H e a d n o t e s

to the Order of the First Senate of 17 December 2013

1 BvL 5/08

1. The legislature may define or clarify the content of applicable law with effect for the past only within the constitutional limits for retroactive legislation.

2. An ex post facto clarification of applicable law by the legislature is a constitutive retroactive regulation if it is intended to clarify a question of interpretation yet unresolved by the regular courts or to preclude a deviating interpretation.

FEDERAL CONSTITUTIONAL COURT

– 1 BvL 5/08 –

IN THE NAME OF THE PEOPLE

In the proceedings
for constitutional review

the Federal Constitutional Court – First Senate –

with the participation of Justices

Vice-President Kirchhof,

Gaier,

Eichberger,

Schluckebier,

Masing,

Paulus,

Baer,

Britz

held on 17 December 2013:

1. § 43 sec. 18 KAGG violates the constitutional principles of legitimate expectations following from Article 20 sec. 3 GG and is void insofar as § 40a sec. 1 sentence 2 KAGG is retroactively applicable to profit reductions that relate to shares of securities investment funds in the 2001 and 2002 tax assessment periods.

Reasons:

The request for judicial review concerns the question whether § 43 sec. 18 KAGG violates the principle of non-retroactivity following from the rule of law in Art. 20 sec. 3, Art. 2 sec. 1 GG inasmuch as it ordered the retrospective application of § 40a sec. 1 sentence 2 KAGG to all tax assessments covered by this provision that were not yet final. […]

A.

I.

[Excerpt from press release no. 12/2014 of 20 February 2014:

In the second half of 2003, the legislature addressed a problem of interpretation, namely whether reductions in the profits on fund shares could be deducted for income tax purposes. The question at issue was whether § 8b sec. 3 of the Corporation Tax Act (Körperschaftssteuergesetz – KStG), in the version in force since 1 January 2001, also applies to investment companies, even though § 40a sec. 1 KAGG did not originally refer to that provision. On 22 December 2003 the “Basket II Act” (Korb II-Gesetz, BGBl I p. 2840) inserted § 40a sec. 1 sentence 2 KAGG, which contains an express reference to § 8b sec. 3 KStG; according to the reasons given for the government’s bill, this was an “editorial clarification”. Under § 43 sec. 18 KAGG, the new § 40a sec. 1 sentence 2 KAGG is “to be applied to all tax assessment periods for which assessments are not final”.

The claimant in the initial proceedings is a bank. It held among its current assets shares of investment funds whose trading prices had decreased on 31 December 2002 below the carrying amounts of the 2001 annual financial statements. The claimant recognised impairment losses against its income and initially treated them as applicable for tax purposes. Because of the Korb II-Gesetz , the claimant lodged an amended corporation tax return with the tax office for the year 2002. In accordance with § 40a sec. 1 sentence 2 KAGG in conjunction with § 8b sec. 3 KStG, it increased its profit off the balance sheet by the write-downs it had claimed, but cited the unconstitutional nature of the retroactive effect. The Finance Court hearing the complaint stayed the proceedings in order to obtain a decision from the Federal Constitutional Court. The Finance Court believes that § 43 sec. 18 KAGG is unconstitutional because the new version of § 40a sec. 1 KAGG is not merely a clarification but has an impermissible “true” retroactive effect.

End of excerpt.]

[…]

II.

[…]

III.

[…]

IV.

[…]

B.

I.

The request for judicial review is admissible.

1. According to Art. 100 sec. 1 sentence 1 GG in conjunction with § 80 sec. 2 sentence 1 Federal Constitutional Court Act (Bundesverfassungsgerichtsgesetz – BVerfGG) the referring court must demonstrate in which respect its decision depends on the validity of the provision in question (cf. Decisions of the Federal Constitutional Court, Entscheidungen des Bundesverfassungsgerichts – BVerfGE, 105, 48 <56>; 105, 61 <67>; 133, 1 <10 and 11>). The order of referral must indicate with sufficient clarity that the submitting court would decide differently in case of the validity of the provision in question than in case it was invalid and how the court would justify this decision (cf. BVerfGE 105, 61 <67>; 133, 1 <11>). As long as it is not manifestly untenable, the referring court’s legal opinion on the ordinary law is decisive for the admissibility of a request for judicial review according to Art. 100 sec. 1 sentence 1 GG (cf. BVerfGE 2, 181 <190 and 191>; 105, 61 <67>; 133, 1 <11>).

