Crowdfunding and the Not-So-Safe SAFE


On May 16, 2016, more than four years following the enactment of the Jumpstart Our Business Startups Act (the “JOBS Act”), the much-anticipated era of retail crowdfunding officially began in the United States.[1] On the very first day that the Securities and Exchange Commission’s (“SEC”) new Regulation Crowdfunding went into effect, seventeen companies launched crowdfunding campaigns on various online platforms—known as “funding portals”—that registered with the SEC to host offerings. Over the past several months, dozens of companies have solicited investments through these portals to finance the development of biodegradable toothbrushes,[2] custom-printed condoms,[3] and glow-in-the-dark vegetation,[4] among other projects.[5]

While it is far too early to pass judgment on the long-term prospects of the crowdfunding project more generally, it is possible at this juncture to assess how certain aspects of crowdfunding are developing and to identify potential pitfalls for the players in this new arena. In at least one area—the menu of financing instruments being offered to prospective retail investors—we believe that early market participants may be unintentionally sabotaging the crowdfunding experiment. Specifically, we believe that the forms of a relatively new startup-financing instrument, the simple agreement for future equity (“SAFE”), currently offered by crowdfunding portals such as WeFunder[6] and Republic,[7] contain terms that are likely to frustrate the ability of investors to share in the upside of successful crowdfunding companies. In other words, crowdfunding investors who purchase SAFEs may discover that these instruments are anything but.

To be clear, we do not argue here that the SAFE has no role to play in providing capital to early-stage companies. Outside of the crowdfunding context, there are situations in which the SAFE may be a sensible instrument for startups to use when fundraising. In the crowdfunding context, however, the vast majority of companies raising money are unlikely to ever raise institutional venture capital (“VC”). Since the SAFE was developed as a means of investing in startups that expect to raise such funding at a later date, it is not the right tool for channeling retail investment capital to crowdfunding companies. Even if the terms of the SAFEs currently offered by WeFunder and Republic were to be rewritten, the use of the SAFE in crowdfunding would still present a number of issues from the perspective of a retail investor. Accordingly, we argue that the most promising solution to the problems we identify in this Essay is for the funding portals to remove the SAFE from their menus of financing instruments.

This short Essay proceeds as follows. Part I surveys the types of securities available to crowdfunding companies via the new funding portals. Part II describes the origins of the SAFE. Part III describes the types of crowdfunding companies that have issued SAFEs to date and argues that many of these companies are unlikely to raise institutional VC. Part IV surveys and criticizes the terms of the SAFEs currently on offer by several funding portals. The Essay concludes by discussing several possible solutions to the problems identified herein.

I. Types of Crowdfunding Securities

The JOBS Act crowdfunding provisions did not include any explicit restrictions on the types of securities that issuers could sell in crowdfunding offerings.[8] The SEC considered regulating the types of crowdfunding securities, soliciting comments regarding whether it should, for instance, only permit crowdfunding issuers to offer plain-vanilla equity securities.[9] Based on feedback the SEC received during the comment period and its interpretation of congressional intent in Title III of the JOBS Act, the SEC decided to allow issuers to offer any type of security in a crowdfunding offering, so long as investors are given adequate disclosure about the structure and terms of the investment.[10] The SEC declined to narrow the list of instruments that companies could offer crowdfunding investors in order to give issuers some flexibility as this new market develops.[11]

As a result, startups looking to raise capital through crowdfunding have had free rein to choose whichever instruments they believe best fit their needs. They have offered crowdfunding investors a variety of securities thus far, including common and preferred equity, debt instruments (with rates of return and payment schedules that are either fixed or vary with the company’s revenues), and convertible securities (such as convertible notes and SAFEs).[12] Most of these instruments have longstanding roles in early-stage technology startup and small-business finance. For many years, early investors in tech startups received the same common stock that a startup’s founders received, until it became more common for those early-stage angel investors to purchase convertible notes.[13] Institutional VC investors have traditionally negotiated for preferred stock with liquidation preferences and other minority protections.[14] Debt instruments with fixed or variable repayment schedules have long been staples of financing for small businesses with revenue models that generate sufficient cash flow to service the debt and provide an adequately attractive risk-adjusted return to the lenders, be they community banks, the Small Business Administration, or high-net-worth individuals. Compared to these other instruments, the SAFE is the new kid on the block, having emerged at the end of 2013 and only more recently becoming widely used as a startup-financing tool.[15] To understand how this new instrument may have an adverse impact on the new crowdfunding ecosystem, it is first necessary to briefly discuss the origins of the SAFE.

II. The Simple Agreement for Future Equity (“SAFE”)

The SAFE was developed by Y Combinator, the well-known startup accelerator based in Silicon Valley, as a means of investing in startups that expected to raise institutional VC at a later date.[16] Although the SAFE resembles a classic seed-stage convertible note in most respects, the SAFE is not a debt instrument. It lacks the convertible note’s maturity date and does not accrue interest while it remains outstanding. The SAFE is also not an equity instrument, and its holders are owed no fiduciary duties until the instrument converts into equity. Moreover, the SAFE does not pay dividends and the SAFE holder has no right to vote on matters submitted to shareholders.[17] The SAFE is, in essence, a contractual derivative instrument. It is a deferred equity investment that will prove valuable to the holder if, and only if, the company that issues it raises a subsequent round of financing, is sold, or goes public.

The SAFE was originally created to facilitate early-stage investments in the companies participating in Y Combinator’s accelerator program. Startups that have been through the semi-annual Y Combinator program include several so-called “unicorns” (startups with private valuations of at least $1 billion), most notably Airbnb and Dropbox.[18] These technology companies aspire to follow a fairly well-defined growth trajectory: They raise significant sums of capital, spend it quickly to achieve as much growth as possible as quickly as possible, raising more money along the way to continue their expansion at breakneck pace until achieving a liquidity event—usually in the form of a sale of the company or, in fewer cases, an initial public offering.[19] Savvy startup investors typically view the outcomes of seed investments in these companies as essentially binary: The companies will either succeed or go bust, leaving the investors with either a lucrative multiple return on their investment or a loss of most, if not all, of their principal. Often, in the downside scenario, the founders and investors try to salvage as much of their investments (and reputations) as possible through a sale or acqui-hire, but modest, middling returns are not what most investors are seeking in the feast-or-famine world of seed-stage startup investing.[20]

In addition to receiving an investment from Y Combinator for participating in the accelerator program, in many cases the initial investments in Y Combinator portfolio companies via the SAFE come from a coterie of high-profile angels and VC investors who routinely fund the accelerator’s portfolio companies with relatively small amounts of seed capital.[21] Y Combinator has marketed the SAFE as being “simple,” in that it is a minimalistic contract of only a few pages, containing little legalese and contractual boilerplate as well as fewer terms than the convertible notes that these parties were already quite familiar with (and which these investors had been using for years to invest in Y Combinator companies). Switching from convertible notes to SAFEs had the added benefit—at least, from the founders’ perspective—of not requiring the additional legal work often needed to extend the maturity date of convertible notes if a subsequent financing had not occurred prior to maturity. Using SAFEs also allowed founders to avoid having difficult conversations with convertible noteholders at maturity if the company was not performing as expected or was having difficulty raising a subsequent round of financing. Effectively, the SAFE purported to improve upon a very specific concern (the maturity feature of convertible notes) encountered by a particular type of company (unfunded tech startups, specifically those participating in Y Combinator’s accelerator program) and a few specific groups of people (founders of hot startups and highly experienced startup investors competing for access to those companies) within the bubble of the Silicon Valley startup ecosystem. Given Y Combinator’s prominence as an influencer in the startup world, startups outside the Silicon Valley ecosystem have since increasingly adopted the SAFE as a seed-financing tool.

SAFEs can be suitable investment instruments for companies—like Y Combinator portfolio companies—that are strong candidates for future VC investment. It is important to bear in mind, however, that SAFEs are highly company-favorable securities—a product of the latest startup-financing frenzy—requiring investors who understand and accept the binary nature of investing in early-stage tech startups and who believe that the company will eventually be in a position to raise institutional VC so that the SAFEs will convert to equity as intended. In the context of crowdfunding, the use of SAFEs has the potential to result in some unexpected and unfavorable outcomes for the uninitiated.

Indeed, for companies and investors outside the clubby startup world of U.S. technology hubs like Silicon Valley, the nomenclature “SAFE” may actually be somewhat misleading. Retail investors, who presumably are used to investing in traditional asset classes, such as publicly traded stocks and bonds, are unlikely to be familiar with the convertible notes and SAFEs that more sophisticated accredited investors use to invest in tech startups. As a result, they are also unlikely to find the mechanics by which SAFEs convert to equity to be particularly “simple.”[22] The safety implied by the clever acronym “SAFE” actually points to the instrument’s safety for the issuing company—which is able to avoid the maturity dates associated with convertible notes—rather than any safety for the investor. A potential problem with using SAFEs in crowdfunding, therefore, is that inexperienced retail investors may mistakenly believe that they are receiving something simple and safe, a security that they believe all of the top startups and investors in Silicon Valley use, and make an investment without fully understanding the risks that they are assuming by purchasing those SAFEs.

