Gauging “Ultra-Vires”: The Good Parts

Abstract

The Federal Constitutional Court’s ultra-vires case law—especially its most recent iterations—has more than its fair share of bad parts. It went from non-existence to global prominence in an alarmingly short period of time. Fortunately, the doctrine contains some extraordinarily good parts. Within the case law, there are three beautiful, elegant, and highly expressive elements that are buried under a massive tower of good intentions and hard luck: First, the principle of distinction between the two concepts of responsibility and accountability, second, the ban on transferring blanket empowerments, and third, the idea of a “program of integration” as a good medium for expressing vague ideas. In combination, these elements constitute a constitutional mechanism that does not play hell with European law, but truly complements any union based on multi-level cooperation. Focusing on the good parts—and avoiding some bad parts—might help prospective ultra-vires reviews to steer clear of wreaking havoc. The subset of good parts can serve to shift the constitutional case law towards reliability, readability, robustness, foreseeability, and, if nothing else, explicability.

“Setting the attractions of my good parts aside, I have no other charms.”

–William Shakespeare, The Merry Wives of Windsor (Falstaff, Act 2, Scene 2)

A. Subset and Stabilize!

The name is “ultra-vires.” That is what the Federal Constitutional Court (FCC) calls this form of review.¹ The where could most readily be described as the realm of “programs of integration” (Integrationsprogramme).² But there is a vast difference between being part of some program and being in charge. The what is easy. Recently, and for the first time, the FCC qualified a total of five ECB Governing Council decisions as ultra-vires acts,³ despite the CJEU’s judgment to the contrary.⁴ That is also the when. As for the why, beyond the obvious judicial protective motivation, it is exceedingly simple: The Basic Law prohibits conferring blanket empowerments for the exercise of public authority (Blankettermächtigung).⁵ Which leaves us only with the how. And therein, as the Bard would tell us, lies the rub.⁶

The recent PSPP judgment resolutely tells us that “the conditions under which the Federal Constitutional Court conducts an ultra-vires review are well-established.”⁷ This is true in as much as the name, the where, the what, and the why are settled. The how, at least, is not.

Despite all the care taken, the FCC has not yet put the how of its provocative ultra-vires review into concrete terms. That is not to say there are no concrete terms at all.⁸ Quite the contrary: Measured by volume, the FCC’s ultra-vires case law has prospered ever since its early evolutionary phase began in 1981.⁹ The Second Senate would learn about more aspects of gauging the ultra-vires character of public exercises of authority and would attempt to apply the new idea in high-profile cases such as the Pershing II weapon systems case (1984),¹⁰ the overseas deployment case (1994),¹¹ the case involving NATO’s new strategic concept (2001),¹² and the case involving the Tornado aircrafts in Afghanistan (2007).¹³ Notable parts of modern ultra-vires review also took shape in the Second Senate’s landmark decisions on the treaties of Maastricht (1993)¹⁴ and Lisbon (2009).¹⁵ Even if the salient backdrop to these European judgments was the delicate nature of Germany’s integration into the European project, the Maastricht and Lisbon cases nevertheless concerned challenges to acts of the German Parliament. Yet the Court meticulously elaborated on its ultra-vires review.

Eventually it became clear that some of those established case law attributes were ailing. Some of them were poorly specified and were more likely to cause application problems.¹⁶ Some resulted in new judgments that were difficult to read, which did not make it easier for the Second Senate to apply them—or so for seasoned members of the Second Senate, impossible to consent to.¹⁷ Some induced the Second Senate to write in a manner that was too tricky and error-prone. And some accrued aspects of the case law were design errors. The Court’s increasingly shaky ultra-vires jurisprudence got a second life on January 14, 2014 when six justices operationalized some version of ultra-vires review¹⁸ leading the two longest-serving justices to object in rare dissenting opinions.¹⁹ One dissenter insisted that “in an effort to secure the rule of law, the court happened to exceed its judicial competence.”²⁰ The OMT Referral case accomplished this much: In trying to more clearly gauge ultra-vires review, (at least) one faction had strayed.

Most lines of constitutional law development contain good parts and bad parts. It is hard for constitutional courts to remove imperfections from their past case law. Courts are usually powerless to do anything except to build upon those wobbly foundations by heaping more concrete terms on top of the existing pile of imperfections. And new terms do not always interact harmoniously, thus producing more bad parts. But constitutional courts have the power to define their own subsets. Lines of constitutional law development can improve over time by relying more and more exclusively on the good parts.

