THE GUARDI ANOFTHE CONS TI TUTI ON: HANS KEL S ENANDCARL S CHMI TTONTHE L I MI TSOF CONS TI TUTI ONAL L AWThis volume provides the rst Englishtranslationof Hans Kelsensand Carl Schmitts inuential Weimar-era debate on constitutionalguardianshipandthelegitimacyof constitutional review. It includesKelsens seminal piece, The Nature and Development of ConstitutionalAdjudication, as well as key extracts fromthe Guardian of theConstitution whichpresentSchmittsargumentagainstconstitutionalreview. AlsoincludedisKelsensreviewofSchmittsGuardianoftheConstitution, as well as some further material by Kelsen and Schmitt onpresidential dictatorship under Article 48 of the Weimar Constitution.These texts show Kelsen and Schmitt responding to one another, in thecontext of adebatefocusedonaconcrete constitutional crisis, thusallowingthereadertoassesstheplausibilityofKelsensandSchmittslegal and constitutional theories.larsvinx is an assistant professor in the Department of Philosophy atBilkent University, Ankara, Turkey. His main areas of interest are legalandpolitical theory, constitutional theoryandthehistoryof politicalthought.cambridgestudiesinconstitutionallawThe aim of this series is to produce leading monographs in constitutionallaw. All areas of constitutional law and public law fall within the ambit ofthe series, including human rights and civil liberties law, administrativelaw, as well as constitutional theory and the history of constitutional law.A wide variety of scholarly approaches is encouraged, with the governingcriterionbeingsimplythattheworkisof interesttoaninternationalaudience. Thus, works concernedwithonlyone jurisdictionwill beincluded in the series as appropriate, while, at the same time, the serieswill include works which are explicitly comparative or theoretical orboth. Theserieseditorslikewisewelcomeproposalsthat workat theintersection of constitutional and international law, or that seek to bridgethe gaps between civil law systems, the US, and the common law juris-dictions of the Commonwealth.Series EditorsDavid DyzenhausProfessor of Law and Philosophy, University of Toronto, CanadaAdam TomkinsJohn Millar Professor of Public Law, University of Glasgow, UKEditorial Advisory BoardT. R. S. Allan, Cambridge, UKDamian Chalmers, LSE, UKSujit Choudhry, Toronto, CanadaMonica Claes, Maastricht, NetherlandsDavid Cole, Georgetown, USAK. D. Ewing, Kings College London, UKDavid Feldman, Cambridge, UKCora Hoexter, Witwatersrand, South AfricaChristoph Moellers, Goettingen, GermanyAdrienne Stone, Melbourne, AustraliaAdrian Vermeule, Harvard, USABooks in the series:The Guardian of the Constitution: Hans Kelsen and Carl Schmitt on theLimits of Constitutional LawTranslated by Lars Vinx, with an introduction and notes by Lars VinxParliamentary Bills of Rights: The Experiences of New Zealand and theUnited Kingdom ExperiencesJanet L. Hiebert and James B. KellyLawyering for the Rule of Law: Government Lawyers and the Rise ofJudicial Power in IsraelYoavDotanBalancing Constitutional Rights: The Origins and Meanings of PostwarLegal DiscourseJaccoBomhoffJudges on Trial: The Independence and Accountability of the EnglishJudiciaryShimonShetreet and SophieTurenneProportionality and Constitutional CultureMoshe Cohen-Eliya and IddoPoratThe Politics of Principle: The First South African Constitutional Court,19952005Theunis RouxThe New Commonwealth Model of Constitutionalism, Theory andPracticeStephenGardbaumSearching for the State in British Legal Thought: Competing Conceptionsof the Public SphereJanet McLeanJudging Social RightsJeff KingProportionality, Constitutional Rights and their LimitationsAharonBarakParliamentary Sovereignty: Contemporary DebatesJeffreyGoldsworthyTHEGUARDI ANOFTHECONSTI TUTI ON:HANSKELSENANDCARLSCHMI TTONTHELI MI TSOFCONSTI TUTI ONALLAWTranslation, introduction, and notes byL A R S V I N XUniversity Printing House, Cambridge CB2 8BS, United KingdomCambridge University Press is part of the University of Cambridge.It furthers the Universitys mission by disseminating knowledge in the pursuit ofeducation, learning and research at the highest international levels of excellence.www.cambridge.orgInformation on this title: www.cambridge.org/9781107092686 Lars Vinx 2015This publication is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place without the writtenpermission of Cambridge University Press.First published in English by Cambridge University Press 2015This work contains translations of extracts from Der Hter der Verfassung, byCarl Schmitt ( 1996, Duncker & Humblot GmbH, Berlin) and Positionenund Begriffe, im Kampf mit Weimar Genf Versailles 19231939,by Carl Schmitt ( 1994, Duncker & Humblot GmbH, Berlin).Printed in the United Kingdom by Clays, St Ives plcA catalogue record for this publication is available from the British LibraryLibrary of Congress Cataloguing in Publication dataThe guardian of the constitution : Hans Kelsen and Carl Schmitt on the limits ofconstitutional law / edited and translated by Lars Vinx.pages cm (Cambridge studies in constitutional law)ISBN 978-1-107-09268-6 (Hardback)1. Kelsen, Hans, 18811973 Correspondence. 2. Schmitt, Carl, 18881985 Correspondence. 3. Constitutional law Philosophy. 4. Executivepower Germany. I. Kelsen, Hans, 18811973, author. II. Schmitt, Carl,18881985, author. III. Vinx, Lars, editor translatorK339.G83 2015342.001dc232014029908ISBN 978-1-107-09268-6 HardbackCambridge University Press has no responsibility for the persistence or accuracy ofURLs for external or third-party internet websites referred to in this publication,and does not guarantee that any content on such websites is, or will remain,accurate or appropriate.Chapters 1, 4, and 6 are printed with the kind permission of the Hans Kelsen-Institutin Vienna.CONTENTSAcknowledgements page viiiA note on the texts ixIntroduction 11 Kelsen on the nature and development of constitutionaladjudication 222 The guardian of the constitution: Schmitts argument againstconstitutional review 793 The guardian of the constitution: Schmitt on pluralismand the president as the guardian of the constitution 1254 Who ought to be the guardian of the constitution?Kelsens reply to Schmitt 1745 Prussia contra Reich: Schmitts closing statementin Leipzig 2226 Kelsen on the judgment of the Staatsgerichtshof of25 October 1932 228Notes 254Bibliography 266Index 274viiACKNOWLEDGEMENTSThe publication of this volume would not have been possible without thegenerous support of Professor David Dyzenhaus and of FinolaOSullivanat CambridgeUniversityPress. I amverygrateful totheHans Kelsen-Institut in Vienna, to Duncker & Humblot in Berlin, andto Professor George Schwab in New York for giving me permission tomakeuse of thetexts that aretranslatedinthis volume. I amalsoindebtedto the two referees of the initial proposal for CambridgeUniversity Press who made a number of very valuable suggestions andto Simon Wigley who read a draft of the introduction. Finally, I wouldlike to thank my copy-editor Jeremy Langworthy for helping to improvethe quality of the English text.viiiANOTEONTHETEXTSThe translations in this volume contain the page numbers of the Germansource-texts in square brackets. References to the translated texts in theintroduction, cross-references in and between the translated texts, andreferences in the notes at the end of the volume all use these bracketedpage numbers. The ellipses in chapters 2 and 3 indicate the beginnings ofpassages of Schmitts Guardian of the Constitution that are not includedin this volume.Anasteriskinthe translatedtext indicates aneditorial note thatprovides background information on a point discussed in the translatedtext. These editorial notes can be found at the end of this volume, startingonp. 254. Theyarelikewiseidentiablebyreferencetothebracketedpage numbers.The bibliography lists all literature cited in the introduction and in theeditorial notes. Kelsens and Schmitts works are referred to by the yearof original publication.ixuIntroductionI The PreussenschlagOn 20 July 1932, the conservative chancellor of the Weimar Republic,Franz von Papen, made use of an emergency decree that the president,Paul vonHindenburg, haddrawnupafewdaysbefore. Thisdecreeauthorizedthe chancellor todepose the government of Prussia, thelargestGermanLandorstate, thenundertheleadershipofthesocialdemocratic prime minister Otto Braun, and to appoint federal commis-sioners to take over the business of the Prussian ministers serving withBraun. Theallegedgoal of theso-calledPreussenschlag (the strikeagainst Prussia) was to restore public security and order in the state ofPrussia. There had been serious unrest and violence in the streets of thePrussian town of Altona a few days before, as a result of clashes betweencommunists, Nazis, and police. But von Papens federal government wasas responsible for this breakdown of public order as the Prussian govern-ment. It hadrecentlyliftedthebanontheSA1andthus helpedtoprecipitate violent clashes betweenNazis andcommunists. The realgoal of thePreussenschlagwastowrest control of Germanyslargeststate from the social democrats and to make Prussias executive poweravailable to the conservative federal government.2The emergency-decree that authorized the Preussenschlag was basedon article 48 of the Weimar Constitution. That article gave the presidentthe power, in its rst paragraph, to compel the states of Germany, if needbe by the use of armed force, to full their obligations towards the Reichunder the federal constitutionandunder federal laws. Moreover, itauthorized the president, in its second paragraph, to take all necessarymeasures to restore order in case of a severe threat to public security. Thepresidents decree appealed to both of these provisions. It claimed thatthe replacement of the Prussian government with a government1The Sturmabteilung, i.e. the paramilitary organization of the Nazi-party.2Clark(2007), 64054; Mommsen(2009), 52948; KolbandSchumann(2013), 1423,2647.1appointed by the chancellor of the Reich was necessary to restore publicsecurityandorder inPrussia, andit alsoaccusedPrussiaof havingviolated itslegal obligations toward theReich, though the decree itselfdid not specify this charge in any way.3Article48oftheWeimarConstitution, initslastparagraph, deter-mined that the president was required to inform the federal parliament,the Reichstag, immediately of any measures taken under article 48. Thepresident hadanobligationtosuspendemergencymeasures at therequest of the Reichstag. This restrictionof the presidents powersunder article 48, though, was no longer fully operative in July 1932, astheReichstaghadlongceasedtofunctioninthewayintendedbytheconstitution.The parties in the Reichstag had been unable, since 1930, toform alegislative majority willing to support a parliamentary government.Germany had instead been governed on the basis of presidential emer-gency decrees issued by appeal to article 48 paragraph 2. The chancellor,as a result, came todependmore onthe presidents trust thanonparliament.4The rst of these presidential governments, that ofHeinrichBrning, inofcefrom1930to1932, hadstill enjoyedthetoleration of parliament, or more precisely of the parties of the WeimarCoalition that had formed the