2. The referring court’s interpretation of § 40a sec. 1 KAGG (former version – f.v.) is justifiable. It may be based in particular on the wording of the provision. The fact that the opposite interpretation of the provision also appears justifiable (cf. the judgments cited under A I 2 of the Munich Finance Court, Finanzgericht München , delivered in 2008 and 2009) and that the relevant legal question has not yet been clarified by the highest courts does not exclude the admissibility of the referral. For the referral to be admissible it is sufficient that the referring court’s legal opinion is not manifestly untenable. Prior clarification by the highest courts of a preliminary question on ordinary law that is decisive for the constitutional assessment is not required for a request for specific judicial review of statutes to be admissible. Alone the fact that first or second instance courts may also apply for a specific judicial review of statutes according to Art. 100 sec. 1 GG speaks against it.

3. In its order requesting judicial review the Finance Court sufficiently dealt with the possibility of an interpretation of the relevant statutory law provisions in conformity with the Constitution and denied […] it. […]

4. There was no obligation of the referring court to complement its order of 22 February 2008 with regard to a number of decisions the Federal Constitutional Court had taken in the meantime (cf. BVerfGE 126, 369; 127, 1; 131, 20; 132, 302) and which contain statements on questions of the constitutionality of retroactive laws that were also relevant to the referral.

Constitutional procedural law neither generally requires the referring court to continuously monitor factual or legal developments relevant to the order of referral but arising thereafter nor to update the referral when necessary. This applies in particular with regard to decisions of the Federal Constitutional Court on constitutional law issues also relevant to the question of referral, which are published only after the order of referral. However, the referring court is entitled to inform the Federal Constitutional Court about new and in its view important findings for the referral proceedings. It may also adopt a supplementary order if it intends to remedy deficiencies of the initial order of referral (cf. e.g. BVerfGE 132, 302 <310>).

5. The question of referral which relates to shares in equity funds is, according to the wording of § 40a sec. 1 sentence 2 KAGG, to be extended to shares in securities investment funds and respective reductions in profits. The purpose of judicial review proceedings to settle conflicts permanently (cf. in this respect BVerfGE 132, 302 <316> with further references), speaks in favour of such an extension of the question of referral. This does not raise any other constitutional questions.

The same applies to the 2001 tax assessment period, to which the question of referral is to be extended. […]

II.

§ 43 sec. 18 KAGG is unconstitutional insofar as it orders the retroactive application of § 40a sec. 1 sentence 2 KAGG in the 2001 and 2002 tax assessment periods to reductions in profits in connection with shares in securities investment funds. In this respect, § 43 sec. 18 KAGG, already in formal terms, has “true” retroactive effect (echte Rückwirkung ) (1). In constitutional terms, the retroacting reference in § 40a sec. 1 sentence 2 KAGG to § 8b sec. 3 Corporate Income Tax Act (Körperschaftssteuergesetz – KStG) must be treated as a constitutive amendment of the legal situation to date and therefore, also with regard to its substance, be examined under the principles governing “true” retroactivity (2). The conditions of a “true” retroactivity, which is permissible only in exceptional circumstances, are not met here (3).

1. § 43 sec. 18 KAGG has at least formally brought § 40a sec. 1 sentence 2 KAGG into force with “true” retroactive effect for the 2001 and 2002 assessment periods.

a) In its established case-law, the Federal Constitutional Court distinguishes between laws with “true” retroactive effect, which are generally incompatible with the Constitution (cf. BVerfGE 45, 142 <167 and 168>; 101, 239 <262>; 132, 302 <318>; with further references respectively), and those with “quasi” retroactive effect (unechte Rückwirkung ), which are generally permissible (cf. BVerfGE 132, 302 <318> with further references).