The Court’s ultra-vires case law has more than its fair share of bad parts. It went from non-existence to global prominence in an alarmingly short period of time. It never benefitted from calm weathers, not to mention an interval in the laboratory of compelling cases in which it could be tested and polished. It went straight to the sensational PSPP judgment of May 5, 2020²¹ just as it was. And it was very rough.

Fortunately, the Court’s ultra-vires doctrine has some extraordinarily good parts. Within the case law, there are beautiful, elegant, and highly expressive elements that are buried under a massive tower of good intentions and hard luck. The best qualities of constitutional ultra-vires review are so effectively hidden that, for many years, some people held the opinion that ultra-vires review was an unsightly, incompetent tool. My intention here is to reveal what’s right with ultra-vires review. It is a constitutional mechanism that does not play hell with European law, but truly complements any union based on multi-level cooperation. But this calls for chipping away at the massive case law block of marble in which this inspired sculpture is presently enshrouded. Removing all attributes that are neither elegant nor needed. Carving out a subset of the terms that is vastly superior to the case law as a whole, being more reliable, readable, and maintainable.

Focusing on the good parts—and avoiding bad parts—might help prospective ultra-vires reviews to steer clear of wreaking havoc. The subset can be used to shift the constitutional case law towards reliability, readability, robustness, foreseeability and, if nothing else, explicability. There is also a side benefit: Highlighting the good parts can eventually reduce the need for unlearning bad patterns. Sometimes case law is subsetted²² to oil the wheels for beginners. But at this juncture, subsetting constitutional case law makes it work better for professionals.

B. Responsibility and Accountability

The best thing about constitutional ultra-vires case law is its principle of distinction between the two concepts of responsibility and accountability. On this point it got almost everything right.

Each of these concepts account for a distinct role of participation in completing tasks for a process of integration. Responsibility and accountability are the fundamental units of gauging ultra-vires. They are used for clarifying and defining roles in unions based on multi-level cooperation. Roles are used to specify design decisions related to interaction among different participants. Generally, there is a difference between a role and individually identified institutions: A role describes an associated set of tasks; several institutions may perform the same role; and one institution can perform different roles.

The responsible institution is the one that does the work to complete a task. There is at least one role with a participation type of responsible, although other actors can be delegated to assist in the work required. The accountable institution is the one ultimately answerable for the correct and thorough completion of the task or process. That is to say: The accountable institution ensures that the prerequisites of the process are met, and it delegates the competences to take action to the responsible institutions. An accountable institution approves, what one (or more) responsible institution(s) provide(s). And there may be only one accountable institution specified for each task.

Institutions are not successively accountable or responsible, but at the same time. An accountable institution, having approved the actions of some responsible institutions in the past, nevertheless stays accountable over time. Approving the actions of some responsible institutions does not suspend the accountability by passing control to the responsible institution. Responsibility and accountability always travel in pairs. And they stay paired even if one institution performs both roles.

So, who is in charge? Who has the last say? After all, that is what ultra-vires reviews are all about? How do we solve this? It turns out, in a purely “responsible-or-accountable”-typed pattern, the interaction dispenses with hierarchies (although there might also be hierarchies for other reasons). We focus instead on the roles. A responsibility-and-accountability-pattern is conceptually simpler than classical hierarchies: An institution can be accountable without being in charge of the responsible institutions, without some kind of hierarchy, without sorting and shelving the accountable on top of the responsible institution. This is perhaps unfamiliar—and it might even sound “unfair” at first glance²³—but it is easy to understand. You simply start with one institution that delegates competences to another institution (creating accountability and responsibility at the same time). You can then create many more responsible institutions that are like the first one. The classification process of breaking the whole institutional architecture based on multi-level cooperation down into a set of nested hierarchies can be completely avoided. German legislative bodies are in no hierarchical relationship with any European institution. But they become the accountable institution (by approving the program of integration). And they remain accountable over time.

The most striking aspect of these heterarchical building blocks is that they act as a bridge between responsible European institutions that are approved by accountable legislative bodies (Bundestag and Bundesrat) and the ultra-vires review itself. The task of gauging ultra-vires, which determines whether an accountable institution has approved, what some responsible institutions do, is just another responsibility. In this setup the European institutions and the FCC are responsible institutions. The German parliamentary bodies, on the other hand, are the accountable one.