last pre-crisis government. While therewas no majority willing to legislate for Brning, there was no majorityeither, due to the tacit support of the democratic parties, for a vote of nocondence against him that would have forced new elections.The centrist Brning, however, had been dismissed by Hindenburg atthe end of May 1932, for reasons unrelated to the lack of direct parlia-mentary support for his government. Hindenburg had then appointedthe ultra-conservative Franz von Papen as chancellor. Von Papen andhis supporters were keentoridthe presidential government of itsdependence on the democratic parties, and in particular of its depend-ence on the social democrats. The newgovernment dissolved theReichstag, a move that triggeredfederal elections withinsixtydays.Intheinterim, vonPapentriedtowinthesupport of theNSDAP,5whichwasexpectedtomakelargegainsinthecomingelections, forhisgovernment. Theplanwhicheventuallycametonaughtdueto3The decree is printed in Brecht (1933), 481.4Mommsen (2009), 32982, 43182; Rossiter (1948), 3173.5TheNationalsozialistischeDeutscheArbeiterpartei, i.e. theNational-Socialist GermanWorkers Party led by Adolf Hitler.2 theguardi anoftheconsti tuti onHitlers stubborninsistence that he be appointedchancellor of theReichwasnottoreturntoparliamentarygovernment, butrathertoget the Nazis to tolerate von Papens presidential government. The banon the SA, which had been put in place under Brning, at the demand ofthe interior ministers of the states, had been lifted to attract the supportof the Nazis. The Preussenschlag, then, was also intended asa concili-atory move towards the NSDAP, as it promised to get Prussian police offthe Nazi partys back.6Though the legality of the Preussenschlag was very much in doubt, thePrussian government under Otto Braun chose not to actively resist vonPapens measures. The constitutional situation in Prussia did not lookmuchbetter, inJuly 1932, thanthat inthe Reich.7The governingcoalition in Prussia, led by the social democrats, had lost its parliamen-tary majority in the elections to the Prussian legislature in April 1932.But the NSDAP, which had won the election and become the strongestparty, asyetdidnothaveenoughvotesinparliamenttoelectanewgovernment. Theoldparliamentarymajorityhad, inadvanceof theelections, changed the parliamentary rules of procedure for the electionof a new government, by introducing the requirement that a newgovern-ment could only be elected with an absolute (and not, as had previouslybeen the case, with a relative) majority of votes. Since the communistswere equally unwilling to support the election of a social democratic or anational socialist prime minister of Prussia, the new Prussian LandtagfailedtoelectaprimeministerandBraunsgovernmentcontinuedinofce in a caretaker role.There are indications that some members of the Prussian governmentwerenotinprincipleaversetotheappointmentofafederalcommis-sioner to take control of Prussias police force. Such a move had alreadybeen contemplated by Brning, not least to ensure that it would not fallinto the hands of the Nazis.8Von Papens Preussenschlag, however, wentmuch further. As we have seen, the decree of 20 July 1932 was justiednot merelyonthegroundthat federal interventionwasnecessarytorestore public order. It accusedthe Prussiangovernment of havingviolatedits legal duties towards theReich. Moreover, thedecreedidnotonlyputPrussiasexecutivepowertemporarilyintothehandsoftheReich. It envisagedacompletetransferof all competencesof thePrussiangovernment to the Reich, andthus appearedto eliminatePrussias independence as guaranteed by the federal systemof the6Mommsen (2009), 52991.7Clark (2007), 64054.8Seiberth (2001), 3758.i ntroducti on 3Weimar Constitution. Thedecree, accordingly, alsoempoweredvonPapen to remove all Prussian ministers from their ofces, a power thathe used to the full on 20 July 1932.9Though the Prussian government chose not to resist the Preussenschlagthrough violent means, it challenged the legality of the decree with anappealto the Staatsgerichtshof (literally the court of justice in matters of state) inLeipzig. It was supported in this appeal by several other German states thatfeared that von Papens Preussenschlag would turn out to be the rst step inageneral abolitionof federalism.10TheStaatsgerichtshof wasnot afullconstitutional court, endowed with a non-incidental and exclusive authorityto review and to annul unconstitutional legislation and acts of government.It was a special tribunal that was convened upon occasion at theReichsgericht (the Weimar Republics highest civil andcriminal court)and empowered, by article 19 of the Weimar Constitution, to adjudicateconicts between the federal government and the states.In its decision of the case, which was issued on 25 October 1932,11thecourt rejected the claim that the Prussian government had violated anyduties towards the Reich and it ruled that the federal government did nothave the power permanently to depose the Prussian ministers or to takeover all competences of a Prussian government. At the same time, thecourt held that the Reichs assumption of Prussias executive power wasjustiedasameasuretoprotect publicsecurity, andthusrefusedtointerfere with the federal governments momentary control overPrussias administrative apparatus.This attempt to split the difference left all parties unsatised. Thoughthe federal government kept control of the Prussian executive, the judg-ment blocked its suspected attempts to turn Germany into a politicallycentralizedstate, bymakingit clearthatthepowersof thepresidentunder article 48couldnot be usedtopermanentlyinfringe ontheprinciple of federalism. Nevertheless, the Preussenschlag did succeed inwrestingpolitical control ofthePrussianstatefromthehandsofthesocialdemocratsandtheircoalitionpartnerswhosupportedthecon-tinuing existence of the Weimar Republic. When Hitler was appointedchancellor inJanuary1933, HermannGringtookover thepost offederal commissioner for Prussia. Goebbels quippedthat vonPapenhad purged the Prussian state so carefully of republicans and democratsthat there was nothing left for the Nazis to do.129Brecht (1933), 4816.10Seiberth (2001), 11179.11Printed in Brecht (1933), 492517.12Mommsen (2009), 543.4 theguardi anoftheconsti tuti onThe Preussenschlag was not just a key event in the disintegration of theWeimar Republic and the rise to power of the Nazis. It also marked theculmination of two of the most important jurisprudential debates thattook place in the Weimar era: the discussion on the nature and limits ofexecutive powers of emergency under article 48 and the debate on thelegitimacyanddesirabilityofconstitutionaladjudication.13Thesetwodebates intersected in the context of the Preussenschlag.There were those, on the one hand, who, like Carl Schmitt, advocated anextensive reading of the presidents powers under article 48. In a situation ofconstitutional crisis, Schmitt believed, onlyapolitical powercapableoftaking a decision on the exception,14to suspend the law altogether, wouldbe able to restore the situation of normality that, in Schmitts view, mustunderpin all legal governance. The power of constitutional guardianship,therefore, must belong to the head of the executive and not to a court, asimplicitly acknowledged, according to Schmitt, by article 48 of the WeimarConstitution.15Other inuential constitutional theorists, among themHans Kelsen, favoured the view that constitutional guardianship ought tobe the preserve of a constitutional court, i.e. of a court empowered to controlall acts of legislation and of high-level executive action for their conformitywith the constitution, and explicitly endowed with the authority to invalid-ate acts deemed unconstitutional.16Unsurprisingly, Kelsen and Schmitt came to different assessments con-cerning the role of the Staatsgerichtshof in the aftermath of thePreussenschlag. InKelsens view, the judgment of the Staatsgerichtshofhad failed out of undue deference to the president, who was constitution-ally responsible for executing the judgments of the Staatsgerichtshof toannul the legal effects of an emergency decree that the court itself appearedto regard as unconstitutional. This confusing outcome, Kelsen suggested,could have been avoided if the case had been decided by a proper constitu-tional court.17Schmitt, who acted as counsel for the federal government atthe trial in Leipzig,18expressed the opinion, by contrast, that the presidentsdecree ought not to have been subject to substantive judicial review in therst place.1913Stolleis (2002), 11418.14Schmitt (1922), 515.15Schmitt (1924); Schmitt (1931a).16Stolleis (2002), 11718.17Kelsen (1932a), 6591. Translation in ch. 6 of this volume.18Seiberth (2001), 78110; Mehring (2009), 281302. See also Schmitt (1932c).19See Schmitt (1931a), 59; Schmitt (1934b), 447; Schmitt (1932d), translated in ch. 5 ofthis volume.i ntroducti on 5Thesedifferingassessmentsof thejudgmentonthePreussenschlagwereonlythepartingshots inalonger debatebetweenKelsenandSchmitt on the problemof constitutional guardianship.20In 1929,Kelsenhadpublisheda paper entitledWesenundEntwicklung derStaatsgerichtsbarkeit (On the Nature and Development ofConstitutional Adjudication) that systematically laid out the case for aconstitutional court as a guardian of the constitution.21Schmitt, in turn,had challenged Kelsens advocacy of a constitutional court in a numberof articles that were eventually integrated into a book-length treatment,which appeared in 1931 under the title Der Hter der Verfassung (TheGuardianof theConstitution).22KelsenrespondedtoSchmittsbookwith a review Wer soll der Hter der Verfassung sein? (Who Ought tobe the Guardian of the Constitution?)23 that is one of the most incisivecriticisms of Schmitts constitutional theory ever written.The aim of the present volume is to make these texts available, for therst time, in English translation.II The KelsenSchmitt debateAs pointed out above, the constitution of the Weimar Republic did notprovide for the institution of a constitutional court. But there was a livelydebate as to whether the competences of the Staatsgerichtshof, created toarbitrateinconictsbetweentheReichandtheLnder, ought tobestrengthened so as toturn it into a full-blownconstitutional court. Inparticular, scholarsandpoliticiansdebatedthequestionwhethertheStaatsgerichtshof shouldbeendowedwiththepowertoannul uncon-stitutional legislation.24The Reichsgericht in Leipzig, in a much noted decision in 1925, hadclaimed that the courts of the Weimar Republic possessed an incidentalright of judicial review of legislation: a right not to apply statutes whichtheyconsideredtobeunconstitutional toaparticularcaseathand.25What is more, a highly developed system of constitutional adjudication20See Dyzenhaus (1997); Diner and Stolleis (1999); Beaud and Pasquino (2007);Gmplov (2011). For the background of the debate in German public law theory seeCaldwell (1997).21Kelsen (1929a). Translation in ch. 1 of this volume.22Schmitt (1931a). Chapters 2 and 3 of this volume offer a partial translation.23Kelsen (1931). Translation in ch. 4 of this volume.24See Schmitt (1931a), 37; von Hippel (1932); Stolleis (2003); Hartmann (2007).25Schmitt (1929b).6 theguardi