eventually export goods, and (4) by building local consumption opportunities that draw highly skilled in-migrants (Table 1).
Pathways from Consumption to Economic Development.
Source: Adapted from Markusen and Schrock (2009).
Regardless of the incessant policy focus on export-oriented and manufacturing industries, the long-term development of regional economies and export capacity has always relied on the development of unique local services, consumption practices and cultural norms (Tiebout 1956; Blumenfeld 1955; Piore and Sabel 1984). Markusen and Schrock’s synthesis of these analytical traditions effectively identifies two sets of measurable changes resulting from shifts in consumption. The first concerns substitution, from imported goods to those made locally, and from goods made by machines to those made by humans. These shifts fit comfortably with economic development practice, and proceed more or less mechanically: The ability of places to shift consumption toward local or high-labor-content goods and the outcomes to those shifts varies, in response to industry and market structure, the production process, and labor market conditions (Fitzgerald and Leigh 2002; Peters and Fisher 2004).
These immediate-term substitutions concern the short-term effects of economic development, commonly represented through measures of industry employment and output. However, the second set of hypothesized consumption shifts manifest in terms of long-run development. Places with distinctive consumption practices grow and change differently over the long term. Over the past fifty years, population migration into the Mountain West and Pacific coast reflects both conventional measures of economic opportunity (jobs, income, real estate) and the valorization of natural, urban, and cultural amenities by mobile professionals (Busch 2015; Jurjevich and Schrock 2012). Western states have also drawn on residents’ enthusiasm for beer, wine, natural foods, and outdoor recreation to develop nationally leading industrial concentrations in brewing, distilling, natural foods, and outdoor apparel (Clark 2013). Here, cities and urban regions benefit from the co-location of production and consumption, which mutually strengthen one another by allowing firms and product developers to test emergent innovations easily and quickly (ibid.).
The four pathways from consumption changes to development thus provide a detailed framework for analyzing the development of legal marijuana manufacture and retail. And, perhaps more important, Colorado’s experiment in legalization fashions a large-scale case through which to test both the extent to which changes in consumption can trigger economic development and the essential and contingent factors—the preconditions, accidents, contingencies, and choices—necessary for them to do so. As a large-scale, data-rich test of these processes, marijuana legalization in Colorado provides the empirical materials with which to begin to qualify how, where, when, and under what conditions particular consumption shifts may generate economic growth.
The Other Green Jobs: Legal Marijuana Comes to the Centennial State
When 54 percent of Coloradans voted for Amendment 64 in the November 2012 election, they formalized local-consumption preferences that had long prevailed in the state. Marijuana legalization emerged from the same consumption practices that had led homebrewers to found New Belgium and Oskar Blues brewing, wildflower enthusiasts to found Celestial Seasonings Tea, and nutritional activists to found gluten-free food manufacturing firms. Similarly, the origins of legalization play a key role in shaping the marijuana industry.
Like brewing in particular, marijuana production benefits from the population’s strong taste for intoxicants. Prior to legalization, Colorado routinely ranked as a top-five state for per-capita marijuana consumption (Hedden et al. 2015). This high aggregate consumption built on the backbone of prodigious daily intake by a small but dedicated community of intensive users (Light et al. 2014). Legalization’s origins in both a strong marijuana culture and a broader culture valorizing local consumption practices make Colorado an ideal site in which to study the legalized industry. The location of legal marijuana in Colorado, Washington, Oregon, and Alaska suggests legalization is endogenous to local consumption cultures (Heying 2010; Jurjevich and Schrock 2012). Yet Colorado stands out as a study site for several crucial reasons.
First, Colorado’s status as the first state to legalize retail marijuana sales gives it the longest experience with legal marijuana, and a first-mover advantage that further magnifies the industry growth and diversification processes at issue. 1 In terms of research strategy, these conditions position Colorado as a critical case—a unique intersection of the circumstances crucial to understanding both the industry and its implications for theories of consumption-driven development (Yin 2008). Data help: The Colorado Department of Revenue’s Marijuana Enforcement Division provides monthly counts of establishment licenses and sales by municipality, as well as total occupational counts for individuals licensed to work in marijuana retail. These data supply reliable measures of a new consumption industry for which no North American Industrial Classification System code exists.
