Opportunity zones: do tax benefits go to the most distressed communities?

Abstract

Purpose

The exact criteria used by state governors for choosing opportunity zones (OZs) are not publicly available. This paper aims to examine whether state governors selected the most distressed communities, or those with the highest proportions of minorities, as OZs.

Design/methodology/approach

This paper compares the distressed communities chosen as OZs in states throughout the country to an equal number of those eligible distressed communities but not selected. Moreover, this paper uses regression analysis to determine whether the poverty rate, median family income, population, percentage of population that is minority and the percentage of population that is African American are significant explanatory factors in the choice of OZs.

Findings

After describing the tax incentives for investing in OZs, this paper documents that governors did not select many of the most distressed communities, or those with high proportions of minorities, in their individual states.

Originality/value

This paper describes in some detail the way in which investors may generate tax benefits by investing in eligible property or businesses in OZs. It also examines the extent to which the degree of poverty and the percentage of the population that is minority (and African American) were key factors in the selection of OZs. It arises an issue that the chosen communities are not necessarily those most in need of more investment or those heavily populated by minorities, particularly African Americans.

Citation

Barth, J.R., Sun, Y. and Zhang, S. (2021), "Opportunity zones: do tax benefits go to the most distressed communities?", Journal of Financial Economic Policy, Vol. 13 No. 3, pp. 301-316. https://doi.org/10.1108/JFEP-06-2020-0125

Introduction

One of the more recent efforts by the federal government to help economically distressed communities throughout the country is the creation of “opportunity zones.” Localities chosen by state governors as opportunity zones (OZs) may achieve faster economic growth and greater gains in employment than otherwise through potentially new investment generated by tax benefits. More specifically, the Tax Cuts and Jobs Act (TCJA) that became law on December 22, 2017, added two new sections to the Internal Revenue Code (IRC). One section provided for the designation of OZs on December 22, 2017. The first chosen set of OZs, among all communities originally eligible as OZs, occurred on April 9, 2018, and covers parts of 18 states. The final list of chosen OZs covers parts of all 50 states, the District of Columbia and five US territories that occurred on July 9, 2018.

The other section of the IRC provides tax benefits to investors who reinvest capital gains in an opportunity fund (OF) within 180 days of a sale in an OZ for a minimum of five years. The tax benefits include the permanent exclusion of up to 15% of the capital gains from inclusion in gross income for federal tax purposes. The remaining 85% of the gain is not included in gross income until December 31, 2026. In addition, the IRC excludes from gross income the post-acquisition gains on investments in OFs held for at least ten years. These tax benefits, deferrals and exemptions provide strong incentives to encourage investment in OZs, thereby potentially spurring new economic development and creating more jobs in distressed communities.

The incentives can be quite substantial. For example, according to the Fundrise Opportunity Fund, if an investor has a gain of US$25,000 to invest in an OF and holds the investment for ten years, the potential returns can exceed the returns on a traditional stock portfolio with an expected annual return of 6% by US$9,184 [1]. Of course, the additional economic development and jobs created because of any new investment in OZs cannot be determined for many years because of the newness of the program. Importantly, however, one can still assess the extent to which the chosen distressed communities providing the investment incentives are among the most distressed communities in each state. In addition, one can assess whether the chosen distressed communities are the most heavily populated by minorities, and particularly African Americans, as there is a high correlation between such communities and the proportion of the population that is minority.

Clearly, this is not the first attempt to use federal tax incentives to stimulate investment in economically distressed communities. The Community Renewal Tax Relief Act (TCJA) on December 21, 2000, provided a tax credit for investment in “low-income communities,” which are census tracts that have a poverty rate of at least 20% or have a median family income that generally does not exceed 80% of the greater of the statewide average or the metropolitan area average [2]. The TCJA also focuses on low-income communities, or census tracts, including some adjacent census tracts. However, the number of chosen OZs is limited to approximately 25% of the total number of originally eligible OZs in each state, US possession and the District of Columbia. Among all census tracts in 50 states, the District of Columbia and five major US territories, 42,160 were originally eligible for designation as OZs. Eventually, state governors chose 8,764, or 21%, of the eligible census tracts as OZs [3].