To appreciate this point is to change the popular image of ultra-vires reviews as a race hazard, forcing different institutions to compete for superiority all at one time. From the perspective of the responsible European institutions, the boundary of sovereign powers accorded to them is an expression of the delegation of competences. From the perspective of the responsible institution gauging ultra-vires, however, the same expression is simply a description of what the accountable institution approves and mandates.

Against this backdrop, the Lisbon judgment’s famous wording simply does not imply any claim for supremacy: “The Federal Constitutional Court examines whether legal instruments of the European institutions and bodies keep within the boundaries of the sovereign powers accorded to them by way of conferral.”²⁴

One weakness of this heterarchical pattern is hidden in the case law: So far, we get no boundaries; there is only “text”—some acts of Parliament. Sometimes that does not matter. But with regard to ultra-vires, it matters a lot. Playing fast and loose with a text can undermine any attempt to deliver dependable and predictable ultra-vires review. Here is the catch. And this is where the second elegant part of the ultra-vires case law comes to the rescue: The concept of “blanket empowerments.”

C. Keep it Classy: The Curious Case of “Blanket Empowerments”

The so-called “ban on transferring blanket empowerments”²⁵ is a lightweight concept at the inner core of the case law. It is based on parliamentary accountability, one of the case law’s best parts. Even though this ban was most noticeably conveyed in the high-profile cases relating to the acts of Parliament approving the treaties of Maastricht²⁶ and Lisbon,²⁷ it is self-sufficient. It can be used to communicate “boundaries” to responsible institutions. The ban insures against dissolution of boundaries. It is a treaty format, so it is readable by Parliament and readable by the Federal Constitutional Court. It is easy to implement and easy to use. But sadly, it also is easily mistaken. By avoiding the mistakes, the FCC makes ultra-vires a more reliable review.

The ban on blanket empowerments is a style quality tool for treaty law. It takes acts of Parliament approving treaties and scans them. If it finds a boundary, it returns a boundary by describing something that is not transferred yet. If it finds no boundary at all, the treaty is banned. Parliamentary approval would be unconstitutional because, after all, Parliament is accountable for the boundary. The problem with “treaties without boundaries” is not necessarily a design flaw within the draft, although it might be. Nor would it be clearly impossible to initiate neat processes of integration based on “treaties without boundaries,” though it might raise a few eyebrows. But Parliament could neither become nor stay accountable without having approved some boundary. From this perspective the “ban on blanket empowerments” in the ultra-vires doctrine serves as a corollary of parliamentary accountability. To undermine the ban is, for better or worse, to undermine parliamentary accountability. To put it another, more objective way, the concept of carte blanche and the constitutional requirement of parliamentary accountability are mutually exclusive.

Being a style quality tool, the ban on blanket empowerments looks at style conventions as well as a treaty’s structural problems.²⁸ Passing the “ban on blanket empowerments”-test, however, does not prove that some treaty’s program of integration is flawless or that the spawned process of integration would evolve as expected. Nothing could be further from the truth. The test may be a tool, but it is no cure-all. Passing the “ban on blanket empowerments”-test simply means that there is a boundary. That result on the test is an entirely different matter from some related messages, such as “there is a reasonable boundary,” or “there is a boundary clarifying every aspect of the competence,” or “there is a boundary illuming every crucial aspect of the competence,” or at least “there is a boundary delineating precisely those areas, some stakeholders are eager to mark-off.”

It should also go without saying that passing the “ban on blanket empowerments”-test does not imply a consensus about “methodological deficits” or “appropriate methodology” with regard to verifiably rendering every coming judgment tenable or untenable. Yet, casually mixing both concepts is more topical than ever.²⁹ The logical implication at hand is an unambiguous one-way street. A treaty consensus on appropriate methodology would most definitely provide for some kind of boundary (or more specifically: a boundary between tenable and untenable judgments). But one cannot necessarily turn the idea around: A treaty consensus on some kind of boundary does not necessarily imply some kind of appropriate methodology. For example, if there is a treaty boundary between monetary unions as “stability communities” and monetary unions as other kinds of communities (that is not even a big if, because the FCC has repeated that boundary like a mantra,³⁰ yet still no one has ever contradicted it), then we do not necessarily know anything about judicial methodology at all (maybe arguably: It implies any methodology, as long as no judgments are rendered saying welcome to “monetary unions as unstable communities”).