Second, the public, exuberant nature of legalization facilitates secondary data collection. Colorado’s marijuana producers and service firms loudly call attention to themselves—an important trait, given the complete absence of marijuana sale and production in federal economic data sets. Within three years of legalization, Colorado was home to an endowed cannabis chair and marijuana law curriculum at the University of Denver law school; the marijuana real estate listing service 420MLS; a daily newspaper, The Cannabist; and dozens of industry-specific business directories, trade publications, and consortia. This makes the growth and diversification of the industry easy to track.
Third, the distinguishing characteristics of marijuana regulation in Colorado accentuate the measurable aspects of consumption-driven economic development. From January through October 2014, Colorado retailers were subject to a vertical-integration rule requiring 70 percent of marijuana sold to be self-supplied. This intensified competition around proprietary cannabis strains, and limited the vulnerability of individual retailers to cheaper, standardized wholesale products. More important, Colorado law stands out for relegating licensing standards and approvals to municipalities. The majority of the state’s municipalities, including Colorado Springs and several large Denver suburbs, have banned retail licenses outright. In April 2016, Denver (population 700,000) passed legislation freezing the total number of marijuana retail licenses at current levels, and mandating an eventual reduction of 15 licenses for grow houses (Murray 2016). State law also limits license holding to individuals with two or more years of residency. Cumulatively, these factors push the market toward quality-based, rather than cost-based, competition, by restricting the total number of sellers, slowing out-of-state ownership, facilitating the early proliferation of retailer-specific cannabis strains, and making license approval dependent on site-specific relationships with City Councils and Mayors.
Following Yin’s prescription for rigorously executed case studies (2008), I treat each of the four developmental pathways identified by Markusen and Schrock (Table 1) as research propositions—falsifiable, directional hypotheses about legalization’s probable effects. My assessment of these propositions focuses simultaneously on measurement and process: To what extent has marijuana legalization yielded the traditional economic development outcomes of job growth, industry growth, and diversification? And which place-specific policy decisions, institutions, practices, and demographic characteristics shape those outcomes?
My answers to these questions draw on a range of secondary sources, including marijuana retail license figures, Secretary of State filings, content analysis of business listings, and newspaper articles. Although these data sources improve on many of the individual examples previously used to develop hypotheses about consumption-driven development, they are necessarily limited by the absence of marijuana industry codes in federal data sets. I use two primary strategies to extract as much reliable information as possible from these limited sources. First, I draw heavily on Colorado Department of Revenue, Marijuana Enforcement Division data measuring sales, revenues, and occupational licenses. Second, I use the unique characteristics of Colorado’s sudden, highly visible, and high-dollar-value legalization to interpret limited and partial information. The resulting analysis makes maximal use of improved, but still limited, data.
Proposition One: To What Extent Did Legalization Steer Consumer Spending toward Goods with a Higher Local Production Content?
In Colorado and elsewhere, marijuana laws require all retail cannabis sales to consist of plants seeded and grown in the state. Thus, calculating the net gain in local production rests on (a) estimating the geographic origin of marijuana sold prior to legalization, (b) estimating likely demand changes after legalization, (c) discounting existing medical marijuana supply, and (d) accounting for the individual-level production authorized by state law.
Marijuana Demand before and after Legalization
Prior to outright legalization in 2014, approximately 17 to 18 percent of Colorado’s adult population consumed marijuana annually, a higher rate than in 45 other states. The estimated 13 percent of Coloradans who consumed marijuana monthly ranked the state second nationally (Hedden et al. 2015), and points to demand’s roots in the one-quarter of users (23 percent) who consume daily (Light et al. 2014). Accounting for frequency and types of use, the Colorado Department of Revenue estimates the size of the marijuana market at 130 metric tons annually (ibid.).
These estimates assume static demand unaffected by legalization. However, removing the legal stigma attached to marijuana use will likely persuade new segments of the population to consume the drug for the first time, or more frequently. Additionally, to the extent that retail marijuana increases Colorado’s appeal to tourists and plays a role in the state’s rapid post-2012 population growth, legalization directly expands the base of possible consumers. As a result, the total volume of Marijuana sales in Colorado reached 149 million metric tons in September 2015. This establishes the Colorado Department of Revenue’s pre-legalization estimates as a de facto demand floor, and calls attention to the likelihood of continued growth in aggregate demand.