One contribution of this paper is to describe in some detail the way in which investors may generate tax benefits by investing in eligible property or businesses in OZs. Another, and far more important contribution, is to examine the extent to which the degree of poverty and the percentage of the population that is minority (and African American) were key factors in the selection of OZs. The examination involves comparing the distressed communities chosen as OZs in states throughout the country to an equal number of those eligible distressed communities but not selected. As we document, the chosen communities are not necessarily those most in need of more investment or those heavily populated by minorities, particularly African Americans. Clearly, the issue that arises is the extent to which the extent of poverty and degree of minority population, which are highly inter-related, were decisive factors in choosing OZs, given the intent of the TCJA.

Unfortunately, the various factors relied upon in choosing “distressed” communities as OZs and the weights attached to them are not publicly available. Nevertheless, we can empirically examine the relationships between both the extent of poverty and the degree of minority population in eligible distressed communities and the likelihood that an eligible distressed community is chosen as an OZ. To our knowledge, despite its importance, our analysis is the first to examine whether state governors selected the most distressed communities, or those with the highest proportions of minorities, as OZs. After considerable time elapses, one can then assess whether the economic gap between the chosen OZs as compared to those eligible but not chosen has widened or not, thereby enabling one to better assess the benefits and costs of the TCJA.

Criteria for choosing opportunity zones

OZs are an economic development tool insofar, as they may spur economic growth and job creation in distressed communities. The process of selecting OZs started with a list of all low-income, or distressed, communities in every state [4]. The census tracts on this list became eligible OZs within each state. Governors in each state could then nominate, in writing, up to 25% of these census tracts as OZs. No later than March 21, 2018, with a possible 30-day extension, governors made nominations. The Secretary of the Treasury could then certify the nominations and designate the nominated tracts as OZs. In addition, up to 5% of the designated tracts in a state could be non-qualifying tracts, but they had to be contiguous with qualifying tracts and have median incomes less than 125% of the contiguous qualifying tracts. The resulting OZs hold that designation for ten years.

There is no detailed and publicly available information from either the federal or state governments describing all the factors used in choosing the low-income communities nominated as OZs from all eligible OZs [5]. However, the Economic Innovation Group (2018) conducted a survey to determine the factors used by states in the nomination process. Based on responses from 41 state officials, “[…] every state augmented the baseline eligibility criteria with additional data analytics customized around their own priorities.” In addition, some states convened an external advisory panel of citizens or public leaders to vet final recommendations, while others consulted with national experts, advisors and peers. Interestingly, the survey revealed that 29 states provided a website describing information about the governor’s selection process, while others solicited public input. Only a few states, moreover, released draft recommendations to the public before making submissions of their nominations to the Treasury Department [6].

The result of this process was the final designation in June 2018 of 8,764 OZs, or 21% of all eligible OZs and 12% of all census tracts in the USA and five US territories. Of all the OZs, 8,566 are low-income communities and 198 are contiguous with low-income communities and chosen as OZs. Moreover, 7,826 OZs are located in the USA, or 89%, while 863 are located in Puerto Rico [7] and 75 in Guam, American Samoa, Northern Mariana Islands and Virgin Islands, accounting for the remaining 11%. Based on the population and the number of census tracts in just the USA, the average number of persons in a census tract is 4,128. This means there are roughly 32 million people in the OZs, or approximately10% of the total population.

The number of OZs within the 50 states and District of Columbia ranges from a high of 879 in California to a low of 25 in ten states. The latter figure is the maximum number of designated OZs because each of these states had fewer than 100 low-income communities. The top five states account for 2,775 OZs, or 35% of the total. At the same time, the same five states account for 119 million people, or 37% of the total population. More generally, the rank-order correlation between the number of OZs in the states and population of the states is 0.99. This indicates that the distribution of OZs tends to reflect the distribution of the population across the states.