A legal provision has “true” retroactive effect if it changes ex post facto a settled matter (cf. BVerfGE 11, 139 <145 and 146>; 30, 367 <386>; 101, 239 <263>; 123, 186 <257>; 132, 302 <318>). This is the case in particular if its legal consequence shall apply, with encumbering effect, to settled matters even before its promulgation (“retroactive impact of legal consequences” – Rückbewirkung von Rechtsfolgen ; cf. BVerfGE 127, 1 <16 and 17>).

In tax law a provision only has “true” retroactive effect if the legislature ex post facto alters a tax liability that has already been determined (cf. BVerfGE 127, 1 <18 and 19>; 127, 31 <48 and 49>; 127, 61 <77 and 78>; 132, 302 <319>). In the field of income tax law this means that an amendment of legal provisions with effect for the ongoing tax assessment period is, at least under formal aspects, to be qualified as “quasi” retroactivity; since, according to § 38 Fiscal Code (Abgabenordnung – AO) in conjunction with § 36 sec. 1 Income Tax Act (Einkommenssteuergesetz – EStG), the income tax duty emerges only after the expiry of the assessment period, that is at the end of the calendar year (§ 25 sec. 1 EStG; cf. BVerfGE 72, 200 <252 and 253>; 97, 67 <80>; 132, 302 <319>; cf. also already BVerfGE 13, 261 <263 and 264, 272>; 13, 274 <277 and 278>; 19, 187 <195>; 30, 272 <285>). The same applies to assessments for corporation tax (cf. § 30 no. 3 KStG).

b) Formally, § 43 sec. 18 KAGG that was introduced through the Korb II-Gesetz promulgated on 27 December 2003 has “true” retroactive effect in the 2001 and 2002 tax assessment periods (cf. BVerfGE 126, 369 <391 and 392>) insofar as it applies to tax assessments which were not yet final in those assessment periods. They expired on 31 December of the respective calendar year and therefore before the promulgation of the Korb II-Gesetz. Accordingly, the new regulation applies ex post facto to a settled matter.

2. In the case at hand, the constitutional principles prohibiting “true” retroactive laws also apply with regard to the substance of the case, because § 40a sec. 1 sentence 2 KAGG, unlike assumed in the reasons given in the government bill (cf. Bundestag document, Bundestagsdrucksache – BTDrucks 15/1518, p. 17), must, in constitutional terms, be treated as a constitutive amendment of the former legal situation.

a) If § 40a sec. 1 sentence 2 KAGG only clarified retroactively what already was the applicable law, the question whether the provision was exceptionally compatible with the general prohibition of “true” retroactivity , in spite of its formal “true” retroactivity would not arise. The legitimate expectation in the applicable law then could, from the outset, not be affected, because the applicable law was not subject to subsequent substantive amendments.

Whether a retroactive amendment has declaratory or constitutive effect with regard to the former law depends on the content of the former and the new law which – with the exception of clear-cut statutory wording – in most cases needs to be established by means of interpretation.

The view taken in the reasoning of the draft law as regards § 40a sec. 1 sentence 2 KAGG, namely that the provision was only of a clarifying nature (cf. BTDrucks 15/1518, p. 17), is not binding on the courts. It neither limits the rights and obligations of the regular courts and the Federal Constitutional Court to exercise their judicial control nor does it relativise the relevant constitutional standards (cf. BVerfGE 126, 369 <392>).