Treaty compliance with the “ban on blanket empowerments” decorates (or, more technically speaking, compliance configurates) the ultra-vires review: The boundary approved by the accountable parliamentary bodies is identical in content to the boundary safe-guarded by the responsible ultra-vires reviewer. Any responsible ultra-vires reviewer faces exactly the same challenges as the accountable parliamentary bodies approving the treaties: It is about grasping the boundary. From the perspective of the accountable institution, getting the boundary wrong would result in approving hollow words, whereas the misguided ultra-vires reviewer would protect anything but the approved boundary.

As conceptually important as the “ban on blanket empowerments”-test may be, the test only ensures a boundary (and therefore accountability). Apart from that it is silent. The silence specifically applies to “the meaning” of any aspect that is no boundary. At a higher level, this becomes crucial and yet aggravated by the need not to confuse boundaries with anything else. Hence one particular weakness of the test pattern we have seen so far is that we get no clear notion of “not-a-boundary.” What actually is “not-a-boundary”? And how does it relate to gauging ultra-vires? Meet “Program and Process”: The third—and last—good part of the FCC’s ultra-vires case law. This elegant two-piece set bridges the remaining conceptual gap.

D. Why a “Program of Integration” Is a Good Medium for Expressing Vague Ideas

“Program of integration” (Integrationsprogramm) is what the FCC calls the “boundary” that the “ban on blanket empowerments” provides. Whereas the second label “Process of integration” refers to the dynamic course of events that is, taken by itself, “not-a-boundary.” This pair of terms dovetails with the ultra-vires domain. High-level expressions of this set of concepts are easy to formulate:

• “the current process of integration infringes the program of integration,”³¹

• “the parliamentary bodies have approved the changes to the program of integration, therefore the process of integration will probably gain momentum,”³²

• “the European Court of Justice is not precluded from refining the law and shaping the process of integration within the program of integration,”³³ or

• “the process of integration is dynamic by nature, European institutions autonomously seize ample opportunities, even so the program of integration is static by nature, until further parliamentary approval it remains unchanged.”³⁴

It bears mentioning that this now seemingly pretty natural pair of terms has been years in the making. The terminological variants of “program of integration” (Integrationsprogramm) include “framework”³⁵ (Rahmen), “program”³⁶ (Programm), “program of alliance”³⁷ (Bündnisprogramm), “crucial structural decisions”³⁸ (wesentliche Strukturentscheidung), “fundamental structure”³⁹ (elementare Struktur), “contents of the treaty”⁴⁰ (Vertragsinhalt), “concept of the treaty”⁴¹ (Vertragskonzept), “conventionary conception”⁴² (vertragliche Konzeption), “decision by the parliamentary act of approval of the treaty”⁴³ (vertragsgesetzliche Entscheidung), “fundamental orientation”⁴⁴ (Grundausrichtung), “basic order”⁴⁵ (Grundordnung), “basic concept”⁴⁶ (Grundkonzept), “political momentous decision”⁴⁷ (politische Grundentscheidung), “core of the concept”⁴⁸ (Kern der Konzeption), “system of competences”⁴⁹ (Kompetenzsystem), “fabric of competences”⁵⁰ (Kompetenzgefüge), “conceptual responsibility with regard do integration”⁵¹ (konzeptionelle Integrationsverantwortung), “context of meaning”⁵² (Sinnzusammenhang), “crucial boundary”⁵³ (wesentliche Grenze), “margin of judicial interpretation”⁵⁴ (Raum für gerichtliche Rechtsinterpretation), “boundaries relevant for judicial refining of the law”⁵⁵ (die der Rechtsfortbildung gesetzten Grenzen), and “system of constitutional distribution of power and distribution of leverage”⁵⁶ (System konstitutioneller Macht- und Einflussverteilung). This poetic richness is not just a consequence of random terminological excess, instead it is a product of a court’s personal statement: The Second Senate’s quest for the suitable gauge. But the mystifying problem with gauging ultra-vires has never been the flowery language, though this (lack of) technical vocabulary always seemed odd. As it stands, the attribution problem is the fundamental one.

The attribution problem refers to the difficulty of finding out whether some legal requirement is either part of the program of integration or part of the process of integration. It always was an error-prone endeavor because, once again,⁵⁷ the logical implication at hand is an unambiguous one-way street: Compliance with the program of integration is always a legal requirement. But not every legal requirement is part of the program of integration. Some legal requirements are easily attributed to either the program or the process, some are not. In any case, ultra-vires-review thoroughly depends on hyper-accurate attributions. This dependence is an unusual and defining characteristic of gauging ultra-vires. Common adjudication in case configurations other than ultra-vires review usually does not require this pinpoint distinction, as long as it is at least established that the requirement at hand is some kind of legal requirement.