anoftheconsti tuti onhadalreadybeenput inplaceintheRepublicof Austria, wheretheconstitutionalcourt hadbeen giventhepower, undertheconstitutionof 1920, tostrikedownunconstitutional federal andlocal legislation,upon appeal by the federal or by regional governments.26The Weimardebateonaconstitutional court, thus, wasoftenphrasedintermsofwhether Germany should adopt the Austrian solution.27In 1928, two presentations at the annual meeting of the Vereinigungder Deutschen Staatsrechtslehrer (the Association of the GermanTeachers of Public Law) by HeinrichTriepel and by Hans Kelsenengaged with the topic of constitutional review.28Both authors afrmedthe need for a constitutional court, though Triepel much more hesitantlythan Kelsen. Kelsens presentation is regarded as the classical plea for aspecial constitutional court endowedwithanexclusive authorityofabstract or non-incidental control of general legal norms issuedbyparliament or government. It has beenextremely inuential intheContinental European context,29while it has so far been largelyneglectedinAnglo-Americandebatesonjudicial review. Apart fromofferingargumentsdelegeferendafortheintroductionofaconstitu-tional court, Kelsens paper also put forward a host oflegal-technicalreections, i.e. of recommendations as to the best institutional design ofaconstitutional court. Theserecommendations weretosomeextentinuencedbythemodeloftheAustrianConstitutionalCourt. Kelsenserved as a judge on that court from 1920 to 1929, and he had, throughhisinvolvementinthedraftingoftheAustrianConstitutionof1920,helped to create it.30Kelsensargument fortheintroductionof aconstitutional court isbasedontheso-calledStufenbaulehre, thetheoryof legal hierarchy,whichKelsenadoptedfromhispupilAdolfJuliusMerkl.31Accordingto the theory of legal hierarchy, the process of the creation of law is to beunderstood as a step-wise sequence of enactments in which the creationof any legal norm is authorized by higher-level legal norms. A judicialdecision, for instance, is seen as the enactment of a particular norm thatis authorized by the statute which it applies. The enactment of a statute,26Heller (2010), 139234; Paulson (2003); hlinger (2003). See also Kelsen (1942).27See Schmitt (1931a), 6.28Triepel (1929) and Kelsen (1929a).29Stone-Sweet (2000), 328.30See Schmitz (1981); Olechowsky (2009); Lagi (2012). For Kelsen as a judge see Walter(2005). Kelsen was removed from the court in the wake of the constitutional reform of1929. See Neschwara (2005).31Compare Kelsen (1934), 5575. See also Koller (2005).i ntroducti on 7inturn, isunderstoodasauthorizedbytheconstitutional normsthatdetermine the proper procedure for the process of legislation and thatperhaps lay down material limitations for the production of general legalnorms.According to Kelsen, any norm-enactment on any level of legal hier-archyispartlydiscretionary: thefullmentoftheconstitutional con-ditions, procedural andsubstantive, for the enactment of a statutetypicallyleavesthelegislatorwithawiderangeof legislativechoices.Similarly, ajudgedecidingaparticularcasetypicallyenjoysacertaindegree of discretion in applyinga statute. Aswe movedown the legalhierarchy fromconstitutional norms towards particular judicial oradministrative decisions the level of discretion enjoyed by the relevantdecision-takers will tend to decrease. Kelsen argues, however, that thereis no qualitative difference, only one of degree, between the activity of alegislator and that of a judge or an administrator. Just as a judge applies astatute in enacting a particular norm that will decide a particular case,legislators, though they have greater freedom of choice, apply constitu-tional norms in enacting statutes.32If that is the case, Kelsen concludes, there is no good reason to holdthat the activity of legislators cannot or should not be subject to constitu-tional review. No one would doubt that the actions of lower-level legalauthorities, of subordinate executive agencies or judges ofrst instance,shouldbesubject toreview, inordertoguaranteethelegalityof therelevant particularnorm-enactments. But if legislation(orhigh-levelexecutive action) is also a form of the application of law, it is as possibleand as necessary to offer a guarantee that legislators or government willabide by the constitutional norms that authorize and limit their activity.If there are no guarantees of constitutional legality, Kelsen argues, thenthe constitution, as the highest and most important level of legal order,will remain a form of second-rate law that lacks full legal force. And asufcient guarantee of constitutional legality, in Kelsens view, can onlybe provided by a constitutional court endowed with the power to annulunconstitutional legislation as well as unconstitutional acts ofgovernment.33A constitutional court, moreover, is of special importance in a dem-ocratic and federal state. Its guarantees of constitutional legality protectminorities against the potential excesses of the rule of a majority; a rule32Kelsen (1929a), 14857.33Ibid., 15246. See Troper (1995); Nino (1996), 18996; Vinx (2007), 14575.8 theguardi anoftheconsti tuti onthat will become tolerable, according to Kelsen, only if it is bound to therule of law. A federal state is in need of a constitutional court, since it is tobe understood as a system in which two mutually independent author-ities are legally co-ordinated on the basis of a constitutional division ofcompetences. Suchco-ordinationrequiresanimpartial arbitrationofconicts of competence between the central and the local authorities thatcan only be offered by a constitutional court.34Though Kelsens argument was on the whole received favourably atthe meeting of the Vereinigung der Deutschen Staatsrechtslehrer, his pleaforthecreationofaproperconstitutionalcourtinWeimarGermanyalsocalledforthstrong opposition. Schmitts 1931 monographDerHter der Verfassung(TheGuardianof theConstitution) is inlargepart a reply to Kelsens arguments on constitutional adjudication.Schmitts argument against Kelsen builds on the claim that constitu-tional adjudication exceeds the legitimate powers of a court.35A judicialtribunal called upon to adjudicate on the constitutional legality of legis-lation or of acts of government would, Schmitt argues, typically have totakedecisionsthatarecontestableandsubjecttoreasonabledisagree-ment. Constitutional provisions, incontrast toordinarystatutes, areoften too vague and open-textured to allow for uncontroversial applica-tion. As a result, a constitutional court would be forced to take politicaldecisions, decisions that are nolonger justiable as applications ofdeterminate legal norms. It would have to act as a constitutional legis-latorandthusviolatetheseparationofpowers. Theintroductionofaconstitutional court, Schmitt concludes, would not de-politicize consti-tutional conict but rather politicize the courts and thus undermine thelegitimacy of judicial activity.Schmitt, however, was as opposed to parliamentary sovereignty as hewastoconstitutional adjudication. Schmitt heldthat modernparlia-ments, as a result of pluralist division, are no longer capable of takinggenuinely political decisions in the name of a people as a whole.36In theconstitutional monarchiesof thenineteenthcentury, Schmitt claims,parliament could claim to be a representative of the people as a wholebecause it opposed a monarchical executive the sovereignty of which wasstill, in principle, uncontested. Parliament acted as the defender of a non-34Kelsen (1929a), 15269. See also Kelsen (1927), 1627.35Schmitt (1931a), 1248; Schmitt (1967).36Schmitt (1931a), 7391. See also Schmitt (1938), 6577.i ntroducti on 9political social sphere against the incursions of an administrative statethat possessed an undoubted monopoly of political decision.With the establishment of a parliamentary system of government, anddue to the accelerating process of modernization, the legislature has, inSchmitts view, come to occupy a very different position. The traditionaldistinction between state and society has disappeared, in modern society,together withthe limitationof the states sphere of activity that itimplied. The state is now, at least potentially, a total state.37There canno longer be any principled limits to the states interference with societyand the economy. At the same time, the state has lost its transcendentposition above the fray of party-political conict. It is now controlled byparliamentary majorities that act in pursuit of their own sectional inter-est. Thestate, evenwhileseeminglyhavinggrownmorepowerful, nolongerexpressesthepolitical identityofthepeopleasawhole. Ithasbecomeaninstrument inthehands of parliamentaryleaders whosebickering has thrown it into political paralysis. Schmitt concludes thatparliament cannot function as a guardian of the constitution.Schmitts reaction to the perceived threat of a pluralist disintegrationof thestatewastwofold. Ontheonehand, Schmitt championedtheclaim, despitehishostilitytoconstitutional review, that theWeimarConstitutionputabsolutelimitationsonthepowersofparliamentarymajorities.In Schmitts interpretation, the Weimar Constitution, as the expres-sion of a constituent choice of the German people, contained an intan-gible core of fundamental political decisions that are legally immune tochange by any constituted power, including parliaments power of con-stitutional amendment.38This view was not supported by the text of theWeimarConstitution, whichdoesnotmentionanymaterial limitsofamendment. Schmitt tried to justify it, rather, on the basis of a generaltheory of what a constitution is. According to this theory, a constitutionis not to be identied with the constitutional laws that are contained inthe written constitutional text. Rather, a constitution,rst and foremost,is a concrete social order or positive constitution, which is put in placebyanexerciseof constituent powerandwhichembedsanumberoffundamental social values. The written constitution, in Schmitts view, isno more than an attempt to codify this antecedent concrete social orderendorsed by the popular sovereign. Its norms and procedures, therefore,37See Scheuerman (1999), 85112; Cristi (1998), 17999.38Schmitt (1928), 724, 7981, 1508; Schmitt (1932a).10 theguardi anoftheconsti tuti onare to be regarded as binding only as long as they help realize and protectthe positive constitution.39The constituent choice that afrms a concrete order or positiveconstitution manifestsanantecedent political unityof thepeople, aunity that Schmitt claims is prior to the politics that takes place underthe regulation of constitutional norms. Schmitts constitutional thoughthas atendencytoidentifytheauthenticwill of thepeoplewiththedecisionsofapre-legalconstituentpower, allegedlyunaffectedbythepluralist division into several competing political parties. True democ-racy, as aresult, is heldtoexist onlyintherealmof constitutionalpolitics, while constituted democracy is denigrated as the scene of plural-ist disintegration.40The function of the constitution, in this perspective,is tohedgeinparliamentarypolitics, throughanappeal topopularsovereignty, evenwhilethelegal illimitabilityof constituentpowerisenthusiastically afrmed.41Schmitts second, related reaction