Tax benefits for opportunity zones

There are three main federal tax incentives to encourage investors to take unrealized capital gains and redeploy those gains into OZs [8]. Given that there is approximately US$2.3tn in untapped capital gains held by investors, the potential certainly exists for such redeployment to contribute importantly to economic development and job creation in distressed communities (DeBolt, 2018). One incentive is that taxpayers can defer the tax on the capital gains due upon the sale of an asset with an unrelated person if all or a portion of the gains are reinvested within 180 days in OFs [9]. A second incentive is that if taxpayers hold investments in OFs for at least five years, the basis on the original gains increases by 10% of the gains on the original assets, which means taxpayers only owe taxes on 90% of the rolled-over capital gains. If the investments are held for at least seven years, the basis on the original gains is increased by an additional 5% of the original gains [10]. To receive these benefits, investors must invest in OFs by 2021 to receive the 10% increase, and by 2019, to receive the additional 5% increase. The third incentive is that the basis of investments maintained for at least ten years, and until at least December 31, 2026, are eligible to be marked up to the fair market value of such investments on the date the investments are sold. This effectively means the investments are not subject to a capital gains tax on any gains earned from the investments in the OFs over ten years when the investments are sold. These tax incentives are in effect from December 22, 2017, through December 31, 2026. There is no deferral or gain available for any sale or any exclusion available for any investments after December 31, 2026. By this date, moreover, the federal government receives tax revenue when investors are required to pay the reduced and deferred capital gains taxes from the original investments in 5–7 years. Until then, investors effectively receive an interest-free loan from the federal government [11].

Opportunity fund investments [12]

Generally, for taxpayers to qualify for the tax benefits, the capital gains must be invested in an OF, which must be an entity treated as a partnership or corporation, whether domestic or foreign, for federal tax purposes and organized in any of the 50 states, District of Columbia or five US territories. An OF then invests all or part of the capital gains in OZ eligible property or businesses. The requirement is that at least 90% of the assets of an OF include eligible property or businesses, which includes stock in an OZ domestic corporation, partnership interest in an OZ domestic partnership and OZ business property. The remaining 10% can include property or businesses located outside of an OZ. The OFs themselves certify that their investments comply with the law as interpreted and implemented by the Internal Revenue Service and Treasury Department.

An OF can invest in the stock or obtain a partnership interest in a new or existing firm located in an OZ. All or substantially all (at least 70%) of the OZ tangible property of the firm must be owned or leased by the OF [13]. The firm must also derive at least 50% of its total gross income from its activities in the OZ. The firm, moreover, cannot be a golf course, country club, massage parlor, hot tub facility, suntan facility, racetrack or other gambling facility, or liquor store. In the case of OZ business property, it is generally tangible property that an OF purchases from an unrelated person in 2018 or later. The purchased property must be in a trade or business in an OZ, and substantially improved by the OF over a 30-month period (or its original use in the OZ commences with the OF). An OF must increase the basis of the acquired property by an amount that exceeds the initial basis to qualify as substantially improving the property, which effectively means more than doubling its basis. The OF, moreover, may borrow funds to purchase or improve the property. Lastly, the final regulations allow some business located outside an OZ to be included in businesses located within an OZ. For example, if a common development plan led by a single qualified OZ business includes two properties in an OZ worth US$7m and one property outside the OZ worth US$3m, the business is an OZ because 70% of its property is “in use” in the OZ (Opportunity Alabama, 2020).

Opportunity zones

In the USA, there are 42,160 eligible OZs. The Treasury Department in July 2018 completed approving all 8,764 nominated as OZs [14]. The chosen OZs represent 12% of all census tracts and 21% of the eligible OZs. Of the 8,764 OZs, 8,566 are low-income, or distressed, communities, while 198 are contiguous with low-income communities and have median incomes less than 125% of the contiguous communities. The selected tracts average between 1,200 and 8,000 in population and vary in size from 199–235,352 acres. The total population of the OZs is 35 million, or 10.6% of the total population.