III. Types of Crowdfunding Issuers Opting for SAFEs

Of the 96 issuers to launch crowdfunding offerings through August 31, 2016, 30 issuers (approximately 31%) chose to offer convertible securities (such as convertible notes, SAFEs, or similar instruments) to prospective crowdfunding investors. Ninety percent of the convertible securities used were SAFEs. The remaining convertible securities were convertible notes.

Two different types of issuers have opted to use SAFEs thus far in their crowdfunding offerings:


1. Tech startups with business models and growth trajectories that are potentially attractive to VC investors; and

2. Non-tech startups with business models that are less likely to attract VC investment.


Many tech startups using SAFEs in crowdfunding offerings to date hail from technology hubs—places like the San Francisco Bay Area, Boston, New York, and Southern California—where the influence of Y Combinator is strongest.[23] For some of these companies, using SAFEs with crowdfunding investors is not likely to cause any serious issues because the SAFE was designed for investing in these types of companies—tech startups that are likely to either raise institutional VC or fail. But even crowdfunding issuers that are tech startups and that have business models which, at first blush, would appear attractive to VC investors (and therefore suitable candidates for using SAFEs) may be less likely to raise future VC financing than the typical tech startup. Due to the additional costs and disclosures required of crowdfunding issuers, most startups that have access to traditional forms of startup fundraising will be loath to undertake a crowdfunding offering. As a result, many of the startups that choose to pursue crowdfunding as a means of raising capital do so because they have no other options, and they may still struggle to raise traditional venture financing down the road.[24] Additionally, some of the startups using SAFEs are not based in technology hubs, and may have turned to crowdfunding because they are outside of traditional angel and VC networks. These factors may mean the SAFE is an inappropriate instrument for these investments, since the SAFE is predicated on the expectation that the issuer will eventually raise a round of institutional VC and otherwise follow the traditional path of a high-tech venture-backed startup.

The second category of crowdfunding issuers using SAFEs—non-tech startups—presents even greater concerns. These are companies with business models and growth trajectories that often look quite different from tech startups. As a result, these companies are less likely to be candidates for VC investment and more likely to evolve into either lifestyle businesses for the founders—providing them with healthy salaries and the ability to distribute any profits to themselves in the form of dividends for the foreseeable future—or companies that rely on debt financing (such as bank loans) and reinvested profits to support additional growth. These companies, even if they are successful, may never raise additional equity capital, be sold, or go public, leaving SAFE holders with no way to receive returns on their investments.[25] The SAFE was simply not designed to be used to invest in this type of company.

IV. Funding Portals and Variations on a SAFE

Thus far, we have been discussing the conceptual concerns with different types of companies using SAFEs in crowdfunding, but there are also more specific issues raised by the forms of SAFEs that actual crowdfunding issuers have offered to prospective investors. These SAFEs have all been based on the forms made available to the issuers by the funding portals they chose to host their offerings. For instance, WeFunder, which has been the most popular funding portal to date thanks to its streamlined disclosure process and industry-low commission (at four percent of funds raised),[26] has a form of SAFE available on its website that every WeFunder company using SAFEs has adopted.[27] The WeFunder SAFE has a number of features that may exacerbate some of the problems we have described with the use of SAFEs in crowdfunding generally.

There are typically three scenarios in which SAFE investors receive cash back from their investment:

1. Post-Conversion Liquidity Event. In this scenario, the company sells priced equity securities following the SAFE financing, and the SAFEs convert into those equity securities based on the discount or valuation cap stated in the SAFE contract. At some point following the conversion, the company is sold or goes public, and the former SAFE holders receive proceeds from those liquidity events just like the other investors (such as VCs) holding those equity securities.

2. Pre-Conversion Liquidity Event. If the company is sold before it raises a subsequent round of priced equity capital (in which case the SAFEs would still be outstanding), the SAFE holders would elect to either (A) convert the SAFEs to equity and receive proceeds from the sale based on their pro rata equity ownership, or (B) receive a cash payout of their original investment amount (plus some pre-negotiated return, such as 1.5x2x) in connection with the sale.

3. Dissolution Event. If the company shuts down and liquidates prior to raising a subsequent round of priced equity financing, the SAFE holders would receive any residual assets up to the amount of their original investments.

One scenario is not anticipated in most SAFEs and is also not addressed in the WeFunder SAFE: a scenario in which a company never raises additional equity capital and never sells itself or goes public.[28] This scenario is not anticipated because it is a rare outcome for venture-backed tech startups. As we have discussed, however, crowdfunding offerings are not undertaken exclusively by tech startups. Imagine a non-tech company that raises capital in a crowdfunding offering using a SAFE. The company uses that capital to launch a product or service, which starts generating significant cash flow before the company needs additional capital. The company is able to use that cash flow to obtain bank financing and may even have profits to reinvest in growing the business. At some point, that company may also have sufficiently healthy profits to start distributing those profits to its owners (the founders). This business, following a path that is extremely common—perhaps the norm—for non-tech startups and small businesses, could continue in this fashion in perpetuity without ever needing additional equity capital or needing to sell. If that were to happen, the SAFE holders would continue to hold their securities, earning no interest, receiving no dividends and never seeing any return of their original investment. We call this the “dividend problem.”[29]

The WeFunder SAFE amplifies the dividend problem because a financing conversion only occurs under the contract when the issuer closes a bona fide preferred stock financing raising any amount at a fixed pre-money valuation. The SAFEs do not convert if the company raises equity capital by selling common stock.[30] The SAFE is often drafted this way because it presupposes that the next financing round will be a traditional VC investment and the typical VC investment is structured as preferred equity. However, crowdfunding issuers (which, as discussed above, could be tech or non-tech companies) raising subsequent equity capital from non-VC sources may choose to issue common stock instead of preferred stock. In that case, the SAFEs issued to crowdfunding investors using the WeFunder form would remain outstanding until the company is sold. Under the terms of the WeFunder SAFE, a company could theoretically raise unlimited amounts of private capital selling common stock and distribute profits to those investors and the founders via dividends without ever triggering a conversion of the SAFEs or allowing the SAFE holders to participate in those dividend payments.

The WeFunder SAFE contains yet another provision that may frustrate the ability of many SAFE holders to share in the upside of successful crowdfunding companies. The issuer can repurchase the SAFEs of non-accredited investors for the fair market value of the instrument, as determined by an independent appraiser of the company’s choosing, at any time prior to conversion. This means that the investors taking the greatest risk (the seed crowdfunding investors) can be prevented from seeing the bulk of the returns from the most successful companies they fund using the WeFunder SAFE. Even if the WeFunder SAFE converts to equity, moreover, the contract provides that the SAFE converts into a non-voting series of preferred stock, leaving the crowdfunding investors at the mercy of the founders and more sophisticated investors who negotiate special rights for themselves (although post-conversion, the former SAFE holders would at least be owed fiduciary duties by the company’s board of directors).[31]

WeFunder is not the only funding portal to create a form of SAFE that adds to the problems inherent in using SAFEs in crowdfunding offerings. Republic, a funding portal created by former employees of the well-known startup investment platform AngelList, created its own form called the Crowd SAFE. Like the WeFunder SAFE, the Crowd SAFE is based on Y Combinator’s version but modified in various ways for use in crowdfunding offerings.[32] The Crowd SAFE converts into stock in connection with any priced equity financing (preferred or common) raising proceeds of at least $1 million.[33] Republic added a new feature to the Crowd SAFE, however, allowing the company to postpone the conversion of the instrument until a liquidity event (in most cases, the sale of the company), while promising investors that they will receive the same economics (that is, the same conversion price) regardless of when they actually convert.[34] The Crowd SAFE effectively allows the company to raise any form of equity capital without triggering the conversion of the SAFEs, while also neglecting (like the WeFunder SAFE) to account for a scenario in which the stockholders of the company receive their return in the form of dividends and not in a liquidity event such as a sale or initial public offering.


The SEC has two competing missions in all of its regulatory endeavors: promoting capital formation and protecting investors. Regulation Crowdfunding has largely been viewed as heavily favoring investor protection over capital formation (particularly the disclosure requirements).[35] When it came to the types of securities available to crowdfunding issuers, however, the SEC took a laissez-faire approach. With many aspects of crowdfunding, such as policing individuals’ annual investment limits and screening prospective issuers for fraudulent schemes, the SEC chose to rely heavily on the funding portals to make crowdfunding as safe as possible for non-accredited investors. When the SEC declined to narrow the list of permissible securities, perhaps its expectation was that the funding portals would help keep issuers from offering retail investors inappropriate securities through their platforms. Unfortunately, this does not seem to be happening in practice.

There are several possible solutions to the problems identified in this Essay. First, the funding portals could seek to limit the use of SAFEs to the “right” sort of companies—those that are likely to raise future capital from institutional investors. Policing the types of securities offered by crowdfunding companies may sound like a lot to ask of the portals but many of them already market themselves as significantly curating the offerings they make available on their platforms. Accordingly, we do not see this additional curation as overstepping.[36] Second, the portals could amend the forms of SAFE currently on offer to address some of the specific issues we have raised. This would be a positive development, to be sure, but it would not address the deeper problems that flow from the fact that many of these crowdfunding issuers will never raise institutional VC. Third, the funding portals could remove the SAFE from their menu of financing instruments. We believe that this last approach represents the simplest and best solution. A crowdfunding company that wants to issue a SAFE-like security could instead issue a convertible note, which is similar to the SAFE in many respects but which accrues interest, has a maturity date, and offers retail investors other protections that are associated with debt instruments. Despite these additional investor protections, however, convertible notes are also less than ideal instruments for most companies in the crowdfunding context because, like SAFEs, they too are intended for use by companies that are likely to raise institutional VC in the near term. Alternatively, and we believe preferably, the company could issue debt, common equity, or preferred equity (the latter two providing investors with the full benefits of being shareholders of the company including, most importantly, the protection of fiduciary duties owed by the company’s board of directors). These alternatives are, in our view, more suitable vehicles for channeling capital to crowdfunding companies than the SAFE.