As a rule, it is the judiciary that is called to provide binding interpretations of the law (cf. BVerfGE 65, 196 <215>; 111, 54 <107>; 126, 369 <392>). This also applies to the question of whether a legal provision is constitutive or declaratory in nature. However, the legislature may also amend the content of a provision it has enacted or further specify its details for clarification and, in doing so, may correct case-law it does not agree with. But in so doing, it must remain within the bounds of the constitutional system of which the limits for retroactive legislation following from the fundamental rights and the rule of law form part. The legislature may not undermine this binding nature and the competence of the courts to review a case by claiming that the provision it has enacted was only clarifying in nature (cf. BVerfGE 126, 369 <392>). The legislature has no authority to provide authentic interpretation of legal provisions (cf. BVerfGE 126, 369 <392>; 131, 20 <37>).

b) It is generally for the regular courts to interpret the ordinary law (aa); it is, however, generally for the Federal Constitutional Court to determine the content of a legal provision submitted in proceedings for specific judicial review (bb). However, when clarifying the question of whether a retroactive provision is constitutive or declaratory in nature, special rules apply; such a provision is, from a constitutional-law perspective, already deemed to be constitutive if it opts for or against a justifiable interpretation of a legal provision and thereby resolves serious doubts as to the interpretation of the applicable law (cc).

aa) It is primarily for the competent regular courts to interpret the ordinary law, to choose the relevant methods to be applied as well as to apply the law to the individual case and the Federal Constitutional Court is generally not competent to review whether this was done correctly (cf. BVerfGE 128, 193 <209>), as long as it is not apparent that there have been errors in applying the law or incorrect interpretations of statutes which result from a fundamentally erroneous view of the meaning of a fundamental right or, in particular, an erroneous view of the scope of its protection, (cf. BVerfGE 18, 85 <93>; established case-law). Apart from that and as long as it is within the bounds of a justifiable interpretation and permissible judicial development of the law, the application of the ordinary law by the regular courts is not constitutionally objectionable. If, however, their interpretation is extremely contradictory to the provisions applied, the courts claim powers that the Constitution conferred upon the legislature (cf. BVerfGE 49, 304 <320>; 69, 315 <372>; 71, 354 <362 and 363>; 113, 88 <103>; 128, 193 <209>).

bb) Insofar as the interpretation and understanding of the ordinary law is of relevance for determining the constitutionality of a law in specific judicial review proceedings, the Federal Constitutional Court itself fully reviews the ordinary law (cf. BVerfGE 2, 181 <193>; 7, 45 <50>; 18, 70 <80>; 31, 113 <117>; 51, 304 <313>; 80, 244 <250>; 98, 145 <154>; 110, 412 <438>; established case-law). In this case, it is not bound by the referring court’s interpretation of the ordinary law. It may fully examine essential preliminary questions of ordinary law and decide upon them as starting point for the constitutional review. This is the only way to avoid that the Federal Constitutional Court feels the need to review the constitutionality of a provision on the basis of an interpretation that is possibly, although still justifiable, partial and not shared by others, namely the regular courts. However, the Federal Constitutional Court can still adopt the regular court’s interpretation as expressed in its order of referral and will mostly do so if there are no doubts regarding its correctness.

cc) (1) Irrespective of the Federal Constitutional Court’s general competence to fully review the ordinary law in judicial review proceedings, to answer the question of whether a retroactive provision is, from a constitutional point of view, to be treated as constitutive, it is sufficient to establish that the amended provision in its original version could be and has been interpreted in a way that the new provision intends to exclude (cf. BVerfGE 131, 20 <37 and 38>).

(a) Under the Constitution, the legislature’s wish to clarify a legal situation retroactively can generally only be permitted within the limits set by the prohibition of retroactivity. By claiming the need for clarification the legislature could otherwise, also beyond these constitutional limits, give a legal situation its own notion of a correct interpretation without the competent courts having clarified if this in fact was the legal situation. Thereby, the protection of the confidence that the law remains stable, which is required by the rule of law, would be severely weakened. In view of the fact that the law is both open to and in need of interpretation, the legislature could otherwise easily justify a need for clarification. A competence to retroactively clarify the applicable law that is largely exempt from requirements of the protection of legitimate expectations would grant the legislature a broad discretion to intervene in already settled legal positions, would open the door for politically opportunistic considerations that were not characteristic for the ordinary law at the time of its interpretation – which later was held to be in need of revision – and would thereby seriously compromise the legitimate expectation that the law remains stable.