Figure 1 shows the number of OZs in the states and territories. California has the most OZs at 879, while the Virgin Islands has the fewest at 14. Puerto Rico has the second largest number at 863, but as noted earlier, practically, the entire territory is an OZ.

Due to the limitation of the census demographic data [15], our analysis focuses on 50 US states, the District of Columbia and Puerto Rico [16]. Of the 8,689 OZs, the mean (median) poverty rate is 32 (31)%. For purposes of comparison, we examine the same number (8,689) with the highest poverty rates from all 33,396 of eligible but not chosen OZs. For this group, the mean (median) poverty rate is 38 (35)%. Both of these figures are slightly higher than for the chosen OZs. The maximum poverty rate for an OZ is 100%, which occurred in three OZs [17]. Seventeen other eligible census tracts not chosen also have poverty rates of 100%. This indicates that there were more eligible OZs with higher poverty rates than those chosen. The minimum poverty rate is zero, which occurred in two OZs, one in Wayne County, Michigan, and another in Bexar County, TX [18]. The population in the two OZs is 9 and 22, respectively.

The chosen 8,689 OZs and the same number (8,689) of eligible but not chosen OZs are now combined and ranked by the poverty rate from high to low. One can then rank the top 8,689 of the 17,378 census tracts in terms of poverty rates. Figure 2 shows that 3,737, or 43%, of all these eligible OZs were chosen as OZs, whereas 4,952, or 57%, of them with higher poverty rates were not chosen. Clearly, this indicates that factors other than only the poverty rate of a low-income community were involved in the selection process.

Instead of ranking chosen and not chosen OZs using the highest poverty rates, we rank them based on the highest percent of minority population. In terms of mean (median), the OZs have a value of 60 (66)%, while the eligible but not chosen OZs have a value of 89 (90)%. This indicates that there are substantially more eligible OZs with higher minority population rates than those chosen as OZs. The chosen OZs have a maximum percentage of minority population of 100% and a minimum percentage of 0%. In the case of the not chosen OZs, they also have a maximum percentage of minority population of 100%, but a minimum of 72%. In addition, of the 586 eligible OZs with minority population rates of 100%, 424 eligible OZs were chosen, while the remaining 162 were not chosen.

Figure 3 shows the percentages of both chosen OZs and eligible but not chosen OZs when 8,689 of 17,378 eligible OZs with the highest minority population rates are considered. The figure shows that 35% of the eligible OZs were chosen as OZs, while 65% of them were not chosen. This indicates that factors involved in the selection process included more than only the minority population rate.

Instead of ranking 8,689 chosen and not chosen OZs based on the highest minority population rates, we rank them on the highest African American population rates. In terms of mean (median), the OZs have a value of 24 (10)%, while the eligible but not chosen OZs have a value of 52 (45)%. This indicates that there are more eligible OZs with higher African American population rates than those chosen as OZs. The chosen OZs have a maximum percentage of African American population of 100% and a minimum percentage of 0%. In the case of the not chosen OZs, they also have a maximum African American population rate of 100%, but a minimum of 21%. In addition, six OZs of the 19 eligible OZs with African American population rates of 100% were chosen, while the remaining 13 were not chosen.

Figure 4 shows the percentages of both chosen OZs and eligible but not chosen OZs when ranking 8,689 eligible census tracts by the highest African American population rates. The figure shows that 29% of the eligible OZs were chosen as OZs, while 71% were not chosen. This indicates that factors involved in the selection process included more than only the African American population rate.

Also, one can consider the mix of the eligible chosen and not chosen OZs that meet both the minimums for the poverty rate (33%) and the minority population rate (85%) for both groups of OZs when ranked separately. In this case, 4,171 OZs satisfy both minimums of 33 and 85%, respectively. As Figure 5 shows, 51% of the total number of eligible OZs are chosen as OZs, while 49% are eligible but not chosen.