Sources of Supply before and after Legalization
Prior to 2012, Colorado was a net importer of marijuana from climates better suited to the plant’s growth. However, the legalization of medical marijuana production in 2000 created a significant amount of licensed in-state production, totaling 60.4 tons on the eve of legalization (Table 2). Additionally, an estimated 3.9 percent of users met their consumption needs from homegrown supply.
Estimated Growth in Annual In-State Marijuana Production after Legalization.
Source: Kilmer et al. 2013.
Source: Light et al. 2014.
Source: Gettman 2006.
A small portion of the remaining demand for illegal marijuana was met by organized, illegal production within Colorado. Gettman’s (2006) estimates of marijuana production by state indicate that Colorado accounted for 0.03 percent of the nation’s illegal marijuana growth, against 1.75 percent of total consumption, meaning that 98 percent of illegal demand was met through imports. Accounting for medical marijuana supply, less than half of Colorado’s demand prior to legalization was met through in-state production.
By legal mandate, consumer demand for retail marijuana in Colorado is met through in-state production. Accordingly, the expansion of total consumer demand to an annual rate of 149.1 tons in September 2015 resulted in 81.5 tons of new in-state production. This equates to $725 million in estimated new sales, representing a 0.3 percent increase in state GDP. The size of this increase is equivalent to the combined size of Colorado’s primary metals and electronic equipment industries (US Bureau of Economic Analysis 2013). While the absence of input–output data on the industry make the calculation of an economic impact multiplier impossible, the mandate to produce in-state ensures nearly zero leakage on backwards demand for the product.
To the extent that it simply reflects the legalization of existing demand for marijuana, the growth of marijuana retail sales does not come at the expense of reduced consumption of other consumer goods. In fact, early studies of consumer spending in Colorado suggest that the net out-of-pocket amount spent on marijuana decreased between 2014 and 2015, because of the high intoxication content and comparatively low price points of legal marijuana (Light et al. 2014). Among new consumers and tourists in particular, consumer-spending surveys suggest that marijuana consumption leads to at least a partial reduction in spending on alcohol, as well as decreased spending on pharmaceuticals. These substitutions will partially mitigate net growth in Colorado economic output, but with limited effect: The mandate for in-state marijuana production ensures that sales in the industry will have a larger economic impact than any other consumption option, including brewing, natural foods, and other goods for which large portions of demand are met by in-state supply.
Because illegal marijuana production skews heavily toward five US states, the federal requirement that all legal marijuana be produced in the states where it is sold ensures massive import substitution and the development of what Markusen (1996) terms a “sticky” industry unable to make or exercise threats of capital flight. This requirement will increase the economic development benefits of legalization in the forty-two US states whose illegal consumption of marijuana relies on imports. However, states with significant illegal production, including California and Hawaii, will see smaller net benefits to legalization than Colorado, as retail marijuana sales will reassign production from illegal to legal status, rather than move it across state lines. For those states, the potential to reproduce other mechanisms of legalization-induced economic development will be more important.
Uneven Local Supply and Demand
While legalization represents significant net business growth to Colorado, the distribution of that growth is highly uneven. State law devolves the decision of whether to legalize sales, and how many retail licenses to grant, to municipalities. Thus, actual supply and demand for retail marijuana follow cultural and political lines. While Denver and the college towns Boulder and Fort Collins have enthusiastically embraced legal sales, politically conservative Colorado Springs and many inner-ring Denver suburbs ban them. Given that possession and consumption remain legal statewide, and that cities banning retail sales count marijuana consumers among their residents, this leads to effective exporting, and highly uneven spatial development for marijuana production, sales, and tax revenue, relative to other retail goods (Table 3).
Retail Marijuana Tax Revenue, 10 Largest Colorado Counties.
Source: Colorado Department of Revenue, Marijuana Enforcement Division 2016b; US Census Bureau Intercensal Population Estimates.