Nor does a right of the legislature to intervene in past matters follow from the principle of democracy; rather there is a tension between these principles. Whereas the prohibition of retroactivity limits the parliament’s discretion to regulate the past, the democratic responsibility of the parliament, however, refers to the present and the future. Legislative decisions adopted in the past have an autonomous democratic legitimation. The historic context of legitimation cannot – at least as long as the effects of the law lie in the past – simply be dispelled by the retroactive intervention of a present-day legislature. This would be obvious if laws were to revise retroactively decisions made during an earlier legislative period and taken under a different political majority. For the past, these decisions are democratically legitimised only by the former and not by the present-day decision-making context. The democratic constitutional state conveys upon the legislature a legitimation at the time. The principle of democracy, too, requires that the legislature’s intervention in the past remains an exception.

(b) In any event, the legislature’s retrospective clarification of the legal situation is to be qualified as constitutive retroactive regulation if the legislature thereby seeks to retrospectively undermine the interpretation of the law as clarified by the highest courts. Regarding the past, the legislature generally has to accept that the courts interpret the law applicable at that time with binding effect, within the constitutional limits of the judicial interpretation and development of the law. If this interpretation does not or not any longer correspond to the political will of the legislature, it may amend the law with effects for the future.

The legislature’s retroactive clarification of the legal situation generally also qualifies as constitutive retroactive regulation if the retroactive regulation decides on a question of interpretation that is controversial amongst the regular courts and has not yet been settled by the highest courts. A clarifying provision is already constitutive if it intends to exclude an interpretation by the regular courts – be it only by the lower courts – by retroactive intervention in a settled matter. By aiming to retroactively clarify a legal situation, which is apparently unclear or at least non-uniform in terms of its application, with the introduction of a law whose relevant statement is now clear-cut, the legislature gives constitutive effect to the retroactive law.

In these cases, the Federal Constitutional Court only decides on the constitutionality of the retroactivity, not on the binding interpretation of the ordinary law that the legislature intended to change retroactively. It is not the task of the Federal Constitutional Court to interpret ordinary law in those cases where its disputed interpretation has not yet been clarified by the highest courts. To establish a constitutive retroactive amendment of the law, it is sufficient that the referring court takes a justifiable view on the interpretation of the former law that the legislature wants to exclude by means of the new retroactive regulation. This legal viewpoint does not require settled case-law or confirmation by the highest courts. What is decisive is that the legislature wants to correct or exclude it.

With its finding that the new regulation has constitutive retroactive effect, the Federal Constitutional Court does not adopt a decision on whether the citizen or the public authority rightly took a particular view on the old legal situation in the dispute in the initial proceedings. If the Federal Constitutional Court – as in the present case – considers the retroactivity unconstitutional, it remains within the competence of the regular courts to clarify the content of the former legal situation by means of interpretation. This corresponds to the division of functions between the Federal Constitutional Court and the regular courts. The further interpretation by the regular courts and the highest courts in particular, may show that the provision is in fact to be understood as the legislature intended to have it established ex post facto . This, however, remains a question of interpretation of the applicable law that is not the task of the legislature, but that of the judiciary and thereby, primarily, the regular courts.

(2) On the basis of these principles, the retroactive “clarification” of the applicability of § 8b sec. 3 KStG (f.v.) in § 40a sec. 1 sentence 2 KAGG is constitutive. When adding sentence 2 to § 40a sec. 1 sentence 1 KAGG, the legislature expressed its intention of resolving the resulted problem of interpretation (cf. BTDrucks 15/1518, p. 17). Before the legislature’s clarification, § 40a sec. 1 KAGG could justifiably be interpreted both in the sense that it allowed for an application of § 8b sec. 3 KStG (f.v.) as well as in the sense of its non-application. The fact that a judicial decision on that question had not yet been taken by the time the Korb II-Act was promulgated in the Federal Law Gazette does not justify a different view with regard to the constitutional standards to be applied. Because also here the legislature inserted the provision intended to provide subsequent clarification in the law retroactively, acting in response to a legal situation it regarded as unclear and thus in a situation of uncertainty. This uncertainty has later been confirmed by the diverging decisions of the finance courts […].