One can also consider the mix of the eligible chosen and not chosen OZs that meet both the minimums for the poverty rate and the African American population rate for both groups of eligible OZs when ranked separately. In this case, 3,519 OZs satisfy both minimums of 33 and 32%, respectively. As Figure 6 shows, 41% of the total number are chosen as OZs, while 59% are eligible but not chosen.

Possible missed opportunities in the choice of opportunity zones

Our analysis shows that the choice of OZs did not depend solely on the highest poverty rates among the eligible OZs. Indeed, there was a significant number of eligible OZs not chosen even though they had higher poverty rates than those chosen. In addition, the choice of OZs did not depend on the eligible OZs with the highest minority population rates or the highest African American population rates. Once again, a significant number of eligible OZs had both higher minority population rates and higher African American population rates than those chosen. Of course, the choice of OZs generally depended on the poverty rate in a census tract being greater than 20%. However, there was no requirement that a minimum percentage of those census tracts with the highest poverty rates be chosen. There was also no requirement that the choice of OZs had to take into account the percent of the population of the eligible OZs that was minority or African American.

As a further analysis, we estimate various regression equations to determine whether the poverty rate, median family income, population, percentage of population that is minority and the percentage of population that is African American are significant explanatory factors in the choice of OZs. Table 1 shows the correlations among these factors based on data for all 31,758 eligible OZs. All the correlations are highly significant and indicate that the poverty rate is negatively associated with both income and population, but positively associated with the percentage of population that is minority and African American. In addition, although not reported, the correlation between the poverty rate and the percentage of the population that is non-minority is negative and highly significant.

Table 2 examines whether these variables are significantly related to the likelihood that an eligible OZ is chosen as an OZ. More specifically, we estimate a logistic regression model so that we can evaluate the marginal effects associated with each variable. In Table 2, an OZ is coded as a 1, and an eligible OZ but not chosen as an OZ is coded as a 0. The sample consists of chosen OZs and an equal number of eligible but not chosen OZs, the latter ranked by the poverty rate. The results indicate that when taken individually, the higher the income and the greater the population, the more likely a census tract chosen as an OZ. By contrast, the higher the poverty rate and the greater both the percentages of population that are minority and African American, the less likely a census tract chosen as an OZ. When the four of the five variables are included together, with minority in one equation and African American in another equation, the results are the same for poverty, while population is no longer significant. In the case of income, it now becomes significantly negative, indicating a higher level of income is associated with a lower likelihood of a census been chosen as an OZ. The percent of the population that is minority enters with a significantly negative sign, while the percent that is African American enters with a significantly positive sign.

Table 3 is similar to Table 2, but instead of the eligible but not chosen OZs ranked by the poverty rate, they are ranked by the percentage of population that is minority. The results are the same for all of the variables when considered individually, except that the poverty rate and population are no longer significant. When four of the five variables are included together, with minority in one equation and African American in another equation, the results are the opposite for poverty, with higher poverty rates positively associated with the likelihood of an eligible OZ chosen as an OZ. Income is now not significant in one case and significantly positive in the other case. Population is now negative and significant in both cases. In the case of both the percentages of population that are minority and African American, they are both negative and highly significant.

In Table 4, the eligible but not chosen OZs are now ranked by the percent of population that is African American. The results differ somewhat from those in Table 3 when the eligible but not chosen OZs are ranked by the percentage of population that is minority. In terms of the five equations with each of the variables included by itself, income is no longer significant, but the poverty rate and population now are both positive and significant. The two equations with all variables included indicate that income changes sign and becomes insignificant in one case, while it does not change sign and becomes significant in the other case. Population changes sign but still remains significant in one case, while remaining the same in the other case. The results are the same for other variables, as in Table 3.

One last thing we do is to examine factors that may have played a role in determining the number of chosen OZs at the county level. Table 5 contains the results of this examination based upon the same explanatory factors used in previous tables. However, the variables are measured at the county level rather than the census tract level. Furthermore, the dependent variable is the number of chosen OZs in a county rather than a dummy variable with a value of either one or zero. As may be seen, the poverty rate and income are significantly negative in the two regressions. Population enters positively and significantly, as do the percentages of the population that are minority and African American.