Even among cities with legal sales, the total amount of retail activity and tax revenue diverges substantially. Denver boasts both the state’s highest population and, at $11, the highest per-capita tax revenues. Here, the core city appears to absorb substantial consumer demand from suburban counties whose development historically strained the city’s tax base. While the statewide 10 percent retail sales tax spreads fiscal benefits statewide, this unevenness indicates substantial potential for intrastate import substitution, and the effective ability of municipalities with legal marijuana sales to draw business, jobs, sales, and tax revenues from the consumption of neighboring cities’ residents.
Proposition Two: Does Marijuana Legalization Lead to the Growth of Comparatively Labor-Intensive Businesses?
Legalizing marijuana changes the capital and labor requirements of both the production and sale of marijuana, but in opposite directions. Compared to illegal marijuana, Colorado’s indoor grow facilities are substantially more capital intensive. However, small establishment sizes and elevated security needs for marijuana retail make the sale of cannabis extremely labor-intensive. Accordingly, the industry’s contributions to job growth develop along two separate lines.
Capital-Intensive Grow Facilities and Two-Tier Labor Markets
Illegal marijuana growers face substantial barriers to scaling up their operations and benefiting from economies of scale. These include the heightened risk of detection in larger facilities, lack of access to legal capital, the lack of access to high-volume distribution channels, and the lack of access to a labor market that facilitates the development, acquisition, and use of cannabis-specific horticultural skills (Hammersvik, Sandberg, and Pedersen 2012). By removing these barriers, legalization allows cannabis producers to rationalize and expand their productions. Colorado’s cold winters further mandate capital-intensive production, by moving cannabis growth to indoor facilities in which industrial design and electricity do the sun’s work.
The resulting need for large industrial facilities, and the costs associated with hydroponic assembly, maintenance, electricity, and humidity control, ensure high levels of capital input into the production process (IBISWorld 2015b). Yet the process also requires higher amounts of labor input per dollar of sales than do automobiles, aerospace, and many other industries coveted by economic development authorities (ibid.). Colorado facilities typically employ a production supervisor with multiple assistant to monitor humidity levels, temperatures, light levels and plant quality (IBISWorld 2015a). In addition to tending plants as they grow, modestly skilled production and harvest workers are responsible for the specialized task of trimming and shaping the unruly plant into aesthetically pleasing buds for sale. For producers of marijuana extracts, wax, and vapor cartridges, these common jobs are complemented by a smaller number of positions for technicians (and supervisors) trained in processes for converting marijuana into consumable chemical compounds and ensuring consistent quality, which is vital to competitive strategies based on the branding of proprietary strains (IBISWorld 2015b). These production characteristics indicate that while legal marijuana production is less labor intensive than illegal production, it is nevertheless more labor-intensive than the typical existing industry.
Labor-Intensive Retailers
The distinguishing characteristics of retail marijuana laws limit the capital-intensiveness of retail facilities in comparison to both other retailers and the economy as a whole. Investor reports on marijuana retail indicate the use of 0.04 units of capital per every unit of labor involved in production—less than half the retail-wide rate (0.1) and substantially less than the economy-wide rate of 0.15 (IBISWorld 2015a).
Several factors combine to make sales uniquely labor intensive. First, high licensure costs (which can reach as much as $400,000 per retail facility) and the absence of the time necessary for industry consolidation generate a retail environment characterized by large numbers of small establishments. In turn, this generates high labor needs: While retailing writ large has moved to capital-intensive warehousing and big-box formats, marijuana sales establishments remain small and poorly suited to substituting machinery—forklifts, for example—for human labor. Similarly, the low volume and high dollar value of marijuana sales limits the large sales-unit volumes needed to realize unit-cost savings in delivery and inventory (IBISWorld 2015a).
Second, security concerns require marijuana retailers to dedicate staff, and portions of otherwise routine jobs, to security. Like diamonds, marijuana’s high price per unit of weight makes it a target for theft. Furthermore, Colorado law classifies the sale of marijuana to minors as a felony, leading retailers to employ full-time security guards, use restrictive inventory systems, and subject all sales to a high level of scrutiny from employees (IBISWorld 2015a). Third, competitive strategies emphasize product differentiation, which is brokered by large and attentive retail staffs. Overall, retail sales entail an unusually high degree of customer interaction, which adds to the industry’s labor demand and places a premium on customer relations skills (IBISWorld 2015a). This organization of labor appears to be unique to Colorado, whose relatively restrictive retail licensing and mandated vertical integration during the first year of legalization have created a smaller number of per capita retails, and a higher emphasis on product quality as opposed to cost, than other states with legal sales (Vermont Legislature 2016).