to the perceived threat of pluralistdisintegration was to play up the role of the president of the Reich undertheWeimarConstitution. Theconstitutionhadindeedendowedthepresident, who was directly elected by the people, with a fairly impressivearray of competences.42The most important of these were thepresidentspowersofdictatorshipunderarticle48paragraph2oftheWeimar Constitution, which, as we have seen, were used by PresidentHindenburg tolegislate by emergency decree andthus topropuppresidential governmentsthat didnot havetheexplicit support of aparliamentary majority. Schmitt had, in an earlier piece, criticized thatpractice.43But in the Guardian of the Constitution, he came around todefending it.44Schmitt argued that the Weimar Constitution had put thepresident into the position of a neutral power above party politics thatwas to form a counterweight against parliament. The president, hence,was to assume legislative power in case of parliamentary dysfunction. Inthuspreservingorderlygovernment, thepresident, inSchmittsview,showed himself to be the true guardian of the constitution.45Incallingthepresident aneutral power, Schmitt didnot meantosuggest that presidential guardianship of the constitution would not be39Schmitt (1928), 7588.40Ibid., 79, 1389.41Ibid., 12535. See Vinx (2013b).42ThepresidenthadthepowertodissolvetheReichstag(art. 25), toinitiateapopularreferendumona statute approvedby the Reichstag (art. 73), andtoappoint thechancellor (art. 53).43Schmitt (1924), 21318.44Schmitt (1931a), 12831.45Ibid., 13259.i ntroducti on 11political. Rather, hispointisthatitwouldnotbeparty-political. Thisclaim is in line with Schmitts general interpretation of article 48, whichholds that the president, in using his powers of emergency, exercises aresidue of sovereign dictatorship.46According to Schmitt, the applica-tionof all law, includingconstitutional law, requires aconditionofnormalitythat canonlybesecured, if it isthreatened, byextra-legalactsof dictatorship.47Sincetheconstitutionhadmadethepresidentresponsible for the exercise of powers of dictatorship, and since it hadimposed no real constraints, in Schmitts view, on the discretion of thepresident in the exercise of those powers, it fell to the president to securethe condition of normality requisite to the functioning of the WeimarConstitution. In deciding whether and how to use his powers of emer-gency, the president thus became the highest interpreter of the WeimarConstitution. In contrast to parliament, the president, elected directly bythe people, was supposedly legitimized by the unied popular will thatalso stood behind the constitution as a whole.48Schmittsargument hereisadevelopment of certainideasof MaxWebers, whose advocacy for a strong president had been inuential inthe making of the Weimar Constitution.49But Schmitt gives a peculiaraccentuation to Webers call for a strong executive supported by plebi-scitary legitimacy. One way to read the Guardian of the Constitution is tounderstandthepresidentspowers, asdescribedbySchmitt, asessen-tially remedial.50The president, under such a reading, is to step in andpreserve the functioning of the constitutional system in the event that theparties in parliament turn out to be unable to form a stable governingmajority.But while it is true that parliamentary government had broken downby 1931, it should be noted that Schmitt had tried to limit the powers ofparliament already in his Constitutional Theory of 1928, well in advanceof the outbreak of thenal crisis of the Weimar Republic. His argumentin the Guardian of the Constitution, moreover, suggests that a presidentasaneutral powershouldberegardedasanecessaryandpermanentfeature of a parliamentary-democratic constitution.51Schmitts histor-ical analysis of the allegedprocess of the pluralist disintegrationof46Schmitt (1924), 2027. See McCormick (1997), 12156; McCormick (1998); Dyzenhaus(2006), 3454.47Schmitt (1922), 13. See Scheuerman (1999), 1180; Hofmann (2002), 3477; Croce andSalvatore (2013), 1176.48Schmitt (1931a), 1569.49Mommsen (1984), 3819.50Compare Schmitt (1931a), 1301.51Ibid., 13240.12 theguardi anoftheconsti tuti onparliament, which portrays the trend towards the dysfunction of parlia-mentary government as a natural consequence of the disappearance of asharp distinction between state and society and of a transcendent state,implies that one should expect a purely parliamentary system of govern-ment to become dysfunctional.Parliamentary government, Schmitt appears to suggest, has turned outto be incapable of providing Germany with strong and coherent politicalleadership. It can be allowed to continue to exist only under the super-visionof astrongpresident whohasastandingpowertoappoint apresidential government and to shield it, if need be, from accountabilityto parliament. Though Schmitt presents his argument in the Guardian ofthe Constitution as an interpretation of the powers of the president undertheWeimar Constitution, theimplicit aim, clearly, is toestablishapolitical systemthat resembles a constitutional monarchy, withthedifference that the positionof the monarchis tobe occupiedbyapopularly elected president whose decisions will provide the people asconstituent power with the necessary political leadership.52The political subtext of Schmitts Guardian of the Constitution was notlostonKelsen, whorespondedtoSchmittsbookwithalongreviewentitledWhoOught tobe the Guardianof the Constitution. Kelsenrefusedtopassjudgement onwhat hetooktobeSchmittspoliticalgoals. But heargued, inthecourseof adefenceof hisownpleaforconstitutional review against Schmitts criticism, that these goals couldnot be supported by way of an interpretation of the Weimar Constitutionand that they ought not to be regarded as democratic.Kelsens response to Schmitt points out that it is impossible to uphold astrict separation between legal and political decisions or between adjudica-tionandlegislation. Everyjudicial decision, not just thosetakenbyaconstitutional court, must be regarded as discretionary, and thus as political,to some extent, as Schmitt himself had conceded in an earlier publication.53The theory of legal hierarchy explicitly incorporates this insight, in holdingthat all legal decision-taking is partly discretionary. It therefore makes nosense, Kelsen argues, to issue a blanket rejection of constitutional adjudica-tion on the ground that it would have to be political. If it shares that featurewith all adjudication, the question ought to be whether it is advisable, in the52See Cristi (2011) and compare Schmitt (1927), 3154. On Schmitts involvement withplans of constitutional reform in the late Weimar Republic see Seiberth (2001); Mehring(2009), 281302; Berthold (1999).53Schmitt (1912).i ntroducti on 13interest of legality, to have conicts over the constitutionality of legislationortheconstitutionalityofactsofgovernmentsettledbyaconstitutionalcourt. Schmitts negativeargument, inKelsens view, simplyrefuses toaddress that question.54Kelsens engagement with Schmitts positive claim that the president isthe guardian of the constitution under the Weimar Constitution isgrounded in a realistic view of democratic politics. Schmitts argument, aswe have seen, assumes that there is a unied will of the people as a whole,above and beyond the legislative will that is formed through the deliber-ations and negotiations of parliamentary parties, a will that is expressed inconstituent choicesandactualizedbythedictatorshipof thepresident.According to Kelsen, this assumption is plainly false. The WeimarConstitution, Kelsen drylyobserves, wasenacted by a parliament staffedwith party representatives, and presidents are elected, by majority vote, in acampaign that can hardly be free from party-political inuence.55InKelsensview, Schmittsassumptionisnot just ctional. It alsocarries autocratic implications. If there is in fact no authentic and uniedpopular will above and beyond the will that is formed within the con-stitutionally constitutedpolitical systemthroughthe deliberation,negotiation, andbargainingof differentpartiesandinterestgroupsthen any attempt to declare one particular organ of the constitution therepresentative of such a will can be little more than a way to mask thepolitical self-empowerment of a particular group. Hence, the claim thatthe president can act as a neutral power is to be rejected as a piece ofimplicitly authoritarian ideology.56Democracy, accordingtoKelsen, istheongoingpeacefulsearchforpolitical compromise among different social groups that are all entitledtoparticipateinthegameof democraticpoliticsunderconditionsofequality. Democratic compromise, in Kelsens view, maximizes the free-dom of all under the system of coercive constraint that we call the law orthe state. It assures, in contrast to any form of autocracy, that the largestpossible number of citizens will see their ownpolitical preferencesrealized, while it holds out the promise to the outvoted that they maybeabletoformpart of afuturemajority.57Democracyisessentially54Kelsen (1931), 153953.55Ibid., 15616.56Ibid., 1572 n. 13. See also Kelsen (1929b), 3546.57Kelsen(1929b). SeeonKelsenstheoryofdemocracyDreier(1990); Herrera(1997),11837; Jestaedt and Lepsius (2006); Vinx (2007), 10144; Ehs (2009); Baume (2012);Lagerspetz (forthcoming).14 theguardi anoftheconsti tuti onconstitutional, sincethepeaceful searchforcompromiserequirestheacknowledged supremacy of a system of procedural rules and minorityrights that give voice and standing to all groups in society. A constitu-tional court actingasanimpartial guardianof theseproceduresandrights, therefore, is especially important in a democratic state.58Kelsen points out that, in making a parliament staffed by representa-tives of political parties the centre of gravity of the political system, theWeimar Constitution clearly sides with democracy, so understood, andagainst autocracy. Schmitts argument for presidential constitutionalguardianship, since it is a covert plea for autocracy, is therefore indefen-sible, Kelsenargues, asaninterpretationoftheWeimarConstitution.Schmitts conceptionof constitutional guardianshipseems less con-cerned with the goal of offering guarantees of democratic constitutionallegalitythan withthecreation of elbow-room foratranscendent stateunboundfromconstitutional law.59Kelsen, bycontrast, believedthatproperly enforced democratic constitutional law can help pacify politicalconict. He therefore held it to be desirable to subject political competi-tion as far as possible to the rule of constitutional law.60Kelsendidnotdeny, however, thattherearepotentiallimitstothelegalization of politics. A political conict, he conceded, may undoubt-edly become so intense that there is no longer any possibility of a legallyregulatedresolutionof thedispute.61ButSchmitt, asKelsenrealized,does more than to point to the evident possibility of such conicts.62Heargues, ineffect, thatajudicial guardianshipofconstitutional legalitywill always inhibit and pervert democracy, as it will inevitably tend toprevent the expression of the unied constituent will of the people thatSchmitt holds to be the only authentically democratic will.According to Schmitt, a peoples political existence is manifested inits