Conclusions

The exact criteria used by state governors for choosing OZs are not publicly available. Clearly, without more information on the criteria used in the selection process for OZs, one does not know exactly why some eligible OZs were chosen, while others were not. Nevertheless, we do examine to what degree a limited group of factors, if any, played a role in the choice of OZs. The results indicate that the choices did not always include the most distressed communities in terms of high poverty rates or low family incomes or even those communities with high percentages of minority and African American populations. This suggests that other factors played a role in choosing communities in which these specific factors did not seem to be the decisive factors. Some leaders in the communities not chosen yet eligible may desire to learn more about all the factors, and the individual weight attached to each factor, that played a role in the selection process. This might allow them to understand the reasons for not choosing their communities. For example, in Alabama, many chosen OZs had an anchor institution like a university. However, the main campus of Auburn University is located in an OZ, while Tuskegee University is not. Comparing the poverty rates as well as the minority and African American population percentages for the census tracts in which these two universities are located may raise some questions. Auburn University is located in a census tract with a poverty rate of 63.2% (which includes off-campus students that skew the poverty rate upward), a minority population percentage of 14.6 and an African American population percentage of 6.8. Tuskegee University is located in a census tract with a poverty rate of 49.9% (with far fewer off-campus students), a minority population percentage of 96.3 and an African American population percentage 88.6. It is not fully clear exactly which factors played the decisive role in choosing an OZ in which Auburn University is located and not choosing an OZ in which Tuskegee University is located. Of course, it may be the case that the census tract in which Auburn University is located may have advantages over other eligible but not chosen census tracts, including that in which Tuskegee University is located, that made it attractive for new investments without the new tax benefits.

The bottom line is that there will be some time before one can evaluate which distressed communities, if any, did indeed benefit and by how much in the form of more economic development and job creation than would have otherwise occurred by being chosen OZs. There is nothing in the TCJA that requires official governmental reports to be issued regarding the impact on the distressed communities because of being designated OZs. However, President Trump issued an order on December 12, 2018, establishing the White House Opportunity and Revitalization Council. One of its purposes is to identify potential actions that federal agencies could take to support investment in OZs. Another is to determine appropriate data, metrics and methodologies to measure the effectiveness of public and private investments in urban and economically distressed communities in OZs. This is important because such information can help periodically assess the benefits and costs associated with the creation of OZs. Importantly, it can help determine whether the chosen distressed communities as compared to those eligible but not chosen were most in need of new investment and investment that stimulated greater economic development and more job creation than otherwise would have taken place. Yet, given the purpose of the TCJA, our analysis raises important questions as to whether the chosen OZs were the most appropriate ones among all the distressed communities eligible.

Figures

Figure 1.

Ranking of states and territories by number of qualified OZs

Figure 2.

Chosen and not chosen OZs based on the highest poverty rates (Total 8,689)

Figure 3.

Eligible chosen and not chosen OZs based on the highest minority population rates (total 8,689)

Figure 4.

Eligible chosen and not chosen OZs based on the highest African American population rates (total 8,689)

Figure 5.

Eligible chosen and not chosen OZs based on the highest poverty rate and minority population rate (total 4,171)

Figure 6.

Eligible chosen and not chosen OZs based on the highest poverty rate and African American population rates (total 3,519)

Table 1.

Correlations among variables

	Poverty	Income	Population	Minority	African American
Poverty (%)	1.00
Income (US$ thousands)	–0.62	1.00
Population (thousands)	–0.11	0.11	1.00
Minority (%)	0.38	–0.37	0.07	1.00
African American (%)	0.26	–0.25	–0.16	0.50	1.00

Note:

All correlation coefficients are significant at the 1% level. Also, the correlations are based on data for all 31,758 eligible OZs. Income is median family income

Table 2.