In closing, it should be emphasized that all of these instruments—SAFEs, convertible notes, common stock, preferred stock, etc.—are simply labels. It is not the name of the instrument that matters so much as the terms set forth within it, that is, the balance struck between issuer and investor. It is possible to issue “common stock” that contains terms commonly used in “preferred stock” financings. It is also possible to issue “SAFEs” that contain terms that make them virtually indistinguishable from “convertible notes.” In this respect, our recommendation that funding portals remove the SAFE from their menu of financing instruments might be criticized as emphasizing form over substance. To be clear, our quarrel is not with the SAFE qua SAFE. Our quarrel is with the terms contained within the SAFEs currently on offer in the retail crowdfunding space as well as the specific context in which these contracts are being used. Unless and until the terms of these instruments are revised to address the concerns outlined above, we do not believe that crowdfunding issuers should use them. The revisions that would be necessary to adequately address these concerns would effectively turn the SAFE into a different instrument (a convertible note, preferred stock, etc.) in all but name, making it better, in our view, to simply remove the SAFE from the menu of financing instruments and use existing instruments that are more fit for this purpose. The SAFE is a financing instrument that was developed to fund early-stage companies that expect to raise institutional VC. This expectation informs the terms set forth in the SAFE. The vast majority of crowdfunding companies are unlikely to raise institutional VC. Accordingly, for all the reasons we have discussed, we believe that SAFEs are not well suited to being used in crowdfunding transactions.


[1]Generally speaking, the term “crowdfunding” refers to the “practice of funding a project or venture by raising money from a large number of people, each of whom contributes a relatively small amount, typically via the Internet.” See Crowdfunding, Oxford English Dictionary Online (3d ed. 2015),‌ntry/429943‌?redirec‌tedF‌rom=cr‌owdfunding&  []. As used herein, the terms “crowdfunding” and “retail crowdfunding” refer to the process of raising capital from non-accredited investors through securities offerings under § 4(a)(6) of the Securities Act of 1933 (codified as amended at 15 U.S.C. § 77d(a)(6) (2012)), and the SEC’s Regulation Crowdfunding, 17 C.F.R. § 227 (2016), and are not meant to include other types of securities offerings or fundraising campaigns that are also often considered forms of crowdfunding, such as online securities offerings to accredited investors under Rule 506(b) or Rule 506(c) of Regulation D, 17 C.F.R. §§ 230.506(b)–(c) (2016), mini-public offerings under Regulation A (as amended by the JOBS Act and now commonly called Regulation A+), 80 Fed. Reg. 21,805 (June 19, 2015), or rewards-based fundraising campaigns on popular platforms such as Kickstarter and Indiegogo.

[2]See Do., LLC, Offering Statement (Form C) (May 16, 2016), https://www‌.s‌ec.g‌ov/Archives/edgar/data/1674379/000167025416000015/0001670254-16-000015-index.htm [].

[3]See Graphic Armor, Inc., Offering Statement (Form C) (May 16, 2016), https://ww‌ [].

[4]See TAXA Biotechnologies, Inc., Offering Statement (Form C) (May 16, 2016), []. 

[5]See Jack Wroldsen, Crowdfunding Investment Contracts, 11 Va. L. & Bus. Rev. (forthcoming 2017) (discussing the wide variety of projects seeking funding from the crowd).

[6]WeFunder, Legal Primer for Founders, [ht‌tps://]

[7]Republic, The Crowd Safe, [].

[8]See Securities Act of 1933, ch. 38, § 4(a)(6), 48 Stat. 74 (codified as amended at 15 U.S.C. § 77d(a)(6) (2012)).

[9]See Crowdfunding, 78 Fed. Reg. 66,428, 66,458 (proposed Nov. 5, 2013).

[10]See Crowdfunding, 80 Fed. Reg. 71,388, 71,427 (Nov. 16, 2015).

[11]See id. at 71,506.

[12]See Wroldsen, supra note 5.

[13]See John F. Coyle & Joseph M. Green, Contractual Innovation in Venture Capital, 66 Hastings L.J. 133, 146–48 (2014).

[14]Id. at 149–51.

[15]Id. at 168–69.

[16]Y Combinator’s form of SAFE is available at https://www.ycombinat‌‌m/do‌cuments/#safe [].

[17]To be clear, the SAFE is not unique in this regard. Common and preferred stockholders of private companies in the tech sector rarely receive dividends, as these companies typically invest all available capital in future growth. Convertible notes also do not grant the holder the right to vote on matters submitted to the shareholders.

[18]Harold J. Krent & Dawn K. Young, Self-Interested Fiduciaries and the Incubator Movement, 66 Syracuse L. Rev. 611, 618 (2016).

[19]The fact that not all startups achieve this growth trajectory should not distract from the essential point that many, if not most, tech startups aspire to it. See Paul Graham, Startup = Growth, []; see also Victor Fleischer, The Missing Preferred Return, 31 J. Corp. L. 77, 90 (2005) (describing the “home run mentality” in the VC industry); Elizabeth Pollman, Information Issues on Wall Street 2.0, 161 U. Pa. L. Rev. 179, 237 n.303 (2012) (same).

[20]See John F. Coyle & Gregg D. Polsky, Acqui-hiring, 63 Duke L.J. 281, 283–84 (2013).

[21]Coyle & Green, supra note 13, at 170.

[22]The SAFE’s “simplicity” presupposes an investor’s familiarity with the terms and mechanics of a convertible note (the instrument on which the SAFE was modeled).

[23]A few crowdfunding issuers are even Y Combinator portfolio companies, so their decision to use SAFEs for their crowdfunding investors is understandable. In addition, the funding portal WeFunder, which has led the way when it comes to issuers employing the SAFE in crowdfunding offerings on its platform, is itself a Y Combinator portfolio company.  Ryan Lawler, Y Combinator-Backed WeFunder Launches to Bring Crowdfunding Startups to the Masses, TechCrunch (Mar. 19, 2013), [].

[24]This adverse selection problem is sometimes described as a “market for lemons.” Darian M. Ibrahim, Equity Crowdfunding: A Market for Lemons?, 100 Minn. L. Rev. 561 (2015).

[25]Outside the context of crowdfunding, in the unlikely event that a startup that raises seed capital using SAFEs never raises a subsequent round of equity financing and instead turns into a lifestyle company, the personal relationship of the founders with angel investors and startup community norms may lead the founders to agree to convert the SAFEs to stock without being contractually required to do so. This type of extracontractual resolution to an unwelcome outcome not contemplated in the investment contract would seem less likely for companies that are not a part of that community, with investors that they do not know personally, as we would expect to be the case for crowdfunding issuers.

[26]WeFunder, Risks, [‌3VG-RJ5C].

[27]For an overview of the WeFunder SAFE and the forms being used by WeFunder’s crowdfunding issuers, see WeFunder, Legal Primer for Founders, https://wefu‌nd‌ [].

[28]WeFunder’s Legal Primer for Founders advises crowdfunding companies that SAFEs are “best for early stage startups – raising with Regulation Crowdfunding – expecting to get acquired or file for an IPO in the future.” Id. Many early-stage companies may think that these outcomes are much likelier than they are, particularly for non-tech businesses, and will end up neither selling the business nor going public but simply continuing to operate as a private concern.

[29]Among the typical seed-stage startup-financing instruments, the dividend problem is uniquely an issue with SAFEs. Investors who purchase the same common stock that a company’s founders hold can rest assured that any dividends declared by the company will be paid to all common stockholders on a pro rata basis. Preferred stockholders, at least in the venture-backed startup context, always ensure that they will receive any dividends paid to the common stockholders, and often negotiate for an additional preferred dividend (despite the fact that these are almost never actually paid). Convertible noteholders typically have the option at maturity to convert into common stock and receive dividends if the company has not yet raised a qualifying round of capital triggering the conversion of the notes into equity. Though it is exceedingly rare for tech startups to pay dividends, the SAFE is the only startup-financing instrument that does not at least account for the possibility and provide the investor with some modicum of protection in this regard.

[30]Twenty-seven of the first sixty crowdfunding issuers offered their investors common equity securities, highlighting the likelihood of these types of issuers opting to sell common stock instead of preferred stock (which was chosen by only six of the first sixty crowdfunding issuers). Practical Law, What’s Market: Federal Crowdfunding Offerings (last updated Sept. 16 2016), [].

[31]See, e.g., WeFunder, WeFunder SAFE—Valuation Cap, Delay Conversion until IPO/Acquisition,‌efunderCr‌owdf‌undingSAFE_IPO.rtf [].

[32]For an overview of the Crowd SAFE and the forms being used by Republic’s crowdfunding issuers, see The Crowd Safe, Republic, [https://p‌er‌].