These shifts, toward capital intensiveness in marijuana cultivation, and labor-intensiveness in its sale, result in occupational bifurcation. To date, the Colorado Department of Revenue has granted more than four times as many occupational licenses for routine, “support” occupations as it has for “key,” skilled occupations (Table 4):
Major Marijuana Occupations and Pay.
Source: Colorado Department of Revenue, Marijuana Enforcement Division 2016c; Job postings onCannajobs.com, MJBizExchange, Indeed.com, 420jobs and Cannajobs, September-October 2015.
The majority of these jobs require some measure of industry-specific skills, an attribute that may help to explain retail marijuana’s wage floor of $11, significantly higher than the Colorado minimum wage of 8.23. Elevated pay levels also likely reflect the industry’s current boom status. With sustained growth in overall demand, the industry has been able to increase monthly sales per establishment, from $27,050 immediately after legalization to $40,000 in October 2015, even as the state added more than seven hundred licensed retailers (Colorado Department of Revenue, Marijuana Enforcement Division 2016b). Given unceasing market growth and the expansion of consumer demand, industries in this phase of the life cycle pay wages above the regional norm, in order to capitalize on expanding product demand (Markusen 1985). As the industry matures, the ability to maintain these high wage rates will rest on the attainment of product monopolies that sustain high employer margins.
These results support the argument that shifting consumption to labor-intensive goods can generate net job growth, at least over the short term. While the prior organization and compensation of marijuana-related labor under prohibition remains unknown, legalization has created a number of jobs paying benefits and hourly wages above the legal minimum. Furthermore, legalization has formalized a labor market for specific skills and experience. While the capital-intensiveness of indoor marijuana growth appears to induce a net decrease in labor content, the growth displaced by legalization occurred almost exclusively outside Colorado’s boundaries. Under national legalization, this substitution of capital for labor may yield a net disemployment effect. Until such a nationwide system is in place, states legalizing marijuana production will immediately see net growth in aboveground production jobs, and in retail jobs requiring extensive customer interaction skills.
Proposition 3: Local Consumer Spending Nurtures the Development of Future Exporting Industries
In addition to generating employment growth by shifting spending toward locally sourced and labor-intensive goods, changes in consumption carry the potential to shift the long-term development of industrial and human capacity within a region. As developmental economics and histories of US manufacturing demonstrate, key exporting industries emerge from local product demand that provides innovators and entrepreneurs with an avid market in which to test and nurture their products (Markusen and Schrock 2009). For marijuana, legalization provides inventors with access to venture capital, occupational skill specialization, and other foundational components of business development
Colorado’s diversified production of marijuana products now includes edible goods ranging from muffins to soups to infused teas; alcohol infusions; resins, waxes, “shatter,” and other intoxicant by-products culled from previously underutilized portions of the cannabis plant; a variety of “vapes” and other smokeless delivery systems; and miscellaneous products ranging from gum to lip balm. Along with these commercial products, Colorado firms have developed a host of specialized producer services, including industrial and retail real estate listings, cannabis accounting services, testing facilities and consultancies, website production, and inventory control. The export of marijuana across state lines remains illegal, but Colorado firms have begun to export these advanced services to states, cities, and tribal reservations in which marijuana has been selectively or completely decriminalized.
To establish the extent of product and service diversification, I developed company profiles for the 150 firms licensed for marijuana manufacture by the Colorado Department of Revenue. I first consulted articles of incorporation and certificates of good standing from Colorado Secretary of State filings to identify the 82 active manufacturers maintaining regulatory compliance and fixed business addresses (Colorado Secretary of State 2016). From these, I used Internet searches to develop an inventory of products. Working from a random sample of 15 manufacturers, I coded products into basic categories, which were later refined and validated via qualitative content analysis (Boettger and Palmer 2010; Tillery and Chresfield 2012). In addition to supplying cultivated marijuana for smoking—measured by pounds of “flower,” in the industry convention—more than 90 percent of Colorado’s licensed marijuana producers manufactured and sold edibles, tinctures, waxes, balms, hash, or other cannabis derivatives (Table 5).