willingness to assert itself, under the leadership of a sovereign dicta-tor, against existing constitutional (and international) legality. Schmittsreply to Kelsens claim that the unied will of the people is a mere ctionwould be to point out that such a will in fact exists for as long as a peopleiswillingtotake(orrathertosupport)genuinelypolitical decisions;decisionsontheexceptionthatconstitutepolitical community, inanextra-legal space, bydrawingthe line betweenfriendandenemy.6358Kelsen (1929a), 1526.59Kelsen (1931), 15703.60See Vereinigung der Deutschen Staatsrechtslehrer (1929), 123; Kelsen (1926), 13942;Kelsen (1929b), 76.61Kelsen (1931), 15334.62Compare Kelsen (1926), 1468.63Schmitt (1932b).i ntroducti on 15To preserve its political existence, which Schmitt elevates into the statusof a supreme political value, a people must avoid becoming ensnared in alegalitythat will forever postponeatrulypolitical decision. Kelsensstrategy of pacication through law is therefore to be rejected, accordingto Schmitt, even where circumstances would render it feasible.64It is wrong, then, to portray Kelsen as a mere normativist who wasoblivious of the limits of legality, and Schmitt as a hard-headed realistwho was courageous enough to reect on the political basis of law. Thedisagreement betweenSchmitt and Kelsen, rather, reects differingassessments of the relative value of constitutional legality. Kelsen advo-catesforalegalpeacethatprovidesanumbrellaforsocialdifference,while Schmitt demands the extra-legal creationandprotectionof asubstantive social homogeneity that he declares to be the basis of truedemocracy.III Kelsen, Schmitt, and the limits of constitutional legalityThesedifferingassessmentsof thevalueof constitutional legalityarereected inSchmitts and Kelsens reactions to the Preussenschlag.Arguing as counsel for the Reich, Schmitt claimed that it had not beenthe aim of the Preussenschlag to permanently abolish the independenceof the state of Prussia or to turn the Weimar Republic from a federal intoa unitary state by presidential decree. However, Schmitt did argue thatthe mere fact that the caretaker government in Prussia was not politicallyaligned with the federal government, and that it was indirectly depend-ent on the tolerance of the communists in the Prussian Landtag, con-stituted a threat to public order that permitted the president totemporarily assign allthecompetencesof thePrussian government toa federal commissioner.65At any rate, the presidential assessment of thestate of public security, Schmitt argued, as well as of the measures neededto protect it, should not be subject to judicial review, provided there wasno abuse of discretion.66If Schmitts positionhadprevailedwiththe Staatsgerichtshof, theindependenceofthestatesintheWeimarRepublicssystemoffeder-alism would clearly have been severely compromised. The governmentsof the Lnder would have been put under a standing threat of suspensionor removal at the presidents discretion. This wouldhave made it64See Vinx (2013a).65See Brecht (1933), 3941, 17781 and ch. 5 in this volume.66Brecht (1933), 1304, 291.16 theguardi anoftheconsti tuti onimpossible for the Lnder to rely on the constitutions system of feder-alism to protect their interests against a federal government of a com-petingpolitical stripe. Thecourt, apparently, wasconcernedtoavoidsuch an outcome, and it tried to limit the consequences of thePreussenschlagfortheconstitutionssystemoffederalismbyclaimingthatthePrussiancaretakergovernmentcouldnotbestripped, underarticle 48 paragraph 1, of its right to represent Prussia in the Reichsrat,the federal council where states exercised inuence on the federal proc-ess of legislation. The court also ruled that the Prussian caretaker govern-ment hadnot, at anyrate, beenguiltyof violatinganylegal dutiestowardstheReich, anditconsequentlyheldthatthepresidentlackedthe power to permanently remove members of the Prussian governmentfrom their ofces.67Nevertheless, the court afrmed the presidents judgement that therehad been a threat to public order and security in Prussia, and itconcluded that the president did have the right, under article 48 para-graph2, totemporarilyassignthecompetencesofthegovernmentofPrussiarelatingtotheinternal governanceof theLandtoafederalcommissioner, in order to restore public security.68In effect, theStaatsgerichtshof readdownthepresidentsdecree, andheldit tobeconstitutional insofar as it focused on the restoration of public order inPrussia.Both Kelsen and Schmitt were scathingly critical of the judgment ofthe Staatsgerichtshof, though for slightly different reasons. Kelsenargued, in his assessment of the judgment, that the courts own reasoningin support of the judgment implied that the decree was unconstitutionalinits entirety. Thecourt, inKelsens view, thereforeought tohaveordered the president to rescind the decree. In trying to split the differ-ence between the parties to the dispute, the courts judgment had failedtogive clear instructions tothe president, whowas constitutionallyresponsibletoexecutethedecisionsoftheStaatsgerichtshof, andthusto stop an emergency action that the court itself appeared to regard asunconstitutional. But the court was not to blame, in Kelsens view, for itsfailure to act decisively in the defence of the democratic republic. Theproblem, rather, wasthelack, intheWeimarConstitution, ofafullydevelopedconstitutional court, explicitlyendowedwiththepowertoannulunconstitutionallegislationandactsofgovernmentthroughitsowndecisions. Kelsenblamedthis legal-technical deciency of the67Ibid., 493.68Ibid., 493.i ntroducti on 17Weimar Constitution which, in his view, contrasted unfavourably withthe Austrian constitutions relevant provisions on the prevailingGermantheoryof publiclawthat continuedtobetoodeferential topolitical power.69Schmitt showered with ridicule the suggestion that the establishmentof a proper constitutional court might have helped preserve the WeimarRepublic.70Inalaterretrospectiveassessment of thetrial inLeipzig,Schmitt claimed that the Staatsgerichtshofs interference with thePreussenschlag helped the Nazis take power. The courts decision cast acloud of suspicion on the law-abidingness of President Hindenburg andthus increased the latters desire to return to parliamentary government.Thisdesire, accordingtoSchmitt, ultimatelyovercameHindenburgsapprehensions against permitting the formation of a government underthechancellorshipoftheBohemiancorporal AdolfHitler. SchmittsowndefenceofthePreussenschlagcouldnowbemadetoappearasadefence of the Weimar Constitution. Tostrengthenthe presidentialgovernment wouldhavebeentheonlyfeasiblewayin1932, Schmittargued, to prevent a national socialist government.71The question which of these two assessments of the Staatsgerichtshofsjudgment, andofthevalueoflegalityingeneral, ismoreappropriatecannot bedenitivelydecidedinthespaceof anintroductiontothepresent volume. It shouldbeclear, though, that thedebatebetweenKelsen and Schmitt on the problem of constitutional guardianship is ofcrucial importance to any assessment of their constitutional theories.ThoughSchmittthrewinhislotwiththeNazisin1933, itisoftenargued that he was not a supporter but rather an opponent of NationalSocialism before 1933. What is more, it is argued that his constitutionaltheory contained the resources to prevent the formally legal destructionof Weimar democracy, to turn it into a militant democracy capable ofdefendingitselfagainstitsadversaries. Schmittsunderstandingoftheconstitution as a substantive order that is more than a mere collection ofconstitutional laws would have justied, so the argument goes, a tempo-rarypresidential dictatorshipdesignedtokeeptheNazis awayfrompower and eventually torestore the Weimar Constitution.72Any suchdictatorship, of course, would have required the violation of some con-stitutional law or other and would thus likely have been rejected as illegal69Kelsen (1932a), 8991; Kelsen (1927), 16775; Dyzenhaus (1997), 12332.70Schmitt (1934b), 447.71Schmitt (1958), 34550.72See Kennedy (2004), 15483; Berthold (1999); Seiberth (2001).18 theguardi anoftheconsti tuti onby the majority of Weimar public lawyers who, like Kelsen, understoodconstitutional legality as conformity with constitutional laws. As a result,Kelsens attempt to develop a purely descriptive and value-neutral sci-ence of law that distinguishes sharply between the law as it is and the lawas it ought to be is sometimes accused, even by authors who harbour littlesympathy for Schmitt, of having unwittingly facilitated Hitlers pseudo-legal Machtergreifung.73Kelsens critique of Schmitts Guardian of the Constitution should giverise to second thoughts about such assessments. To present Schmitt as adefender of the Weimar Republic, one has to accept Schmitts constitu-tional theory, as well as the rather peculiar interpretation of the WeimarConstitutionthat ows fromit; aninterpretationthat was roundlyrejected by the majority of democratic constitutional lawyers inGermany at the time74and that, as Kelsen makes clear, intends a trans-formation of the parliamentary system into a constitutional monarchy inplebiscitarian guise. It may well be true that Schmitt was not a supporterof Nazismbefore1933andthat hehoped, instead, fortheindenitecontinuation of a conservative presidential dictatorship unaccountableto parliament that might have been justied on the basis of his theory ofconstitutional guardianshipandthat wouldhavebeenabletorecon-struct a strong state holdingrm control of society. But that obviouslydoesnt imply that Schmitt aimed to defend what we would recognize as atrulydemocratic constitutional order. It alsodoesnt implythat hisconstitutional theory didnt, ineffect, help to destroy the WeimarRepublic and pave the way for Nazism.75There can be little doubt, on the other hand, that Kelsen understoodhis advocacy of constitutional adjudication as an attempt to help realize ademocratic republics promise of a peaceful and equitable resolution ofsocial conict, which he took to be dependent on sufcient respect forconstitutional legality. Kelsens argument is addressed to the friends andnot to the enemies of liberal democracy, whom it aims to enlighten ontheinstitutional or legal-technical conditionsof liberal democracysproper functioning. It is true that Kelsen did not think that the prefer-ence for democracy over autocracy is capable of legal-scientic justica-tion. Buthedidnotdesertthecauseof democracy. Kelsensattitude73See Dyzenhaus (1997), 15760; Radbruch (1946).74See e.g. Anschtz (1933), 4046; Grau (1932), 27495.75See Dyzenhaus (1997), 70101; Cristi (1998), 17999; Scheuerman(1999), 1180;MacCormick (2004), Breuer (2012), 14371.i ntroducti on 19towards the crisis of democracy in the 1930s is summed up well in a shortpiece entitled Verteidigung der Demokratie (Defense of Democracy):76In the light of this situation, the question also arises whether . . . democracyshouldnotdefenditself, evenagainstapeoplethat does notwant it anymore, even against a majority that is united in nothing more than in the willto destroy democracy. To pose this question is already to answer it in thenegative. A democracy that attempts to maintain itself against the will of