[33]Including this type of de minimis threshold on a financing that triggers conversion as a protection for investors is common in convertible note deals.

[34]WeFunder now offers a form of SAFE that similarly allows companies to delay conversion of the SAFEs in this manner. WeFunder, WeFunder SAFE—Valuation Cap, Delay Conversion until IPO/Acquisition, https://wefunder-production.s3.amazonaw‌‌tat‌ic/WefunderCrowdfundingSAFE_IPO.rtf [].

[35]See, e.g., Abraham J.B. Cable, Mad Money: Rethinking Private Placement, 71 Wash. & Lee L. Rev. 2253, 2256–58 (2014) (discussing “investment caps” that limit to $25,000 the amount that any one individual can invest in crowdfunding offerings annually).

[36]If funding portals are unwilling to provide this type of curation on their own, perhaps the SEC and/or the Financial Industry Regulatory Authority (“FINRA”) should mandate it. Their approach in doing so could be modeled on what these regulators currently require of broker-dealers in the context of purchasing other types of derivative instruments akin to the SAFE. Retail brokerages—such as Fidelity, Vanguard et al.—are required to obtain certain information from customers seeking to purchase options or security futures through their brokerage accounts to enable the brokers to assess the suitability of these derivative instruments for those particular customers, given their financial position and investment experience. See FINRA Rule 2360(b)(16) (2014), http://finra.complinet.‌com/en/disp‌lay/display_m‌ain.html?rbid=2403&element_id=6306 []; FINRA Rule 2370(b)(16) (2011),‌n/display/display_main.htm‌l?rbid=2403&el‌ement_id=6309 []. The FINRA rules require the broker to determine whether the transaction is suitable for the customer based on the “customer’s investment objectives, financial situation and needs” and the broker’s judgment of whether “the customer has such knowledge and experience in financial matters that he may reasonably be expected to be capable of evaluating the risks of the recommended transaction, and is financially able to bear the risks of the recommended position.” See FINRA Rule 2360(b)(19) (2014),‌isplay/display_main.html?rbid=‌2403&e‌lement_id=6306 []; FINRA Rule 2370(b)(19) (2011), http://fin‌ [https‌://p‌er‌]. Since the SAFE is effectively a prepaid forward—the private company version of a future—perhaps the requirements placed on brokers allowing customers to trade security futures provide the best analogy (although the suitability assessment and required diligence are largely the same for options and futures under the FINRA rules).

    Unlike registered broker-dealers, funding portals are actually not permitted to “offer investment advice or recommendations” to the investors in crowdfunding offerings conducted through their platforms. See 17 C.F.R. § 227.402(a) (2015).  Funding portals are, however, allowed to “[d]etermine whether and under what terms to allow an issuer to offer and sell securities in reliance on section 4(a)(6) of the Securities Act (15 U.S.C. 77d(a)(6)) through its platform” within the SEC’s safe harbor from the broker-dealer registration requirements of § 3(a)(80) of the Securities Exchange Act of 1934, 48 Stat. 881 (codified as amended at 15 U.S.C. § 78(c)(a)(80) (2012). 17 C.F.R. § 227.402(b)(1) (2015).  Imposing a requirement on the funding portals similar to those already required of broker-dealers in the option and security future trading context—namely requiring the portals to pass on the suitability of financing instruments (and particularly derivative contracts like the SAFE) offered by issuers to retail crowdfunding investors through their platforms—could be an intermediate regulatory response to the issues we have raised in this Essay, short of an outright restriction on the types of securities available to crowdfunding issuers and investors to plain-vanilla equity and debt instruments.

Retooling the Amicus Machine

The U.S. Supreme Court’s 2015–16 Term may have finished with a diminished number of Justices and a smaller-than-usual docket,[1] but one thing that did not change was the large number of friend of the court, or amicus curiae, briefs filed in some individual cases and for the docket as a whole.[2] The large number of amicus briefs filed is a continuation of a longer trend. For over twenty years, increasing numbers of amicus briefs have been filed before the Court, at both the certiorari and merits stages. Inevitably, one or more high-profile cases each Term attract an extraordinarily large number of such briefs, but even the lower-profile cases often have a healthy number of briefs filed.[3] Amicus briefs have become a ubiquitous presence in the litigation of cases before and the rendering of decisions by the Court,[4] with frequent references to the briefs in oral arguments and in citations appearing in the Court’s decisions.[5] They are now a familiar feature of the coverage and discussion of the Supreme Court.[6]

For a not-unrepresentative example from the Term, consider Whole Woman’s Health v. Hellerstedt, where a 5-3 majority held unconstitutional certain restrictions Texas had placed on facilities providing abortions.[7] The majority opinion by Justice Stephen Breyer observed that in applying the “undue burden” test, the Court would rely “heavily on the District Court’s factual findings and the research-based submissions of amici in declaring a portion of the law at issue unconstitutional.”[8] The majority proceeded to do just that, citing and quoting from at least five amicus briefs at some length in the text.[9] Justice Ruth Bader Ginsburg wrote a brief concurrence which similarly cited and summarized three amicus briefs.[10]

“Much academic ink has been spilt on the study of amicus curiae in the Supreme Court”[11] by law professors and political scientists, exploring the reasons why large numbers of amicus briefs are filed (often by interest groups) and attempting to measure their influence on the Court.[12] In their recent article, Professors Allison Orr Larsen and Neal Devins usefully augment this literature by describing and evaluating what they call “The Amicus Machine.”[13] Their article adds two important components to the study of the amicus phenomenon. First, they surveyed lawyers who regularly practice before the Court and document that litigants and their attorneys regularly solicit amicus briefs to be filed for their side. Thus, such briefs are not spontaneously filed by individuals and interest groups who might be thought to be coincidentally following the Court, but rather are often the result of an orchestrated campaign by litigants and their agents.[14] Second, they step back and critically evaluate the pros and cons of such briefs for the Court as an institution, and conclude that on balance it is a good thing. Among other things, they contend, the briefs provide useful information to the Court about the importance of cases in which review is sought and aid the Court in deciding cases after certiorari is granted.[15]

Larsen and Devins have made an important contribution to the literature on amicus briefs by shedding light on the nonrandom processes that lead to many such briefs being filed in the Supreme Court. They are less convincing, in my view, in their normative and largely positive take on the influence of amicus briefs in general, and orchestrated amicus briefs in particular, on the Court. Like them, I think that the briefs can provide useful information to the Court that for whatever reason the litigants’ own briefs do not, but I find the downside of amicus briefs more problematic. The large number of such briefs, particularly when they are orchestrated, can in some instances influence the Court too much, be detrimental to the traditional adversarial process, and could also lead some attentive members of the public to conclude that the Court is just another political institution, capable of being lobbied like any other.

This Essay proceeds as follows. Part I summarizes the large number of amicus briefs filed in the Court in the last several decades, their apparent increasing influence on the Justices, and the new research and insights of Larsen and Devins. Part II critically examines the less benign impacts of these briefs generated by the amicus machine, both on and off the Court. Part III suggests several possible ways that the machine might be retooled, such as making it easier for parties to respond to amicus briefs, limiting the numbers of such briefs, permitting the parties to pick such briefs to be filed, and encouraging the Court to modulate deference doctrines when relying on such briefs. Those reforms might allow for the pros identified by Larsen and Devins while limiting the cons.

I. The Political Economy of Amicus Curiae Briefs

Amicus curiae briefs have been filed in the Supreme Court for many decades, and likewise it has long been recognized that despite the nominal reference to a friend of the court, the principal point of almost all such briefs has been to support one of the parties.[16] Recent Terms have been characterized by the large and increasing number of such briefs filed at both the certiorari and merits stages. As recently as 1979, only half of the Court’s decisions on the merits had at least one amicus brief filed.[17] That percentage steadily rose, and in recent years almost 95% of the cases had at least one (and often many more than that) amicus brief filed.[18] The increase is particularly noticeable given (and perhaps not unrelated to) the Court’s steadily diminishing docket over the same period.[19] In past decades, progressive interest groups like the ACLU or the NAACP were prominent in filing amicus briefs. More recently, conservative groups like the U.S. Chamber of Commerce have frequently filed briefs.[20]

Of particular note has been the frequent amicus activity of the Solicitor General of the United States (“SG”), and of state attorneys general (“SAG”). Per the Supreme Court’s rules,[21] the SG and SAGs are the only groups that may file the briefs without the Court’s or the litigants’ permission.[22] Since the 1950s the SG has filed briefs in many cases, at both the certiorari and merits stages, including uniquely by invitation of the Court itself.[23] The amicus briefs filed by the SG have come to earn an excellent and by all accounts highly deserved reputation for legal probity and helpfulness. The Court frequently agrees with the SG’s recommendation, usually at the 75% level in any given Term, if not higher.[24] In a somewhat parallel fashion, the SAGs have come to enjoy a noticeable presence in amicus briefs, also highly valued by the Court. SAGs often file the briefs in cases involving states and federalism issues, and have found particular, recent success by joining together in large groups that constitute a supermajority of the states.[25]