Principal Varieties of Manufactured Marijuana Products.
Source: Content analysis of data from Colorado Department of Revenue, Marijuana Enforcement Division 2016c; Colorado Secretary of State 2016; and commercial websites.
Colorado’s marijuana manufacturer census also helps to identify the role that prior medical marijuana authorization played in establishing leading industry firms (Colorado Department of Revenue, Marijuana Enforcement Division 2015). Of the eighty-two active marijuana producers in Colorado in November 2015, twenty-eight started as medical dispensaries prior to the passage of Amendment 64. Company histories of these firms indicate gradual diversification of their product offerings, from simple flower, to edibles, balms, and vapor products. This prior experience, combined with the presence of mature flower and proprietary strains prior to legalization in January 2013, positioned medical marijuana facilities with a distinct first-mover advantage in the market. Today, medical marijuana sales also play an important supporting role for retail vendors, twenty-six of which have added medical sales—characterized by a lower tax rate and steadier demand than retail sales—to their operations.
These firms are owned disproportionately by Colorado investors, despite the leading status of California’s medical marijuana industry as a site of strain, extract, and edibles development because of the state’s early adoption of medical marijuana laws in the 1990s (Stateman 2009). Of the sixty-four active marijuana manufacturers from which ownership information was available, sixty started or operated as Colorado companies, indicating that prior knowledge and business incorporation developed in California did not displace local ownership for Colorado firms.
The diversification of Colorado marijuana production extends beyond manufacture itself. Since legalization, Denver has gained The Cannabist, a daily, online marijuana industry newspaper; a marijuana law program and endowed professorship at the University of Denver law school; and a $9-million annual state budget line to fund academic research on marijuana’s cultivation and social impacts. Denver is home to the national legislative advocacy bodies the Marijuana Industry Group, The Cannabis Trade Council, and the National Cannabis Industry Association, as well as Vicente Sederberg, LLC, a nationally active law firm specializing in marijuana law. This growing web of organizations is complemented by 420MLS, a marijuana real estate service, and a proliferation of businesses specialized in marijuana consulting, finance, facility design, and real estate (Table 6).
Marijuana Industry Producer Services.
Source: Keyword searches for terms 420, high, Cannabis, green.
Table 6 provides an overview of commonly available marijuana industry services. While many of these services are oriented toward consumption within Colorado, Colorado has also become an exporter of advanced industry services. A growing number of medical marijuana dispensaries and retailers located in states with partial or full legalization draws on hydroponic, financial and market consulting from Colorado entrepreneurs. Examples include a South Dakota Sioux tribe building a legal marijuana resort facility, which has contracted with Colorado firms to design both the resort itself and the production of the cannabis to be consumed by visitors (Cano 2015; Osborn 2015). In Illinois, the establishment of the state’s first medical marijuana facility in 2015 was so heavily dominated by one Denver entrepreneur that Illinois merchants filed a lawsuit against him (Associated Press 2015). While a full accounting of Colorado marijuana services exports requires data and transaction records unavailable to the public, these developments, as well as Colorado companies’ prominent place in trade industry lists and the growing use of Colorado firms to design marijuana facilities in Oregon, Hawaii, and California, indicate a degree of service exporting.
Given that industries cluster in space and inevitably form headquarters regions (Markusen 1985), the expanding national cannabis industry will have an advanced services headquarters somewhere. Colorado’s advantage in developing these functions is too tentative to be certain, and may be destabilized by legalization in California or other states with larger consumer markets. Regardless, states that develop marijuana services will develop some capacity to export those services, in addition to benefiting from the advantages of mandated in-state production.