OZs and eligible but not chosen ranked by poverty rate

The dependent variable is a dummy variable equal to 1 for a chosen OZ, 0 otherwise
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Poverty (%)	–0.0565*** (0.0016)					–0.0582*** (0.0020)	–0.0626***(0.0020)
Income (US$ thousands)		0.0225*** (0.0013)				–0.0075*** (0.0015)	–0.0026* (0.0015)
Population (thousands)			0.0519***(0.0089)			0.0069 (0.0096)	0.0114 (0.0098)
Minority (%)				–0.0123*** (0.0005)		–0.0064*** (0.0007)
African American (%)					–0.0042*** (0.0006)		0.0027*** (0.0007)
Constant	2.0892***(0.1320)	–0.7920*** (0.1230)	–0.1744 (0.1165)	0.4027*** (0.0348)	0.2356** (0.1182)	2.7708*** (0.1724)	2.1818*** (0.1701)
N	15,631	15,338	15,631	16,466	15,631	15,338	15,338
Pseudo R2	0.0679	0.0170	0.0019	0.0041	0.0024	0.0722	0.0691

Notes:

The regression results are based on data for the chosen census tracts and an equal number of eligible but not chosen tracts ranked by the poverty rate. State dummies are included. Standard errors are in parentheses. *, ** and ***indicate significance at the 10%, 5%, 1% levels, respectively

Table 3.

OZs and eligible but not chosen ranked by percentage of minority population

The dependent variable is a dummy variable equal to 1 for a chosen OZ, 0 otherwise
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Poverty (%)	0.0017 (0.0013)					0.0467*** (0.0025)	0.0301*** (0.0021)
Income (US$ thousands)		0.0124*** (0.0013)				–0.0032 (0.0024)	0.0246*** (0.0021)
Population (thousands)			–0.0137 (0.0087)			–0.0219** (0.0109)	–0.0712*** (0.0095)
Minority (%)				–0.0641*** (0.0012)		–0.0765*** (0.0014)
African American (%)					–0.0191*** (0.0007)		–0.0210*** (0.0007)
Constant	–0.0586 (0.1211)	–0.4692*** (0.1229)	0.0465 (0.1163)	4.7809*** (0.1679)	1.2591*** (0.1247)	4.2626*** (0.2412)	–0.2978 (0.1882)
N	15,628	15,489	15,631	15,631	15,631	15,489	15,489
Pseudo R2	0.0004	0.0049	0.0005	0.2314	0.0419	0.2687	0.0552

Notes:

The regression results are based on data for the chosen census tracts and an equal number of eligible but not chosen tracts ranked by the minority population rate. State dummies are included. Standard errors are in parentheses. *, ** and *** indicate significance at the 10%, 5%, 1% levels, respectively

Table 4.

OZs and eligible but not chosen ranked by percentage African American population

The dependent variable is a dummy variable equal to 1 for a chosen OZ, 0 otherwise
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Poverty (%)	0.0106*** (0.0013)					0.0384*** (0.0021)	0.0352*** (0.0021)
Income (US$ thousands)		–0.0016 (0.0012)				–0.0080*** (0.0018)	–0.0024 (0.0018)
Population (thousands)			0.0342*** (0.0088)			0.0631*** (0.0098)	–0.0210** (0.0101)
Minority (%)				–0.0285*** (0.0007)		–0.0382*** (0.0009)
African American (%)					–0.0373*** (0.0008)		–0.0440*** (0.0009)
Constant	–0.3623*** (0.1212)	0.0397 (0.1213)	–0.1152 (0.1164)	2.0531*** (0.1315)	2.5156*** (0.1388)	1.5128*** (0.1869)	1.9071*** (0.1913)
N	15,630	15,454	15,634	15,634	15,634	15,454	15,454
Pseudo R2	0.0035	0.0005	0.0011	0.0794	0.1178	0.1155	0.1455

Notes:

The regression results are based on data for the chosen census tracts and an equal number of eligible but not chosen tracts ranked by the African American population rate. State dummies are included. Standard errors are in parentheses. *, **, and *** indicate significance at the 10%, 5%, 1% levels, respectively

Table 5.