themajority, possibly even by the use of violence, has ceased to be ademocracy . . . he who is in favour of democracy must not allow himself tobe drawn into a fatal contradiction and reach for the method of dictatorshipin order to save democracy. One has to stay loyal to onesag, even if theship is sinking; and, while sinking into the depths, one can only hold on tothe hope that the ideal of freedom is indestructible and that it will come tolife again the more passionately the deeper it has sunk.77Endowed as we are with historical hindsight, we may be inclined to takeissuewiththissentiment. Surely, inlight of thehorrorsof NationalSocialism, itwouldhavebeenmuchbetterifKelsenandthemajorityof positivist Weimar-era public lawyers had understood the concepts ofdemocracy andconstitutionality differently, moresubstantively, soas to allow for a more forceful defence of democracy?Still, Kelsensclaimthat it will likelyturnout tobeimpossibletomaintaina democratic political systemif a majority of the peopleemphaticallyreject it turnsout, onreection, toberatherplausible.Would we want to call a political system democratic in which an absolutemajorityofthevotersaredeniedtherighttovoteforapartyoftheirchoice, because the parties in question have been banned as unconstitu-tional? What would we say if the system had turned into an executivedictatorship, bolstered by the threat of the use of military force, wherelaws are made by the president and not by a parliament in which differ-ent social groups are represented? Perhaps there are situations in whichanelitist dictatorshipof thekindadvocatedbySchmitt in19312isnecessary to prevent a form of pseudo-popular rule that is even worse.But in that case it would be preferable, arguably, not to sell that alter-native under the label of democracy or to claim that it is constitutional; ifonly because it will become rather difcult to put up principled intellec-tual resistance to a fascist system if one has already convinced oneselfthat thereisnoessential differencebetweendemocracyandpopulistdictatorship.7876Kelsen (1932b).77Ibid., 237.78See Schmitt (1926a), 1617.20 theguardi anoftheconsti tuti onConstitutional guardianshipforademocracybecomesirrelevant ifdemocracy has already failed, by losing the support of the majority of thepeople. Whatistobedoneinthatcase, asKelsenacknowledged, isapurely political question, not one of democraticconstitutional law. Aslong as democracy still enjoys the peoples support, on the other hand,the reasons for vesting the power to guard it in a judicial institution, andnot in the executive, would appear to be strong indeed. Though Kelsenfavoured the greatest possible extension of constitutional legality he maywell have had a sounder understanding of its political limits thanSchmitt. This suggestion may come as a surprise to somereaders, butonly because Kelsens political and constitutional-theoretical works donot yet, in contrast to Schmitts, receive the attention they deserve.i ntroducti on 211Kelsen on the nature and developmentof constitutional adjudicationTranslation of Hans Kelsen (1929a) Wesen und Entwicklung derStaatsgerichtsbarkeit in HansR. Klecatsky, Rene Marcic, andHerbertSchambeck(eds.), DieWienerrechtstheoretischeSchule. SchriftenvonHans Kelsen, Adolf Merkl, AlfredVerdross, 2 vols. (Vienna: Verlagsterreich, 2010), II, 1485531.THENATUREANDDEVELOPMENTOFCONSTITUTIONALADJ UDICATIONI The problem of legalityAdjudicationinmattersofstateisconstitutional adjudicationand, assuch, a judicial guarantee of the constitution. It is a limb of the system oflegal-technical measures whose purpose it is to secure the legality of thestates functions. The functions of the state themselves have legal char-acter; they appearaslegal acts. Theyare acts through which law (andthis means legal norms) is created, or through which created law (legalnorms already enacted) is executed. Consequently, the functions of stateare traditionally divided into legislation and execution, and the applica-tion of law, as a mere reproduction, is contrasted with legislation, as thegenesis, the creation, or the production of law.The problem of the legality of execution, in the sense of its conformitywith statute, and thus the problem of guarantees of this form of legality,is on the whole well-recognized. The legality of legislation, by contrast,understood as the demand for the legality of the creation of law, as well asthe idea of guarantees of this form of legality, seems to run into certaintheoretical difculties. Does it not amount to a petitio principii to want tomeasure the creation of lawby the use of a standard that is only producedtogether with the object to be measured? And the paradox that seems tobe contained in the idea of the legality of law becomes the greater the22more following the traditional view one identies legislation with thecreation of law and thus statute with law as such; so that the functionsthat arecomprised under the nameof execution, namely adjudication(judicature) and administration, and in particular the latter, appear tostand outside of the law, so to speak, and not to be genuine legal acts, torepresent nothingmorethanapplications or reproductions of alawwhose productionis somehowalready nished, whose creationhasbeen brought to completion, in advance of acts of execution. [1486] Ifone takes the viewthat the lawis exhausted by statute, the meaning of theidea of legality will reduce to conformity with statute. And in that case,an extension of the concept of legality will hardly be self-evident.This conception of the relationship between legislation and execution,however, is false.* The two functions are not opposed to each other in thesensethatthereisanabsolutecontrastbetweenthecreationandtheapplication of law: rather, each of them, on closer inspection, turns out tobe both creation as well as application of law. The relevant contrast is arelative, not an absolute, contrast. Legislation and execution are not twoco-ordinatefunctionsofthestate, butonlytwodifferentlevelsoftheprocess of the creation of law, a process which neither begins nor ends atthelevel of statute, thatstandinarelationshipofsubordinationandsuperiority. Thisprocesscontinuestothebottomviathelevel oftheadministrative decree, the level of the judicial decision, andof theindividual act of administrationuntil it arrives at the acts that putthelattertwo intoeffect (thesearetheactsofso-called enforcement).It continues to the top until it arrives at the level of the constitution, inorder eventually to reach, transcending the area of domestic legal order,the sphere of the order of international law that stands above all partic-ular domestic legal orders. Withthis successionof levels, whichisrelevant, fornow, onlyinsofarasitunfoldsitselfwithintheareaofaparticular state, weof courseintendtodonomorethantooffer aschematic representationof the major steps of the process throughwhich the law, while concretizing itself, regulates its own creation; andthrough whichthestate, togetherwith the law, continuously recreatesitself. Constitution, statute, decree, act of administration, judicial deci-sion, and enforcement: these are simply the steps in the formation of thewill of community that are typical, given the way in which positive laworganizes the modern state. Reality may of course depart from this idealtype. For instance, it is not necessary for a decree, i.e. for a general normissued by an administrative agency, to interpose itself between a statuteandtheactofenforcement; anditisthereforepossible, undercertainkelsenonconsti tutionaladjudi cati on 23circumstances, for adecreetobeissueddirectlyonthebasis of theconstitution, instead of being enacted as a means for the execution of astatute, and thus to stand alongside the statute, as a general normenactedby a representative, parliamentary body. And other modications of thetypical process of the creation of law are similarly possible. But for now,the discussion will presuppose the most common type.Sincetheconstitutionlargelydeterminesthewayinwhichstatutescome into existence, legislation is an application of law, if seen in relationtotheconstitution. Inrelationtothedecree, however, aswell as inrelationtootheractsstandingbelowthelevelofstatute, legislationiscreation of law. And a decree, similarly, is application of law in relationto statute, and it is creation of law in relation to the judicial decision oradministrative act that applies the decree. This decision, inturn, isapplication in relation to the levels above, but creation of law in relationto the level below, to the matter of fact that constitutes its enforcement.The route that the law travels, on its way from the constitution to thematter of fact that enforces a judicial or administrative decision, is one ofcontinuous concretization. While the constitution, statute, and thedecreerepresentthegeneralnormsofthelaw, [1487]whicharepro-gressively more saturated with content, the judicial decision or admin-istrativeactaretoberegardedasindividuallegalnorms. Alegislator,who stands only under a constitution that determines his procedure oflegislation, is bound by law only to a relatively limited extent. His free-dom, hisopportunitytoengageincreativedesign, isrelativelylarge.With every further step downwards, the relation between freedom andconstraint shifts towards the latter. In other words, of the two compo-nents, which determine the respective function, the application of law isstrengthened, and that of free creation of law is weakened. Every level oflegal order does not just represent a production of law, in relation to thelevel below, but also a reproduction of law, in relation to the level above.And insofar as it is application of law, reproduction of law, the idea oflegality is applicable to it since legality is nothing more than the relationof conformity in which the lower level of legal order stands to the higher.Hence, the demand for legality, and for specic legal-technical guaran-tees of legality, exists not just with respect to the act of enforcement, in itsrelation to the individual norms of the administrative command, of theadministrative decision, and of the judicial decision, or with respect tothese acts of execution in their relation to the general norms of the decreeor the statute. The demand applies as well with respect to the relation-ship of the decree to the statute, and with respect to the relationship of24 theguardi anoftheconsti tuti onthe statute to the constitution. Guarantees of the conformity of a decreewith statute and of the conformity of a statute with the constitution aretherefore as possible as guarantees of the legality of individual legal acts.Aguarantee of the constitution, hence, is a guarantee of the legality of thelevels of law that stand immediately below the constitution. That is,rstand foremost, a guarantee of the constitutionality of statutes.That the demand for guarantees of the constitution is still raised today or rather, that it has not been raised before today, and has therefore onlyrecently acquired urgency and become a subject of scientic discussion isdue not merely to the mistaken theory we characterized earlier on, whichlackedafull insight intothehierarchical structureof thelawor, whatamounts