Proposition 4: Superior Local Consumption Alternatives Lead to In-Migration of Skilled Workers
The most uncertain pathway from consumption changes to economic development may also be the most important. While the distributive merits of amenities-focused urban policies rightly remain contested, the ability of superior local amenities and consumption options to draw highly mobile in-migrants is clear (Jurjevich and Schrock 2012). In addition to contributing to local industries, these migrants often go on to start their own firms (Markusen 2004). To the extent that it marks a place as unique, and draws migrants, legal marijuana can contribute to this accrual of human capital (Clarke and Gaile 1998).
The available evidence on in-migration to Colorado is favorable to marijuana legalization. Because of the long periods needed to measure migration, it is also extremely limited. In the first two years following the passage of Amendment 64 in 2012, Colorado ranked among the top three in the United States in population growth, a standing driven by high rates of in-migration (Svaldi 2015). Limited supporting evidence, such as a 33 percent spike in out-of-state applications to the University of Colorado (Kuta 2014) and news accounts of migrants lured by legalization and Medicaid expansion under the Affordable Care Act (Milkman 2015; Rothenberg 2015), suggests that recent Colorado policy changes appeal to footloose migrants.
This evidence is far too speculative to substantiate any kind of claim that legalization has made Colorado a migration destination. Perhaps more important, the endogeneity of legalization to Colorado’s long-term appeal to professionals looms as an interpretive hurdle, regardless of the volume or reliability of migration data on offer. Legal cannabis both builds on and originates from the lifestyle factors that have long made Colorado a migration destination. To an extent uncommon among the other regions lumped into the conceptual ambiguity of a Southern and Western sunbelt, economic boosterism in Colorado’s Front Range, and the Denver metro area in particular (1940 population: 322,000) always emphasized climate and lifestyle over comparatively commonplace appeals to a good business climate (Busch 2015; Judd 1983). After diversifying its once energy-dependent economy during the oil price shocks of the 1980s, the region began to draw large shares of college graduates, at the ninth highest rate of any US metro region during the 1990s (Franklin 2003).
For the past decade, Colorado has fallen annually in the top five states for population growth, due in significant part to so-called quality-of-life draws (sunshine, mild winters and summers, limited traffic congestion compared to similar metropolitan areas) and outdoor amenities (Whisler et al. 2008). Rather than driving or initializing the state’s appeal to skilled migrants, legalization’s value may instead lie in its ability to confirm the state’s allure to college graduates seeking alternatives to conventional career ladders (Schrock and Jurjevich 2012), and to differentiate Colorado from competitors that advertise natural amenities.
For policymakers and academics alike, the chase for mobile, scarce human capital occupies an area of growing concern. It represents the increasingly unquestioned foundation for policies endorsing arts-based development, urban amenities, and spending on the favored class of individuals possessing advanced skills and the potential ability to innovate (Markusen 2004; Clark 2013). The limited evidence of Amendment 64’s impact on migration is positive, and if these results can be sustained and reproduced, the ability of marijuana legalization to draw in-migrants to later adapters and to places that have otherwise struggled to draw migrants will become a crucial question.
Conclusion: Development by Fiat
As a large-scale experiment, marijuana legalization substantiates, with volume and at scale, the broader applicability of developmental pathways whose assessment has to date been limited to individual sites, historical accounts of industry development, and ex-post assessments of local economic change. Colorado’s billion-dollar sales volumes, proliferating industry specializations, and growing demand for skilled and unskilled labor represent economic gains that far outstrip those usually hoped for or claimed in economic development. Significant in their own right, these policy-specific results also shed light on both the future economic potential of marijuana legalization, and on the types of consumption-driven development most likely to yield long-term economic benefits.
As the first state to witness recreational marijuana sales, Colorado’s experience with legalization—overwhelmingly positive in economic terms—carries disproportionate weight in efforts to evaluate marijuana’s economic development potential. Thus, it is crucial to note that the clearest economic benefits emerge from the mechanics of legalization itself. The largest short-term gains to legalization, and the most easily measured gains regardless of magnitude, emerge from the federal requirement to grow and sell marijuana within the same state. This amounts to a massive import substitution mandate, one likely to have similarly large effects in other states that are currently net importers of illegal marijuana. Beyond this basic, significant change, the composition and extent of industry growth after legalization result from the specific details of Colorado law, and not the broader mandate to liberalize. Both the labor-intensiveness of marijuana retail and the distinctive quality-based character of retail competition result from legislative measures that limit license availability, require vertical integration, and mandate heavy in-store security. Place-contingent, these regulatory decisions have not been emulated elsewhere: Washington and Oregon both feature comparatively permissive licensing, and have incentivized mass production, to the point that cheap, interchangeable strains of marijuana dominate the market in each state. In distinction to Colorado’s quality-based competition, this indicates the emergence of a cost-competitive industry likely to be less amenable to high-wage jobs and continued product differentiation.