These attribution concerns are true for all legal requirements and are especially true for courts refining the law. The mandate to uphold the law in the interpretation and application of the Treaties (Article 19.1 sentence 2 TEU) has, of course, never precluded the European Court of Justice from refining and further developing the law.⁵⁸ The same holds true for the FCC refining and further developing the Basic Law. The subtle point to notice is how these refinements and developments are to be attributed: The CJEU’s refinements are part of the process of integration, they cannot touch upon the program of integration, as any refinement to the program needs approval from the parliamentary bodies. By contrast, restating the existing program of integration is obviously feasible, but this goes without refinement or development (and, also, without renewed parliamentary approval). In short, the CJEU can refine and develop the process, but neither refine nor develop the program. The role of the FCC’s refinements is even more restricted: Changes to the program of integration as well as program supplements are clearly beyond reach. When gauging ultra-vires, the FCC is limited to restating the existing program of integration.⁵⁹

Yet another attribution problem arises as a result of the constitutional requirements with regard to predictability and legal certainty: Well-established case law demands that the program of integration “envisaged in the treaty provisions can be predicted and determined by the German legislative bodies.”⁶⁰ The concepts of predictability and legal certainty may seem to be well known—especially within the context of criminal law. But there is a caveat: In general, and especially in the criminal law field, FCC’s case law holds that uncertain laws can become certain over time, that less predictable laws can turn into more predictable versions of themselves. These important transformations allegedly rest solely upon jurisprudence.⁶¹ Be this as it may. Programs of integration do not share this flux. Their measure of certainty and predictability remains stable over time. At least, they remain unaffected by jurisprudence. Since the accountable parliamentary bodies approve the program of integration by law as a preliminary point,⁶² subsequent improvements by jurisprudence cannot be known beforehand and are therefore not approved by the accountable institution. Relying on eventual jurisprudential improvements would be just another “blanket empowerment” (only this time empowering courts).⁶³

The last attribution problem is closely linked with the not-uncommon idea that processes of integration can do only what the program of integration tells them to do. This very idea of “imperative programs” alienates a certain audience. For, eventually, we come to the question of what to do with open processes,⁶⁴ when we want to express a program of integration but our idea of what the open process of integration should accomplish exactly, or how the open process might proceed, isn’t specified in a complete manner—and therefore it is explicitly “open.” One could be forgiven for thinking that programs of integration are not expressive of vague ideas. But that does not add up. This belief in impossibility would be based on a confusion between form and content. An inflexible program of integration need not make for precision in describing processes of integration. The program of integration can be restated very precisely (and it should be, at least due to ultra-vires reviews), but the content of the program remains free. The accountable parliamentary bodies approving the program do not have to be exact in the idea of the open process. They may approve a range of acceptable paths of the process, being content if the “open” process of integration does not step out of this reach. The program does not have to precisely determinate particular processes, although it could. In a range of uncertainty, the program of integration may ask the process of integration to evolve inventively. The attribution problem enforces the need to be very precise in restating the program. No errors are allowed. Still, this does not imply a precise idea of future shapes of the process of integration. Gauging ultra-vires always prompts the question of whether some aspect of the process lies exactly over this range of uncertainty. And reliable answers always depend on reliable programs of integration.

Precisely restating programs of integration is no easy way to express vague ideas. But it is the only way at hand. When gauging ultra-vires, courts are challenged to develop their expertise in this field—intellectual and architectural. Restating the program of integration, the FCC is the honest messenger—not an envoy, not an ambassador, not a minister plenipotentiary, and certainly not delivering some kind of Ems Telegram. But yet there are problematic parts within the ultra-vires case law that cannot be easily avoided.

E. Problematic Parts, Not Easily Avoided, Or: Making Ultra-Vires Reviews Safe for the European Union

The ultra-vires case law—especially its most recent iterations—contains a large set of weak or problematic terms that can undermine the FCC’s attempts to perform reliable and predictable review. The Court should avoid the worst aspects of the case law. Surprisingly, perhaps, the FCC also should avoid some courses of conduct that are useful in general, but hazardous in the matter of ultra-vires. Such patterns are attractive nuisances and, by avoiding them, a dangerous class of potential ultra-vires errors is avoided.