County-level regressions

The dependent variable is the number of chosen OZs in a county
	(1)	(2)
Poverty (%)	–0.0476** (0.0206)	–0.0598*** (0.0192)
Income (US$ thousands)	–0.1245*** (0.0092)	–0.1246*** (0.0091)
Population (thousands)	0.0253*** (0.0003)	0.0252*** (0.0003)
Minority (%)	0.0142** (0.0061)
African American (%)		0.0497*** (0.0082)
Constant	7.1422*** (0.9282)	6.4918*** (0.9256)
N	3,220	3,220
Adjusted R2	0.7866	0.7887

Notes:

The regression results are based on data for each county. State dummies are included. Standard errors are in parentheses. *, ** and *** indicate significance at the 10%, 5%, 1% levels, respectively

Notes

1. The Fundrise Opportunity Fund provides a calculator to estimate how much money an investor could potentially save on a capital gain by investing through its fund versus a standard stock portfolio. See https://fundrise.com/education/blog-posts/tax-incentives-of-investing-in-opportunity-zones, accessed February 17, 2020.

2. This applies to a low-income community located in a metropolitan area, while only the statewide average applies to a low-income community located outside a metropolitan area.

3. See www.cdfifund.gov/Pages/Opportunity-Zones.aspx, accessed February 17, 2020.

4. The IRC provides that the states can choose up to 25 census tracts as OZs containing fewer than 100 low-income communities. This was the case in ten states and four US territories.

5. Gelfond and Looney (2018) point out not all OZs were truly distressed and state governments, with broad discretion to select OZs faced a conflict between selecting distressed areas versus already gentrifying areas that were more likely to provide tax benefits to qualifying investors.

6. Frank et al. (2020) examine the role of political affiliation during the selection of OZs. They find that census tracts with the same political affiliation as the governor are, on average, 7.6% more likely to be selected as OZs.

7. Practically, all of Puerto Rico is designated an OZ.

8. The tax benefits are available to individuals even if they do not live, work or own a business in an OZ.

9. Eligible taxpayers include individuals, C corporations, regulated investment companies (RICs), real estate investment trusts (REITs), partnerships, S corporations and trusts and estates. In addition, the sponsors of OFs may require relatively high minimum investments for investors.

10. To receive the full benefit of the 15% basis increase, taxpayers would have to invest in OFs on or before December 31, 2019, and for the 10% basis increase, on or before December 31, 2021.

11. On October 19, 2018, the Internal Revenue Service published proposed regulations (REG-115420-18) providing some guidance to taxpayers interested in investing in OZs. A second set of proposed regulations (T.D. 9889) was issued on May 1, 2019, and the final regulations (REG-120186-18) were issued on December 19, 2019. The final regulations addressed comments received in response to the first two sets of proposed regulations. However, they retained their basic approaches and structures, but included clarifications and modifications to the earlier proposed regulations (see Miller et al., 2020). For more details on the provisions of the final regulations and comparisons to earlier proposed regulations, see Szilvas (2020). One might also consult Miller et al. (2019) and Bertrand et al. (2020).

12. For more information, see Miller et al. (2018) and Zhang (2019, 2018).

13. For a clarification of the term “substantially all” in the final regulations, see Shearman and Sterling (2020), Opportunity Zones: Final Regulations Provide Additional Flexibility, Perspectives.

14. We include the two opportunity zones that the Treasury Department designated on December 14, 2018.

15. 2011–2015 American Community Survey (ACS) five-year data from the Census Bureau.

16. Also, see Lester et al. (2018) for summary statistics of selected OZs.

17. Two of the OZs are located in Puerto Rico and one is located in Macomb County, Michigan. The population in these three OZs ranges from a low of 15 to a high of 44 persons.

18. Some census tracts with a median family income that generally did not exceed 80% of the greater of the statewide average or the metropolitan area average were chosen as OZs.