In contrast to these legislation-specific benefits, Colorado’s gains in industry diversification and producer services emerge primarily from historical accident and place luck, and the state’s first-mover status. Legalization is not a randomly occurring policy event. In Colorado, it was pushed by marijuana activists who were themselves drawn to the state by its natural amenities, receptiveness to migrants, and affection for consuming then-illegal cannabis. The short, six-month window in which Colorado was the only US state with legal retail sales sped the development of advanced and producer cannabis services, giving Colorado firms an advantage in capturing expanding markets in other states. This advantage is already apparent in the significant contracts Colorado firms have received to design, implement and oversee medical and legal marijuana delivery in states and legal spaces with liberalized laws. Thus, legalization in other states is likely to provide ongoing benefit to Colorado firms embedded in Denver’s emergent agglomeration of marijuana services firms, testing facilities, and equipment manufacturers. Similarly, the reputation effect of early legalization—its affective value as a signal of social liberalism and place distinctiveness—will likely continue to benefit Colorado, which remains a focal point for the industry’s growth despite full legalization expanding to one-sixth of the states.
Beyond the immediate question of legalization, Colorado’s experience also helps to identify the broader applicability of the consumption-driven development processes. Novel when first published, Markusen and Schrock’s arguments linking consumption shifts to conventional economic development goals appear to understate the complementarity between conventional and consumption-based approaches to economic growth and diversification. Colorado’s evolution since legalization mirrors traditional economic development in terms of process as well as outcomes. The regulatory, competitive, and market-fragmenting factors that shape job growth in the state’s legal marijuana industry are the same factors economic development analysts consider when evaluating manufacturing plant location, industrial land use, and inevitably, tax incentives. While the potential for legally driven consumption gains appears to be limited to marijuana law’s unique in-state production mandate, the broader applicability of these mechanisms is nonetheless significant: subjecting consumption industries to the same analytical rigor typically reserved for exporting industries can open new pathways for policy intervention, regional competitive advantage, and the development of high-wage jobs (Doussard 2013).
Yet the more substantial gains to consumption-driven development appear, appropriately enough, harder to attain. While every state can legalize marijuana production and sale, the expanding marijuana industry can functionally have only one headquarters region. In Denver, the industry’s growth responds to the development of a state-level market that has already generated nationally active industry institutions, a diversification of services, and mounting export activity. These gains will grow as Denver-area firms continue to export design, horticulture, testing, marketing, and legal services to the growing body of states planning or experimenting with legalization. To the extent that legalization or other shifts in consumption draw mobile human capital, or renew Colorado’s desirability as a destination for lifestyle migrants, the gains may be even greater. Whether executed through putative “creative class” approaches, support for the arts, or the targeting of skilled occupations by economic development planners, the competition for human capital and the development of distinctive, amenities-rich urban environments have become staples of economic development policy. The only agreement about this competition is that it is extremely difficult to win. Marijuana legalization, and the resulting development of specialized economies, appear to be more a result of such success than its source: like prior consumption-driven industries (natural foods, brewing, outdoor apparel), legal marijuana emerged from Mountain West and pacific coast states whose climates and cultural mores already made them winners in the contest for mobile capital. The ongoing importance of this basic, interregional contest for mobile human and investment capital suggests that consumption-driven development represents a variation, rather than a wholesale innovation, on the typical, desultory economy fight between regions. The principal difference, however, indicates a rare domain of policy leverage: unlike climate and natural amenities, municipalities retain the power to shape their own consumption industries.
Footnotes
Acknowledgements
I wish to thank Phil Ashton, Laura Wolf-Powers and three anonymous reviewers for helpful comments that improved this article. Extra thanks go to Dena Raposa for her excellent research assistance.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.