Whether and to what extent this amicus activity has had an effect on Supreme Court decision making has been a matter of controversy.[26] Some Justices and their clerks suggest that on the whole the briefs are little read and have little effect, with rare exceptions like those filed by the SG.[27] More recent research, however, undermines this skepticism.[28] Subverting the thesis of the modest impact of amicus briefs is not simply based on the apparent success of some amici (again, like the SG) in convincing the Court to reach a particular result. Correlation is not causation, and the causal arrow may run in the other direction; the SG or interest groups may be inclined to file briefs in cases where they predict the Court will rule in what they deem a favorable way. That said, “not all interest groups are equal” when it comes to amicus briefs, and the Court appears to more likely to favor amici that have earned a reputation for submitting briefs of high quality, or groups that collaborate in filing with other groups.[29] Individual Justices frequently refer to the arguments of amici in oral arguments, and often cite amicus briefs in their opinions.[30] Indeed, Professor Larsen has documented that the Court frequently relies on amicus briefs for supportive facts, referenced in opinions, which are not found in the record.[31] More than that, political scientists (using plagiarism-detection software) have shown that the Justices often borrow language from amicus briefs, especially from those filed by the SG, SAGs, and such “elite” interest groups like the ACLU and the U.S. Chamber of Commerce.[32]

As a general matter, why have so many amicus briefs been filed in recent years? Outside of the Court, it is often said that interest groups have recently engaged in more lobbying (for example, before Congress or the executive branch),[33] and amicus activity may simply be a reflection of that.[34] Such groups may perceive that Court decisions are particularly consequential, especially in an era of party polarization and relative gridlock in the other branches of the federal government. The relatively smaller docket of the Court may accentuate the perceived importance of the fewer cases that are decided.[35] There appear to be synergistic effects between the Court seemingly relying on and citing amicus briefs more often, and interest groups filing those briefs.[36] Such groups may also file the briefs to solicit membership and funds, and to justify their existence. The increase may reflect an “arms race” among interest groups.[37]

In their new article, Larsen and Devins summarize and extend the empirical study of amicus briefs filed in the Supreme Court. They observe that the conventional wisdom was that most amicus briefs were filed in an ad hoc, uncoordinated manner by interest groups and other entities that became aware of a case on the Court’s docket and wanted to press their policy positions before the Court.[38] This lobbying was more or less similar to any other lobbying activity by these groups.[39] There were only anecdotal accounts of the litigants themselves lobbying interest groups to file amicus briefs in a coordinated way on their behalf.[40] Larsen and Devins considerably change this picture. Going beyond anecdotal evidence, they interviewed twenty-six lawyers who have collectively argued over 400 cases before the Court, and have written or coordinated the filing of thousands of amicus briefs.[41]

What accounts for the recent rise[42] of this amicus machine? The authors point to several factors. One is the rise of an elite Bar, a relatively small group of private lawyers who often argue the smaller number of cases decided each Term by the Court. This smaller, sophisticated group of lawyers has in turn made recruitment of amici a regular part of their practice.[43] This is accentuated, as noted before, by the SG and SAGs who separately have regularly appeared as amici. Another factor is “the Court’s new hunger for information outside the record,” with amici helpfully satisfying that hunger.[44] Finally, the Court “itself embraces the work of the amicus machine,”[45] since the Court seems to prefer the work of a specialized Bar (for both litigants and amici) that makes it more comfortable in marshaling arguments and facts and “facilitates the declaration of broad legal rules rather than resolving narrow disputes.”[46]

Moving from a descriptive to an evaluative mode, Larsen and Devins argue that the rise of the amicus machine is on balance a good thing. They contend that the machine alters the role of the SG and appropriately dilutes the influence of that office in favor of a larger group of lawyers; assists the Justices (and their clerks) at the certiorari stage by supplying information about the importance and consequential nature (or lack thereof) of cases where review is sought; and helps the Court to take care in pronouncing broad legal principles.[47] The authors are not oblivious to the potential downsides of the machine. They observe that it is elitist and subject to capture by interest groups, or a specialized Bar.[48] But on balance they pronounce it a good thing, as it aids the Court to deal rationally with massive amounts of information in deciding what cases to decide, and then deciding (and explaining the rationale for) those cases.[49]

II. The Problematic Impact of the Amicus Machine

To a degree, evaluating Larsen and Devins’s careful embrace of the amicus machine covers well-trod ground. For decades, various observers have defended amicus briefs as providing helpful facts that aid federal judges in making sound decisions, and appropriate democratic input into an unelected branch of government.[50] In contrast, some critics argue that amicus briefs, especially in large numbers, are inappropriate lobbying tools that improperly undermine the traditional adversarial process.[51] The Justices themselves seem ambivalent, sometimes claiming not to read many of them yet citing them when they seem to be useful.[52] For example, Chief Justice John Roberts recently remarked that amicus briefs can be “great” if they help the Court understand technological issues, or are devoted to the history of a particular constitutional provision at issue in a case. On the other hand, he continued, such briefs are “less helpful” if they are “me, too,” and simply “give you the same legal analysis you get in the party’s brief . . . . so they can say they won a case when it comes out their way.”[53]

As Larsen and Devins point out, analysis of the proper use by the Court of amicus briefs often begins with an appraisal of the proper role of federal courts in resolving legal controversies.[54] One contender is the dispute resolution model, which suggests that judges should focus on the resolution of concrete disputes by parties directly affected by the outcome and use traditional arguments and facts assembled by those parties in the record, all to avoid trampling on the prerogatives of the other branches of government. In contrast, the law declaration model posits that federal courts do and should possess the ability to, in the context of a particular case, announce broad principles of law, and if necessary go outside the strict record prepared by the parties in the case.[55]

To be sure, both models are “stylized and oversimplified” and cannot “capture the full historical or functional complexity of the role of the federal judiciary.”[56] Larsen and Devins are in my view correct in not finding it necessary to embrace either model, but they are also right in arguing that the Roberts Court’s apparent embrace of the amicus machine is a de facto endorsement of the law declaration model (at least on this issue).[57]

Recall that Larsen and Devins are largely not offended by that embrace, and suggest several reasons why on balance it is appropriate for the Court to increasingly rely on the large number of amicus briefs.[58] They argue that judicial reliance on amicus briefs alters, for the good, the role of the SG as an amicus, aids the Justices (and their clerks) in evaluating large numbers of certiorari petitions, and helps the Court in its law declaration mode.[59] Their reasons are largely instrumental, to make the Court function better (as they see it) as a judicial institution, as opposed to arguments grounded on political theory, that amicus briefs make federal courts more democratic.

I do not have a deep quarrel with much of their evaluation. Indeed, I particularly agree with their critique of the SG as an amicus. They point out, as many have, the high and deserved reputation of the SG as amicus at both the certiorari and merits stages.[60] But they argue that many of the practitioners who frequently argue before the Court (and recruit and file amicus briefs themselves) are alumni of the Office of the SG, and have reputational interests to defend. So the Court can be confident in relying on their briefs as much as those filed by the SG.[61] I agree for somewhat different reasons. I too find the SG’s amicus briefs to be of high quality and often justifiably relied on by the Court. Where the United States is not a party, but the federal interest is apparent (such as the application of federal statutes that the executive branch also enforces), then the filing of and reliance on SG amicus briefs is appropriate.[62] But in my view the SG is sometimes too influential, and has filed amicus briefs where there is no serious federal (or executive branch) interest to argue for. Where the federal interest is at best attenuated, the SG acting as an amicus is doing little more than making a political statement to the Court on the desirability of a certain result.[63] So for these reasons, the dilution of the impact of the SG as amicus, if only at the margins, is a good thing overall.

That said, I wonder if the additional information provided by all amici, touted by Larsen and Devins, is an unalloyed positive. Interestingly, the authors themselves have previously worried about the Court relying on amicus briefs for facts outside the record assembled by the parties,[64] or for other reasons.[65] They acknowledge their previous reticence in their new article, but suggest that the benefits of the amicus machine, as they see it, can mitigate those concerns.[66]

For several reasons, I am not as confident with that conclusion. First, consider the effect of amicus briefs within the Court. In principle I agree that more information and reasoned arguments is better than less for decision makers, and that the Court can benefit from amicus briefs that supply that information.[67] This can be particularly true for certain topics where the Court, a tribunal of general jurisdiction, lacks expertise (for example, intellectual property or legal history), or has traditionally deferred to some degree (without abdicating) to the views of the other branches (for example, foreign affairs). The Court itself has acknowledged this in their opinions.[68] Moreover, it is an overstatement to claim that the Court has been captured by amici. While some amici appear to be more influential than others, the Court overall seems to read (or skim) and evaluate most such briefs with a discerning eye. That is, the Court seems to give more or less weight to amici based on the relative expertise of the attorneys writing the brief;[69] the composition of a large number of entities that may join in one brief;[70] or whether the amicus brief is taking a counterintuitive, and thus perhaps a more credible, position.[71]

The problem is not one of kind, but of degree. As I have already mentioned, Professor Larsen has demonstrated that the Court’s opinions selectively use facts from amici outside the record, and others have shown that the Court frequently borrows language from such briefs. So comfortable are the Justices with amicus briefs that they sometimes openly worry when such briefs are not filed.[72] Maybe they are too comfortable. The Court can be the victim of too much information. At some point the Justices (and no doubt their clerks) are inundated in many cases with so many amicus briefs that they tune out most of the briefs entirely, or rely on reputational signals to pay attention to or credit a few briefs (that is, give particular attention to, say, the SG’s amicus briefs). Chief Justice Roberts implied they can easily ignore the “me, too” briefs, but that category may be in the eye of the beholder, and presumably someone (a Justice or a clerk) needs to examine a brief to determine what type it is (helpful or not helpful).[73]