to the same thing, into the thoroughly legal nature of the functionsof the state* and of their respective relations to each other. The fact that thelegal ordersof modernstatesexhibit anabundanceof institutionsthatensure the legality of execution, while they do not provide for any, or onlyfor very poor, guarantees of the constitutionality of statutes (as well as of theconformity of decrees with statutes), is to be attributed to political motives.These motives, in turn, do not remain without inuence on juristic theory,which, after all, would have to take the principal initiative in enlighteningthe public on the possibility and necessity of such guarantees. This obser-vation is valid in particular for those modern parliamentary democracies ofEurope that grew out of constitutional monarchies. The doctrine of publiclaw of constitutional monarchy is still of great inuence today, though thisform of state has largely been pushed into the background. The constitu-tionalist doctrine still determines the juristic theory of the state to a highdegree in part consciously, where one wants to develop the republic, afterthe pattern of monarchy, in the direction of strong presidential power, andinpartunconsciously. Constitutional monarchy[1488]developedfromabsolute monarchy, and its doctrine is therefore often guided by the aim tomakethereductionofpowerthattheformerlyunlimitedmonarchhadsufferedasaresultofthechangeoftheconstitutionappearassmallorinsignicant as possible, or even to veil it altogether. Even in an absolutemonarchy, it is theoretically possible to distinguish between the levels of theconstitution and that of statute. But this distinction is practically insignif-icant, for the reasonthat the constitutionexhausts itself inthe basicprinciplethateveryexpressionof themonarchswill isabindinglegalnorm. And since the legal order therefore lacks a determinate constitutionalform, that is, since there are no differentiating legal norms that regulate theenactment of statutes in a way that differs from the procedure for changingthe constitution, the constitutionality of statutes is not a problem of anykelsenonconsti tutionaladjudi cati on 25importancewhatsoever. Withthetransitiontoaso-calledconstitutionalmonarchy, adecisivechangetakesplaceprecisely atthis point, whichissignalled in a very characteristic way by the designation of the newsystemasa Verfassungs-Monarchie orconstitutional monarchy. The decisive legalshift expresses itself in the strengthened signicance that is nowaccorded tothe concept of constitution: in the legal norm that statutes may only comeabout in a certain way (namely with the participation of a body of repre-sentatives of thepeople), andinthe fact (andthis is the rule of theconstitution) that this norm cannot be changed as easily as other generalrules of law, namely the statutes, i.e. in the fact that a change of this normnecessitates observance of a special, more difcult form, the constitutionalform (heightened majority, repeated decision, special constituent assembly,etc.), which differs from the ordinary form of statute. One would think thatconstitutionalmonarchy, inparticular, shouldhavebeenthegroundonwhichtheproblemof theconstitutionalityof statute, andthus of theguarantees of theconstitution, ought tohavemadeitself felt withthegreatest imaginable energy. The case was precisely the opposite! The con-stitutionalist doctrine veiled the legal shift so threatening to the position ofpower of the monarchies. In contradiction to the legal reality of the con-stitution, it presentedthemonarchastheuniqueor decisivefactoroflegislation, by declaring statute to be the sole will of the monarch, and byportraying the function of parliament as that of a mere assent more orless marginal, inessential, and inferior. To give an example of the methodthat was employed here: the well-known monarchical principle,* which isnotdeducedfromthepositiveconstitution, butimposedonitfromtheoutside, as it were, in order to interpret it in a particular political light, ormore correctly in order to reinterpret the positive law with the help of anideology alien to it; or the famous distinction between the statutory com-mand, which issues solely from the monarch, and the content of statute,which is agreed upon between the monarch and the representatives of thepeople. Thefruitofthismethod: thatitisnolongerconsideredtobe atechnical imperfection, but rather seen as its deeper meaning, that a statuteis to be regarded as valid once it appears in the ofcial gazette, with thesignature of monarch, and irrespective of whether the prescriptions relatingto the involvement of parliament in legislative decision-taking [1489] aresatisedor not. The decisive progress fromabsolute toconstitutionalmonarchyisthus, atleastintheory, almostcompletelynullied; atanyrate, the problem of the constitutionality of statutes and of its guarantees isaltogether avoided. The unconstitutionality of a statute signedbythemonarch, or even its nullication on grounds of unconstitutionality, cannot26 theguardi anoftheconsti tuti oneven penetrate into juristic consciousness as a practical legal question. Whatismore, theconstitutionalist doctrinebasedlessontheletterof theconstitution than on its aforementioned ideology claims for the monarchnot only the sanction of legislative decisions but also, with it and in it, theexclusive right to promulgate statutes. In signing the decision of parliament,the monarch is to conrm the constitutionality of the process by which thestatute came about. According to this doctrine, then, there is at least onepart of the process of legislation that is protected by a kind of guarantee.However, the function of control is claimed precisely by the one power thatis itself most in need of being controlled. Admittedly, the act of the monarchis put under responsibility, byvirtueof therequirement of ministerialcountersignature. Butministerial responsibility,*insofarasitisdirectedagainst acts of the monarch, is without practical signicance in a constitu-tional monarchy. And it is altogether irrelevant where defects of the processof legislation must be attributed to parliament, since it can only be enforcedby parliament itself.Theview, still widelyacceptedtoday, andstill defendedwiththegreatest diversity of arguments, that any scrutiny of the constitutionalityof statutes must remain off-limits to the organs tasked with the applica-tion of law, that the courts may at most claim the right to inquire whethera statute has been duly published, that the constitutionality of the genesisof a statute is sufciently guaranteed by the right of promulgation of thehead of state, as well as the realization of this legal-political view in thepositive law of the constitutions of the republics of today, all this derivesnotleastfromthetheoryofconstitutional monarchy, whosepoliticalideas, moreorlessconsciously, still inuencedthedesignofmoderndemocracies.II The concept of constitutionIf the question is to be answered whether and in what way the constitu-tion can be guaranteed, i.e. whether and how it is possible to ensure thelegalityofthelevelsoflegal orderthatstanddirectlybelowthecon-stitution and are immediately related to it, then it is necessary, above all,to gain a clear concept of the constitution.* And it is precisely the insightdevelopedhere that the structure of legal order is hierarchical which is alone capable of fullling this task. We do not go too far if weclaimthat the immanent meaning, the meaningthat was fromthebeginningintendedbythefundamental conceptof constitution, asitwas already in use in the legal and political theory of antiquity, [1490] iskelsenonconsti tutionaladjudi cati on 27accessible only if one starts out from the theory of legal hierarchy, sincethe idea of a succession of steps in the creation of law is implicit in theconcept of constitution.If, peeling off the many modications that the concept of constitutionhas undergone, one isolates itsrm and intangible core, what results isthe ideaof ahighest principle that determines the whole legal andpolitical order, a principle that is decisive for the nature of the commun-ity constituted by that order. However one denes the concept of con-stitution, it always appears with the claim to encompass the foundationof thestate, on which therest of itsorderis constructed. If onelooksmore closely, it becomes apparent that the concept of constitution,which, in this respect, overlaps with the concept of form of government,primarilyandalwaysreferstoafoundational principleinwhichthedistribution of political power nds its legal expression. It is the rule thatdetermines the genesis of statutes, of the general norms whose executionconstitutes the activity of the organs of state, namely of the courts andadministrative agencies. This the rule for the generation of the legalnorms that primarily form the order of the state, the determination of theorgans and of the procedure of legislation is the essential, original, andnarrowconcept of constitution. Thepositingof thisbasicruleistheindispensable condition for the genesis of the legal norms that regulatethe reciprocal behaviour of the human beings that form the communityof the state, as well as for the genesis of those legal norms that determinethe organs and procedures that are necessary for the application and theenforcement of those rules. The idea that the basic rule of the constitu-tion forms the foundation of all order of the state, and that it is thereforetobeas rmandunchangingaspossible, leadstotheviewthatitisnecessarytodifferentiatebetweenconstitutional normsandstatutorynorms; the former are not to be as easily changeable as the latter. Thisgivesrisetotheconcept of constitutional form, asdistinct fromtheordinary form of statute: the procedure of constitutional legislation (orconstitutional amendment) that differs from the ordinary procedure oflegislation, insofarasitistiedtospecial, inhibitingconditions. Intheideal case, this specic formis restrictedtothe constitutioninthenarrowerandessentialsense, withtheresultthataswetendtosay,thoughnot veryfelicitouslytheconstitutionalinthematerialsense,and only the constitution in the material sense, is also the constitution inthe formal sense.