The worst of all bad ultra-vires practices is the laesio enormis fallacy, that is taking some boundary for the program of integration, just because the (alleged) violation of the boundary would seem important to the Court. This fallacy does not sufficiently give effect to the program of integration and does not take the attribution problem seriously. Use of this fallacy severely degrades the reliability of ultra-vires reviews. Unions based on law cannot safely coexist vis-à-vis the laesio enormis fallacy.

The attribution error at the core of the laesio enormis fallacy is based on a gut feeling according to which “more important” boundaries are more likely to be part of the program of integration. But this is not the case. First, the parliamentary bodies deliberately approve boundaries, or don’t. Second, the process of integration can create important boundaries by itself (notably, the CJEU refining and further developing the law). Third, there is no such thing as a free abstract table of importance, guiding this attribution.

It is not entirely obvious that the FCC recently kept a safe distance from this fallacy.⁶⁵

A very dangerous pattern is the idea of methodologically reconstructing some judgment in a bid to find some kind of “untenability.” This Last Judgment fallacy does not necessarily result in improper ultra-vires review, although it certainly does most of the time. The problem with methodologically reconstructing some judgment is not that there would be no “true” methodology at all, though some might raise this quarrel. Instead, the Last Judgment fallacy is, also, a version of not sufficiently giving effect to the program of integration and not taking the attribution problem seriously. The Last Judgment fallacy likewise degrades reliability and is not safely reconcilable with unions based on law. The attribution error at hand is the idea that methodology would be part of the program of integration. If that were the case, any future change to methodological details would require parliamentary approval. The idea is not necessarily wrong, but highly unlikely: First, it is not clear that parliamentary bodies have ever approved any methodological change. Second, it is unclear, why—if there was a parliamentary approved ultimate methodology—this magnum opus has not yet been broadly distributed (it would certainly attract wide interest in academia and could revolutionize university teaching). Third, the obvious choice would be to task the CJEU with refining and developing methodological details just like refining and developing any other legal requirement—within the boundary of the program of integration—thus attributing the methodology to the process, not the program.

One might argue that the FCC recently walked into that trap.⁶⁶

Finally, the attractive nuisance: Refinement and development of the law is generally accepted to be one of FCC’s most honorable assignments. The FCC is famed for developing and refining the Basic Law (creating, for example, new fundamental rights⁶⁷ and, quite recently, even inventing completely new constitutional reviews on the basis of EU fundamental rights).⁶⁸ In ultra-vires matters, however, refinement and development lead nowhere. Refining the program of integration or developing this boundary would have to change the existing program, or else would fall short of any refinement or development. Performing this—as one might expect, tempting—task would degrade the reliability of ultra-vires review and would undermine unions based on law. First, the FCC would disregard the requirement of new parliamentary approval. Second, the Court would dismantle the concept of parliamentary accountability, because the new “refined” program of integration would not be the program approved by the parliamentary bodies (therefore any competence of executing refined or developed programs of integration would count as “blanket empowerment”). Third, as a single EU Member State lacks the competence to unilaterally change the existing program of integration, it is not immediately obvious where the FCC could derive this competence from.

Though it may be slightly subtle, the question remains: Why did the FCC make a referral to the CJEU in the first place,⁶⁹ asking many questions about boundaries for the ECB, if, recently, the FCC could just restate⁷⁰ the program of integration without any refinement or development?

F. Without a Solid Mental Model of How the Review Actually Works, All Judgments Will Be Brittle

In unions based on law, the decision to declare anything to be an ultra-vires act is one of objective reasonableness. The calculus of reasonableness must allow for the fact that programs and processes of integration are sometimes hard to distinguish. If a court reasonably believes that the process of integration takes over from the program of integration and therefore shatters parliamentary accountability, then declaring parts of the process to be an ultra-vires act is reasonable. But the standard any reviewer should employ is derived from serious questions: Does the court believe that the boundary in question is part of the existing program of integration (and does the court not covertly repair a flawed program)? Does the court believe that the existing program of integration is not refined or developed in passing, but is merely restated (and does the court not secretly negotiate a new treaty)? Does the court believe that the existing program and therefore parliamentary accountability would otherwise be undermined (and does the court not stealthily reshape the program)? If the answer to these questions is yes, then the second determination is whether it is reasonable for the court to hold that belief: Would any reasonable other constitutional court in light of the facts and circumstances confronting them hold the same belief? If so, gauging ultra-vires can protect parliamentary accountability without possible risk of harm to the process of integration.