Next consider the impact of the amicus machine on the public’s perception of the Court. Whether and to what extent the Court is or should be a “political” institution has been the subject of innumerable pages of commentary, and is a topic far beyond the scope of this Essay. But it is fair to say, I think, that most observers conclude that the Court (indeed, any court) should so far as is humanly possible not be overtly “political” (however you define that term) in its decision making. Most recently, Chief Justice Roberts has argued that it is a mistake to treat the Court as a political institution, like any other.[74] Even majorities of jaded lawyers in surveys will say that the Court is a legitimate institution, one that is “at least somewhat political and ideological in the manner in which it renders decisions,” but is usually not “activist or as overly influenced by external political actors.”[75] Perhaps this is a reason that polls usually show the Court has a better reputation and greater legitimacy than the other branches of the federal government.[76]

So far as I know, there are no studies examining whether the increased (or any) number of amicus briefs filed in the Court affects its perception and approval (and legitimacy) by the public in general, or by legal elites in particular. But I wonder and worry, if only to some small degree, that the amicus machine, especially if it sustains its operation and becomes more widely known, would degrade the standing and legitimacy of the Court. If it were widely known that interest groups routinely file many amicus briefs before the Court,[77] and they seemingly have an impact in (some) cases, it would seem to resemble the sort of lobbying that occurs within the other branches of government.[78] The Court would then indeed seem to be just another political institution. I imagine many readers of this Essay would find this conclusion entirely unremarkable and might applaud its candor. I am not sure all would, though, and I think (though cannot prove) that sustained, open, and routine lobbying of the Court would sooner or later seriously erode the Court’s support and legitimacy both in the public and among legal elites.[79]

III. Retooling the Amicus Machine

Assuming one has at least some concern with the amicus machine what, if anything, should be done about it? Reforming amicus activity in the Supreme Court is also well-trodden ground, and I will summarize some of those ideas and sketch a few of my own. Here the goal is to steer a path between doing nothing and simply banning all or most such briefs.

In previous work, Larsen, focusing on amicus briefs supplying extra-record factual information, has suggested that such briefs be subject to notice and comment requirements.[80] At least that would to a degree replicate an adversarial proceeding, as compared to the present state where an amicus brief can be filed, with no response, and then relied upon by the Court. Rather than limited to certain categories of briefs, this requirement could apply to all amicus briefs. Similarly, the Court could more rigorously enforce requirements that amici disclose their financial backers in their briefs.[81]

Another step would be to limit the number of amicus briefs. Again, Larsen has suggested that the parties could be permitted to select a small number of “their” amici, presenting extra-record factual arguments.[82] Why not extend this to all amicus briefs? Each party (at both the certiorari and merits stages) could be permitted five (or perhaps ten, if five strikes you as too low) amicus briefs to be filed on their behalf. This would prevent the numerous “me, too” briefs from inundating the Court and hijacking the Court’s docket for the amici’s publicity purposes. (It should go without saying that interest groups and others, excluded by this proposal, could still proclaim their arguments in any manner they wish, outside of litigation.) I would exclude the SG and SAGs from this limit. The Court for decades has excluded them from the need-for-permission requirement, and by and large they have earned the Court’s trust by usually filing excellent and helpful briefs.

Finally, the Court could revisit the explicit deference, if any, it gives any amicus brief. The Court has been at best inconsistent in the deference (as revealed by that word or its synonyms in opinions) that it gives to amicus briefs filed by the SG[83] or SAGs,[84] and others. More recently, it seems that the Court has not been explicitly giving deference to any amici as much as in the past, and is treating all amici in opinions on an equal footing (which isn’t to gainsay the apparent influence of the large number of amici in general, and the SG in particular). Perhaps this is due to the recent concern of individual Justices, and the Court as a whole, with the deference (if any) due federal agencies on legal issues.[85] In any event, the Court could confine explicit deference to narrow categories,[86] or simply not give any deference at all.


In their new article, Professors Larsen and Devins add to their prior important contributions to the literature on the increasing filings and apparent influence of amicus curiae briefs in the Supreme Court. Their article documents how the amicus machine is now characterized by the lawyers for many litigants proactively assembling amicus briefs to be filed on their behalf. They provocatively argue that this activity is largely a good thing by providing important information to the Court. In this Essay I have questioned some aspects of their latter conclusion, and instead suggest that the Court limit the large number of filings of such briefs and take greater care in its use of the amicus briefs that are filed. I concede that such restrictions might keep some potentially valuable information and legal arguments from the Court. But that is likely to have a marginal impact since the parties themselves and the remaining amici can provide what is needed. The potential upside of restrictions may dampen the influence of large numbers of amicus briefs, make the Court less politicized, and appropriately refresh the adversarial system.


[1]I of course refer to the death of Justice Antonin Scalia in February 2016. The Court issued only sixty-seven decisions on the merits. Kimberly Strawbridge Robinson, SCOTUS by the Numbers: Odd Votes, New Roles, 84 U.S. L. Wkly. 1938, 1938 (2016).

[2]For examples of some high-profile cases, see Whole Woman’s Health v. Hellerstedt, 136 S. Ct. 2292 (2016) (82 amicus briefs filed at merits stage); United States v. Texas, 136 S. Ct. 2271 (2016) (per curiam) (41 amicus briefs filed at merits stage); and Fisher v. University of Texas, 136 S. Ct. 2198 (2016) (85 amicus briefs filed at merits stage). For data on these cases and on the Court’s docket as a whole, see Anthony J. Franze & R. Reeves Anderson, In Unusual Term, Big Year for Amicus Curiae at the Supreme Court, Nat’l L.J., Sept. 26, 2016 (discussing data from 2011 through 2016 Terms); Adam Feldman, Inferences From Amicus Briefs and How Justice Kennedy Continues to Rule Supreme, Empirical SCOTUS (July 13, 2016), [https://pe‌rm‌].

[3]See Kimberly Strawbridge Robinson, Some Supreme Court ‘Friends’ Are Better Than Others, 84 U.S. L. Wkly. 1689, 1690 (2016) [hereinafter Robinson, Some Friends] (quoting Adam Feldman that “between 700 and 900 amicus briefs [were filed] in the early years of the Roberts Court,” while “1300-1500 briefs [were] filed in the last few years.”); Franze & Anderson, supra note 2.

[4]And apparently in lower courts, too; see, for example, Paul M. Collins, Jr. & Wendy L. Martinek, Judges and Friends: The Influence of Amici Curiae on U.S. Court of Appeals Judges, 43 Am. Pol. Res. 255, 257 (2015), though my focus in this Essay will be on the U.S. Supreme Court.

[5]See infra Part I.

[6]See, e.g., Linda Greenhouse, The Roberts Court’s Reality Check, N.Y. Times (June 25, 2015), [ht‌tps://]; Adam Liptak, Study Shows How Much Work It Takes to Be Supreme Court’s Friend, N.Y. Times Sidebar (Mar. 7, 2016), http://www.ny‌time‌s.c‌o‌m/2016/03/08/us/politics/study-shows-how-much-work-it-takes-to-be-supreme-courts-fri‌en‌d.html []; Noah Feldman, The Dark Side of All Those ‘Friends’ at the Supreme Court, Bloomberg View (Mar. 9, 2016, 9:11 AM),‌o‌‌em‌e‌-court [].

[7]136 S. Ct. 2292, 2300 (2016).


[9]Id. at 2312, 2315. The Court also cited an amicus brief of law professors to support its conclusion, earlier in the decision, that the suit was not barred by res judicata. Id. at 2309.

[10]Id. at 2320–21 (Ginsburg, J., concurring). The dissents by Justices Clarence Thomas and Samuel Alito (the latter of which was joined by Chief Justice John Roberts and Justice Thomas) did not cite any amicus briefs.

[11]Helen A. Anderson, Frenemies of the Court: The Many Faces of Amicus Curiae, 49 U. Rich. L. Rev. 361, 364 (2015).

[12]For a sample of important contributions to the literature on amicus briefs, see, for example, Paul M. Collins, Jr., Friends of the Supreme Court: Interest Groups and Judicial Decision Making (2008) (political scientist); Stephen M. Shapiro et al., Supreme Court Practice 749–60 (10th ed. 2013) (Supreme Court practitioners); Joseph D. Kearney & Thomas W. Merrill, The Influence of Amicus Briefs on the Supreme Court, 148 U. Pa. L. Rev. 743 (2000) (law professors). For a general overview of the literature, see Anderson, supra note 11, at 364–65.

[13]Allison Orr Larsen & Neal Devins, The Amicus Machine, 102 Va. L. Rev. 1901 (2016).

[14]Id. at 1903–05.

[15]Id. at 1906–08.

[16]The classic source documenting the trend is Samuel Krislov, The Amicus Curiae Brief: From Friendship to Advocacy, 72 Yale L.J. 694 (1963).

[17]See Lee Epstein et al., The Supreme Court Compendium: Data, Decisions, and Developments 721 (5th ed. 2012) (providing data from the 1946 to 2001 Terms).

[18]See id.; Franze & Anderson, supra note 2 (discussing more recent data); Robinson, Some Friends, supra note 3 (same); Feldman, supra note 2 (same).