*Oncepositivelawprovides for aspecicconstitutional formthatdiffers from the form of ordinary statute, nothing stands in the way of28 theguardi anoftheconsti tuti onusingthisformalsofornormsthatdonotfall undertheconcept ofconstitution in the narrow sense; rst and foremost, for norms that donot determine the mode of creation but the content of statutory norms.Inthis way, aconcept of constitutioninawider sense comes intoexistence. This wider concept is inplaywhenmodernconstitutionscontainnotmerelynormsthatconcerntheorgansandtheprocedureof legislation, but also a catalogue of basic rights and rights of freedom.The primary, though perhaps not the exclusive purpose of such a cata-logueistoputupbasicprinciples, guidelines, andlimitationsforthecontent of statutes to be enacted in the future. If the equality [1491] ofcitizensbeforethelaw, thefreedomoftheexpressionofopinion, thefreedom of religion and of conscience, and the inviolability of propertyare enacted in the typical form the form of a constitutional guarantee ofthe subjective rights of subjects to equality, freedom, property, etc. thismeans, above all, that the constitution determines that statutes must notmerely come about in the prescribed way, but, in addition, must not havea content that violates equality, freedom, property, etc. The constitution,in that case, does not merely have the character of a law of process orprocedure, but also the character of material law. The unconstitution-ality of a statute can consist not only in the fact that the procedure bywhichthelawcameaboutwasdefective, butalsointhefactthatthecontent of the statute contradicts the basic principles or guidelines laiddown in the constitution, or that it exceeds the limitations imposed by it.If one wants to distinguish, in consideration of this point, between thematerial and the formal unconstitutionality of a statute, one should bemindful that this is permissibleonlywiththequalicationthat anymaterial unconstitutionality is also a formal unconstitutionality, insofarasastatutethat, byvirtueofitscontent, comesintoconictwiththerelevant provisions contained in the constitution will lose the defect ofunconstitutionalityonceitisenactedasaconstitutional statute.*Theissue, hence, always reduces to whether a norm has been enacted in theform of an ordinary statute or in constitutional form. Of course, if thereisnodifferentiation, inpositivelaw, betweenconstitutionalformandstatutory form, only the observance of the latter can ever be at issue. Andin that case, the proclamation of basic principles, guidelines, or limita-tions for the content of law will be meaningless from the legal-technicalpoint of view; it will be no more than a misleading appearance created forpolitical ends. Suchisthecasewithfreedomsthat areguaranteedinspecic constitutional form as soon as the constitution, as is often thecase, authorizes the ordinary legislator to restrict these freedoms.*kelsenonconsti tutionaladjudi cati on 29The constitutional provisions that relate to the procedure of legisla-tion, aswell asthosethat put upbasicprinciplesforthecontent ofstatutory law, can only concretize themselves in statutes. Guarantees oftheconstitutionarethereforegiventhisscopeoftheconstitutionnothing but means for the preventionof unconstitutional statutes.However, as soon as the concept of constitution through the mediationof the idea of constitutional form is extended to objects other than theprocedureoflegislation and thebasicdeterminationof thecontent ofstatutory law, there is a possibility for the constitution to concretize itselfin forms of law other than statute; in particular in decrees, and even inindividual legal acts. The content of the constitution can make the levelof statute superuous, just as the statute may be drawn up in such a waythat it does not stand in need of a decree in order to be applied throughindividual adjudicative or administrative acts. Aconstitutioncoulddetermine, forexample, that general legal normsmay, undercertaindeterminate conditions, be enacted not through a decision of parliamentbut rather through an act of government, as in the case of the so-calledemergency-decrees, whichstandimmediatelybelowtheconstitution,alongsidestatutes, replacingormodifyingthelatter, andare endowedwith the same legal force. [1492] Hence, such decrees are immediatelyrelated to the constitution (by contrast to decrees that merely execute astatute), andtheycanthereforebedirectlyunconstitutional, justlikestatutes, sothat theguaranteeof constitutionalityhastodirect itselfagainst themas muchas against unconstitutional statutes. Nothing,moreover, rules out the possibility that norms are enacted in constitu-tional form that are not just basic principles, guidelines, or limitationsfor the future content of law which can be concretized only by a corre-sponding statute. It is possible for norms enacted in constitutional formto regulate a subject matter so comprehensively that they can be imme-diatelyappliedtoconcretecases, throughactsofadjudicationandinparticular through administrative acts. Such is the case if the constitutionin this extended sense determines how certain of the highest organs ofthe executive, the head of state, ministers, or the members of the highestcourts, etc., are to be selected, and does it in such a way that the creationof theseorgans canproceedwithout anyfurther norms (statutes ordecrees)thatexplicatetheconstitutionalprovisionsinmoredetail, sothat the constitution is immediately executed in the act of appointment,beit anomination, anelection, oraselectionbylot. Thesesubjectsindeedappeartohavebeenadmittedintotheconceptofconstitutiontypically used by legal theory. One traditionally understands by a30 theguardi anoftheconsti tuti onconstitution (in the material sense) not just the norms concerning theorgans and the procedure of legislation, but also those which concern thepositionof thehighest executiveorgans, andinadditionthosethatdetermine the basic nature of the relationship of subjects to the authorityof the state or, in other words, the catalogue of basic rights and rights offreedom, a catalogue that, to put the matter in juristically correct form,amounts to certain basic principles, guidelines, and limitations for thecontent of statutes. The practice of modern states, whose constitutionalcharters, as a rule, exhibit these three elements, typically conforms to thisunderstanding of constitution. If this is the case, not only general norms,like statutes or decrees, but individual acts as well can have the characterof being immediate to the constitution, and may therefore turn out to beimmediately unconstitutional. The class of individual acts that areimmediate to the constitution can, of course, be extended as far as weplease, aslongaslegal normsdirectlyapplicabletotheconcretecaseare for one political motive or another clothed in constitutional form;so, for instance, if the legal norms that govern the law of associations, orthose that regulate the position of religious denominations, are enactedas constitutional statutes. Despite the fact that a guarantee of the legalityof actsthatexecutesuchstatutesformallyexhibitsthecharacterofaguarantee of the constitution, it is nevertheless evident that the specicform of guarantee of the constitution whose legal-technical design is tobepresentedinwhatfollows, namelyconstitutionaladjudication, willnot easilynd a place here. The concept of constitution has now beenextendedtoofarbeyonditsoriginal scopethat wederivedfromthetheoryoflegal hierarchy. Theindividual characteroftheunconstitu-tional act wouldgive rise toanopencompetitionof constitutionaladjudication with the administrative courts, which form part of a systemof measuresthat istoguaranteetheconformityof execution, andinparticular of administration, with statute.In all cases discussed thus far, we dealt exclusively with acts immediatetotheconstitution, andthereforewithcasesofimmediateanddirect[1493] unconstitutionality. There is a clear contrast between such actsand acts that are not immediate to the constitution, and that can there-fore only be mediately or indirectly unconstitutional. If the constitutionexplicitlylaysdown, ingeneral terms, theprincipleof thelegalityofexecution, andespeciallyif it raisesthedemandforaconformityofdecreeswithstatute, thelegalityofexecutionwill atthesametimeindirectly constitute a form of constitutionality, and vice versa. Let mehighlight in particular, because we are dealing with general norms, thekelsenonconsti tutionaladjudi cati on 31decree that executes a statute. The aim to secure the legality of such adecree, for reasons to be discussed later on, can still be included in thetasks of constitutional adjudication. Apart from this, it should be notedthat direct andindirect unconstitutionalitycannot alwaysbesharplyseparated from one another, for the reason that there may occur mixedor transitional forms between both types: for example, if the constitutionimmediatelyauthorizesall orsomeorgansofadministrationtoenactdecrees within their sphere of responsibility, in the course of the concreteimplementation of the statutes that they are to apply. The administrativeorgans, in that case, have the authority to enact such decrees of imple-mentation directly from the constitution. That they are at all permittedtoenactdecreesdirectlyresultsfromtheconstitution. However, whatthey have to decree, i.e. the content of their decrees, is determined by thestatutesthat standbetweentheconstitutionandthedecreesthroughwhich they are implemented. (It probably does not have to be especiallyemphasizedthat these decrees of implementationespecially withregard to their proximity to the constitution differ from the aforemen-tioned decrees which replace or change statutory norms. The latter areimmediatetotheconstitutionandcanonlybeunconstitutional, butcannot violate astatute.) Adifferent case: if the constitutionpositsbasicprinciples, guidelines, or limitations concerningthecontent ofstatutes tobeenactedinthefuture, for exampleintheformof theaforementioned catalogue of basic rights and rights to freedom, then itwill be possible for administrative acts to be unconstitutional in a differ-ent sense than the indirect one according to which every administrativeact that violatesastatuteisunconstitutional. If theconstitution, forinstance, determines that an expropriation may only take place in returnforcompletecompensation, andif anexpropriationtakesplace, inaconcrete case, pursuant to a statute of expropriation that conforms to theconstitutionindemanding full compensation, but inviolationof thestatutes as well as the constitutions determinations concerning compensa-tion, the relevant administrative act will not merely be unconstitutional inthenormal, indirect sense, namelyas violatingastatutorynorm. Theadministrativeact inquestionviolatesnot onlyastatute, andthusthegeneral constitutional principle of the legality of administration, but also aspecial principle explicitly enshrined in the constitution: the demand for fullcompensation in cases of expropriation. The act, hence, oversteps a speciallimit that the constitution itself imposes on legislation. It would thereforemake sense to mobilize an institution that serves to guarantee the constitu-tion as well against unlawful acts of this kind.32 theguardi anoftheconsti tuti onThe constitutional principle of the legalityof executiondoes notmerelydemandthat everyexecutiveact must beinconformitywithstatute. Aboveall, it implies that [1494]an act ofexecution may onlytake place pursuant to a statute, only as authorized by statute, and thusnever without a statutory basis. Hence, if a public authority, a court or anadministrative agency, posits an act without any statutory basis, the actin question is not really in violation of a statute, as there is no statute,after all, that could be used to check the acts conformity with statute.Rather, the act is lawless andthus immediatelyunconstitutional. Itmakes no difference here whether the lawless act in question does noteven appeal to a statute or whether the appeal is evidently made in badfaith, as for instance in a case where an administrative agency expropri-ates an urban apartment building pursuant to a law that authorizes theexpropria