[19]See Ryan J. Owens & David A. Simon, Explaining the Supreme Court’s Shrinking Docket, 53 Wm. & Mary L. Rev. 1219, 1228–29 (2012).

[20]Collins, supra note 12, at 50–56 (providing analysis of interest groups as amici over several decades); David M. O’Brien, Storm Center: The Supreme Court in American Politics 229–31 (10th ed. 2014) (same).

[21]Sup. Ct. R. 37.

[22]That distinction has come to be without a difference, since with the rarest of exceptions the Court and litigants routinely grant permission for all such briefs to be filed. Shapiro et al., supra note 12, at 516–17.

[23]Michael E. Solimine, The Solicitor General Unbound: Amicus Curiae Activism and Deference in the Supreme Court, 45 Ariz. St. L.J. 1183, 1186 (2013) [hereinafter Solimine, Solicitor General].

[24]See Ryan C. Black & Ryan J. Owens, The Solicitor General and the United States Supreme Court: Executive Branch Influence and Judicial Decisions 24–28 (2012); Solimine, Solicitor General, supra note 23, at 1185–94.

[25]See generally Margaret H. Lemos & Kevin M. Quinn, Litigating State Interests: Attorneys General as Amici, 90 N.Y.U. L. Rev. 1229, 1231–35 (2015) (analyzing partisanship in SAG amici briefs); Michael E. Solimine, State Amici, Collective Action, and the Development of Federalism Doctrine, 46 Ga. L. Rev. 355, 357–62 (2012) [hereinafter Solimine, State Amici] (discussing how much weight the Court should give SAG amici briefs in various contexts).

[26]A.E. Dick Howard, The Changing Face of the Supreme Court, 101 Va. L. Rev. 231, 274 (2015).


[28]Lee Epstein, Some Thoughts on the Study of Judicial Behavior, 57 Wm. & Mary L. Rev. 2017, 2066 (2016).

[29]Janet M. Box-Steffensmeier, Dino P. Christenson & Matthew P. Hitt, Quality over Quantity: Amici Influence and Judicial Decision Making, 107 Am. Pol. Sci. Rev. 446, 458 (2013).

[30]For examples of references to amicus briefs in oral arguments, see Larsen & Devins, supra note 13, at 1954–56. For citation rates, see O’Brien, supra note 20, at 230–31.

[31]Allison Orr Larsen, The Trouble with Amicus Facts, 100 Va. L. Rev. 1757, 1758–59 (2014) (study of Court decisions from 2008 to 2013 that cited amicus briefs).

[32]Paul M. Collins, Jr., Pamela C. Corley & Jesse Hamner, The Influence of Amicus Curiae Briefs on U.S. Supreme Court Opinion Content, 49 Law & Soc’y Rev. 917, 936–37 (2015). While this study used plagiarism software, it did not attempt to determine how often the borrowed language was or was not accompanied by citations to the briefs. Id. at 928 n.5.

[33]See Lawrence Lessig, Republic, Lost: The Corruption of Equality and the Steps to End It 104 (2d ed. 2015); Joseph P. Tomain, Gridlock, Lobbying, and Democracy, Wake Forest J.L. & Pol’y (forthcoming) (manuscript at 10–11) (on file with author); Thomas B. Edsall, The Lobbying Bonanza, N.Y. Times (June 10, 2015), http://www.nytime‌‌/1‌0/opinion/the-lobbying-bonanza [].

[34]Cf. Katie Zuber, Udi Summer & Jonathan Parent, Setting the Agenda of the United States Supreme Court? Organized Interests and the Decision to File an Amicus Curiae Brief at Cert, 36 Just. Sys. J. 119, 130–31 (2015) (finding interest groups are more likely to file amicus briefs at the certiorari stage when the SG has been requested to file an amicus brief, and when the case has received attention in the national media). Amicus briefs are not cheap; some suggest that a typical amicus brief costs about $50,000. Id. at 121.

[35]See Owens & Simon, supra note 19, at 1228–29.

[36]Thomas G. Hansford & Kristen Johnson, The Supply of Amicus Curiae Briefs in the Market for Information at the U.S. Supreme Court, 35 Just. Sys. J. 362, 363–64 (2014).

[37]Howard, supra note 26, at 274.

[38]Larsen & Devins, supra note 13, at 1910–14.

[39]Id. at 1910–11.

[40]See, e.g., Kristen E. Eichensehr, Foreign Sovereigns as Friends of the Court, 102 Va. L. Rev. 289, 304 (2016) (arguing participation by foreign sovereigns as amici “may be the result of recruitment by the parties they support, particularly if such parties are represented by experienced Supreme Court practitioners, who often coordinate amici support for their clients”); Patricia A. Millett, “We’re Your Government and We’re Here to Help”: Obtaining Amicus Support From the Federal Government in Supreme Court Cases, 10 J. App. Prac. & Process 209, 222–26 (2009).

[41]Larsen & Devins, supra note 13, at 1915.

[42]Larsen and Devins point out earlier cases where there was evidence of amici recruitment by attorneys for the litigants, id. at 1904–05 (discussing Grutter v. Bollinger, 539 U.S. 306 (2003)), but their main focus is on more recent cases like King v. Burwell, 135 S. Ct. 2480 (2015), suggesting that the machine they describe is a relatively recent phenomenon.

[43]Id. at 1904–05, 1919–31.

[44]Id. at 1906.

[45]Id. at 1907.


[47]Id. at 1908, 1940–57.

[48]Id. at 1908, 1957–58.

[49]Id. at 1908, 1958–65. The leading treatise for practitioners also approves of the coordination described by Larsen and Devins. After pointing out that the Supreme Court’s rules neither discourage nor require disclosure of the practice, it adds that “[o]ften some form of consultation and communication is both appropriate and essential if the amicus brief is to be confined, [as Rule 37.1 states], to ‘relevant matter not already brought to [the Court’s] attention by the parties.’” Shapiro et al., supra note 12, at 755 (third alteration in original).

[50]See, e.g., Anderson, supra note 11, at 361–62 (summarizing praise).

[51]See, e.g., id. at 365–66 (summarizing criticisms).

[52]Shapiro et al., supra note 12, at 757–58.

[53]Chief Justice John Roberts, Remarks at Fourth Circuit Judicial Conference (May 25, 2016), []. In his remarks, Chief Justice Roberts gave as an example of helpful amicus briefs those that explain how to apply the Fourth Amendment to searches of information on iPhones. He was likely alluding to his opinion for the Court in Riley v. California, where he cited five amicus briefs. 134 S. Ct. 2473, 2486–90 (2014).

[54]Larsen & Devins, supra note 13, at 1908, 1952–54.

[55]For overviews and critiques of the models, see Richard H. Fallon, Jr. et al., Hart & Wechsler’s The Federal Courts and the Federal System 73–76 (7th ed. 2015); Neal Devins & Saikrishna B. Prakash, Essay, Reverse Advisory Opinions, 80 U. Chi. L. Rev. 859, 862–63 (2013); Henry Paul Monaghan, Essay, On Avoiding Avoidance, Agenda Control, and Related Matters, 112 Colum. L. Rev. 665, 665–69 (2012).

[56]Fallon et al., supra note 55, at 75.

[57]Larsen & Devins, supra note 13, at 1953. For similar conclusions, see Anderson, supra note 11, at 409–11; Solimine, Solicitor General, supra note 23, at 1204.

[58]They are not alone in speaking favorably of the Court’s use of amicus briefs, or even in advocating more use. See, e.g., Andrew Manuel Crespo, Regaining Perspective: Constitutional Criminal Adjudication in the U.S. Supreme Court, 100 Minn. L. Rev. 1985, 2023–26 (2016) (arguing that the Court should invite organizations representing interests of criminal defendants to file amicus briefs to counter the expertise of the SG); Abbe R. Gluck, Comment, Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62, 101 (2015) (asking whether the Court should give “a heightened role [to] amici” as it is called upon to interpret statutes that are the product of an increasingly complicated and unusual legislative process).

[59]See supra note 47 and accompanying text.

[60]Larsen & Devins, supra note 13, at 1941–42.

[61]Id. at 1943–44.

[62]Solimine, Solicitor General, supra note 23, at 1198–201.

[63]Id. at 1203–08. I have previously argued that for these reasons the SG should not have filed amicus briefs in cases involving state abortion restrictions and state same-sex marriage bans. Id. at 1196–99, 1207–10. In my view, this was true for some recent amicus briefs on those cases by the Obama administration. The putative federal interest, such as it is, can be gleaned from the required “Statement of Interests” section of any amicus brief. Thus, in the same-sex marriage case, the SG’s amicus brief stated little more than that “[t]he United States has a strong interest in the eradication of discrimination on the basis of sexual orientation.” Brief for the United States as Amicus Curiae Supporting Petitioners at 2, Obergefell v. Hodges, 135 S. Ct. 2584 (2015) (Nos. 14-556, 14-562, 14-571, and 14-574), 2015 WL 1004710, at *2. Similarly, in the abortion case, the SG stated (correctly) that prior SGs had filed amicus briefs in similar cases involving state laws, and that Congress “has enacted laws relating to abortion, and may legislate further in that area.” Brief for the United States as Amicus Curiae Supporting Reversal at 1–2, Whole Woman’s Health v. Hellerstedt, 136 S. Ct. 2292 (2016) (No. 15-274), 2016 WL 67681, at *1–2 (footnote omitted). Keeping in mind that the constitutionality of state laws was at issue in these cases, it is very unclear what, precisely, the interest of the federal government or the SG is in the resolution of that question. It is no wonder that such briefs are interpreted as being mere political statements. See Jess Bravin, Obama Administration Opposes Texas Abortion Restrictions, Wall St. J.: L. Blog (Jan. 4, 2016, 8:21 PM), [] (“In a politically charged issue such as abortion, both Republican and Democratic administrations have felt compelled to weigh in, although in opposite directions.”). Interestingly, the SG’s amicus briefs were not cited by any of the opinions in Obergefell or Whole Woman’s Health.

[64]Larsen, supra note 31, at 1758.

[65]Devins & Prakash, supra note 55, at 885 (arguing that troubling questions are raised by the Court routinely requesting the SG to file amicus briefs giving advice on how the Court should rule on writs of certiorari, and the Court frequently (though not always) following that advice).

[66]Larsen & Devins, supra note 13, at 1944–46.

[67]See Frederick Schauer, Our Informationally Disabled Courts, 143 Dædalus, J. Am. Acad. Arts & Sci. 105, 105–07 (2014).

[68]For an example involving both legal history and foreign affairs, see Zivotofsky v. Kerry, 135 S. Ct. 2076, 2091 (2015) (“The briefs of the parties and amici, which have been of considerable assistance to the Court, give a more complete account of the relevant history . . . .”).

[69]See John Szmer & Martha Humphries Ginn, Examining the Effects of Information, Attorney Capability, and Amicus Participation on U.S. Supreme Court Decision Making, 42 Am. Pol. Res. 441, 442–43 (2014).

[70]Greg Goelzhauser & Nicole Vouvalis, Amicus Coalition Heterogeneity and Signaling Credibility in Supreme Court Agenda Setting, 45 Publius 99, 100 (2015) (arguing that the Court is more likely to follow an amicus brief at the certiorari stage that was jointly filed by ideologically different states); Karen Swenson, Amicus Curiae Briefs and the U.S. Supreme Court: When Liberal and Conservative Groups Support the Same Party, 37 Just. Sys. J. 135, 138 (2016) (arguing that the Court is more likely to follow amicus briefs filed on behalf of both liberal and conservative groups).

[71]Solimine, State Amici, supra note 25, at 379.

[72]Consider such recent examples as Justice Elena Kagan asking during one oral argument why a party did not have any amicus briefs filed on its behalf, Transcript of Oral Argument at 53–54, Bullard v. Blue Hills Bank, 135 S. Ct. 1686 (2015) (No. 14-116), or Justice Breyer lamenting in a concurring opinion that the SG had not filed an amicus brief in J. McIntyre Machinery, Ltd. v. Nicastro, 564 U.S. 873, 893 (2011) (Breyer, J., concurring in the judgment).

[73]Cf. Paul M. Collins, Jr., Pamela C. Corley & Jesse Hamner, Me Too? An Investigation of Repetition in U.S. Supreme Court Amicus Briefs, 97 Judicature 228, 234 (2014) (arguing that plagiarism software shows that language in most amicus briefs is not repetitious of language in other information sources, such as lower court opinions, or the briefs of the parties or other amici).

[74]Robert Barnes, The Political Wars Damage Public Perception of Supreme Court, Chief Justice Roberts Says, Wash. Post (Feb. 4, 2016), https://www.washingto‌‌s/c‌ourts_law/the-political-wars-damage-public-perception-of-supreme-court-chief-justice-ro‌b‌e‌r‌ts-says/2016/02/04/80e718b6-cb0c-11e5-a7b2-5a2f824b02c9_story.html [https://per‌ma‌.c‌c/D‌2YY-74F7]. Lest the reader think that this concern is restricted to Chief Justice Roberts, consider the widespread negative reaction to the comments of Justice Ginsburg, criticizing Donald Trump during the 2016 presidential campaign. Michael D. Shear & Maggie Haberman, Donald Trump Calls Ruth Bader Ginsburg’s Remarks a ‘Disgrace to the Court,’ N.Y. Times (July 12, 2016),‌ld-trump-criticism.html [].

[75]Brandon L. Bartels, Christopher D. Johnston & Alyx Mark, Lawyers’ Perceptions of the U.S. Supreme Court: Is the Court a “Political” Institution?, 49 Law & Soc’y Rev. 761, 789–90 (2015).

[76]For a summary of the considerable literature on public (that is, non-lawyer) perception of the Court, see id. at 761–63.

[77]Interest groups have not hesitated to tout their amicus filings in the Court on their websites. For examples, see Solimine, State Amici, supra note 25, at 384 n.132; Zuber et al., supra note 34, at 126.

[78]It is interesting that some lawyers who admit that they assemble amici in support of their clients insist on remaining anonymous. Eichensehr, supra note 40, at 304 n.68. Perhaps this is due to their taking part in the arguable violation of anti-lobbying norms for American courts.

[79]In a similar fashion, Professor Devins has previously criticized the Court’s practice of calling for the views of the SG via amicus brief, in part because it would be akin to the Court routinely and openly seeking advice from the Chamber of Commerce in business cases or the ACLU in First Amendment cases. Devins & Prakash, supra note 55, at 885. Professors Devins and Prakash were focusing on SG amicus briefs requested by the Court, but their concerns are applicable to all amicus briefs, whether requested or not.

[80]Larsen, supra note 31, at 1812–15. For similar proposals, see Brianne J. Gorod, The Adversarial Myth: Appellate Court Extra-Record Factfinding, 61 Duke L.J. 1, 68–77 (2011); Rebecca Haw, Amicus Briefs and the Sherman Act: Why Antitrust Needs a New Deal, 89 Tex. L. Rev. 1247, 1284–87 (2011).

[81]Anderson, supra note 11, at 413.

[82]Larsen, supra note 31, at 1810.

[83]Solimine, Solicitor General, supra note 23, at 1212–14 (giving examples).

[84]Solimine, State Amici, supra note 25, at 359, 367–69, 395 (giving examples).

[85]See, e.g., Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2148 (2016) (Thomas, J., concurring) (arguing that the Court should revisit decisions where the Court gives explicit deference to federal agency interpretation of federal statutes or regulations, on the basis that such deference raises separation of powers concerns). For an overview of that debate, see Cass R. Sunstein & Adrian Vermeule, The New Coke: On the Plural Aims of Administrative Law, 2015 Sup. Ct. Rev. 41, 43. Sometimes the views of a federal agency are revealed in the SG’s amicus brief, if the agency is not a party to the suit. Solimine, Solicitor General, supra note 23, at 1215–16.

[86]See Solimine, Solicitor General, supra note 23, at 1217–22 (arguing that the Court should only give some deference to SG amicus briefs on foreign relations, and in cases where private parties are seeking to enforce federal law also enforceable by the executive branch). Cf. Eichensehr, supra note 40, at 296 (arguing that the Court should give the amici of foreign sovereigns greater weight than the SG on issues of foreign law). While the joining of many SAGs in one amicus brief is an ongoing phenomenon, see supra note 25 and accompanying text, I have argued that the Court should only give deference, if at all, to such briefs when they constitute a supermajority of the states, see Solimine, State Amici, supra note 25, at 391–93. Somewhat cutting the other way, more recently many such briefs have been characterized by SAGs of the same party joining in one amicus brief, sometimes producing dueling SAG amicus briefs in the same case. See Paul Nolette, Federalism on Trial: State Attorneys General and National Policymaking in Contemporary America 186–92 (2015). One reaction to this development might be to only give deference (if at all) to amicus briefs joined by significant numbers of SAGs from both political parties. In the 2015–16 Term, a unanimous Court seemed to go out of its way to do so in a high-profile case involving the extent to which federal law defines the bribery of state officials. McDonnell v. United States, 136 S. Ct. 2355, 2372 (2016) (noting that 77 former SAGs filed an amicus brief arguing for a narrow definition of the statute, consisting of “41 Democrats, 35 Republicans and 1 independent”). 

In Defense of the Secular Purpose Status Quo

The secular purpose rule, one prong of the Supreme Court’s interpretation of the Establishment Clause of the First Amendment, requires that government action be justified by a primary, genuine secular purpose. Government actions supported only by religious beliefs, therefore, are unconstitutional. A debate about the morality of the secular purpose rule has emerged, with the main arguments tending to view religious beliefs as either permissible or impermissible. This Note argues that rather than decide purely for or against the secular purpose rule, courts should maintain the current status quo, which is underenforcement of the rule.

To justify this approach to resolving the secular purpose debate, this Note analyzes common arguments made for and against the rule, and distills each argument to its core animating political value. The arguments against the secular purpose rule are motivated by the value of political access, while arguments for the secular purpose rule are motivated by the value of political legitimacy. Underenforcement creates equilibrium between these political values.

Some may worry that underenforcement will change the underlying meaning of the secular purpose rule. But a constitutional requirement can retain its full meaning and be legally binding even if underenforced. Another possible objection is that underenforcement would be tantamount to nonenforcement. To respond to that objection, this Note attempts to canalize underenforcement by marking out situations in which the secular purpose rule should be fully enforced. When, for example, underenforcement would allow discrimination against vulnerable groups, the secular purpose rule should be enforced.