Also in these constellations in which the question of whether a new provision is declaratory or constitutive in nature, the Federal Constitutional Court is competent to interpret the ordinary law itself as basis of its decision, for example because the legal point of view of the referring court is constitutionally objectionable. However, it is not obliged to do so. In the present case, there is no reason for the Federal Constitutional Court to interpret the ordinary law on its own, for there is no indication that the legal standpoint taken in the order of referral on the non-applicability of § 8b sec. 3 KStG (f.v.) could be unconstitutional.

3. The burden imposed by the constitutive effect of § 40a sec. 1 sentence 2 KAGG is unconstitutional insofar as it has, pursuant to § 43 sec. 18 KAGG, “true” retroactive effect with regard to the 2001 and 2002 assessment periods.

The principles of legal certainty and legitimate expectations, which are anchored in the rule of law and the fundamental rights, generally oppose laws with “true” retroactive effect (a). In the present case, none of the exceptions recognised by the relevant case-law are applicable (b). Furthermore, there is no other apparent reason for a justification of the “true” retroactivity here (c).

a) The constitutionality of a retroactive law is only questionable if the law poses a burden upon the citizen (cf. BVerfGE 24, 220 <229>; 32, 111 <123>; 50, 177 <193>; 101, 239 <262>; 131, 20 <36 and 37>). The general prohibition of “true” retroactive burdensome laws is based on the principles of legal certainty and legitimate expectations (cf. BVerfGE 45, 142 <167 and 168>; 132, 302 <317>). It protects the confidence in the reliability and predictability of the legal order established under the Basic Law and the rights acquired on its basis (cf. BVerfGE 101, 239 <262>; 132, 302 <317>). If the legislature subsequently changes, in a burdensome manner, the legal consequence of a conduct that lies in the past, this requires a special justification with a view to the rule of law and the fundamental rights of the Basic Law (cf. BVerfGE 45, 142 <167 and 168>; 63, 343 <356 and 357>; 72, 200 <242>; 97, 67 <78 and 79>; 132, 302 <317>). Both the fundamental rights and the rule of law conjointly guarantee the reliability of the legal order, which is an essential precondition for the self-determination of one’s life plan and thus a fundamental condition of liberal constitutions. It would seriously undermine the individual freedom of the parties concerned if the public authority were easily able to impose, ex post facto , legal consequences that are linked to the parties’ conduct or personally relevant factors and that are more burdensome than those applicable at the time of the legally relevant conduct (cf. BVerfGE 30, 272 <285>; 63, 343 <357>; 72, 200 <257 and 258>; 97, 67 <78>; 105, 17 <37>; 114, 258 <300 and 301>; 127, 1 <16>; 132, 302 <317>). Against that background, laws with “true” retroactive effect are generally not compatible with the Constitution (cf. BVerfGE 45, 142 <167 and 168>; 101, 239 <262>; 132, 302 <318>; established case-law).

b) aa) However, there are exceptions to this general prohibition of “true” retroactive laws (cf. BVerfGE 13, 261 <272 and 273>; 18, 429 <439>; 30, 367 <387 and 388>; 50, 177 <193 and 194>; 88, 384 <404>; 95, 64 <86 and 87>; 101, 239 <263 and 264>; 122, 374 <394 and 395>; 126, 369 <393 and 394>; 131, 20 <39>; established case-law). The principle of the protection of legitimate expectations not only forms the basis for the prohibition of retroactivity, but also determines its limits (cf. BVerfGE 88, 384 <404>; 122, 374 <394>; 126, 369 <393>). The prohibition of retroactivity does not apply insofar as legitimate expectations in the persistence of the applicable law could not be developed (cf. BVerfGE 95, 64 <86 and 87>; 122, 374 <394>) or where confidence in a particular legal situation was not objectively justified and therefore not worthy of protection (cf. BVerfGE 13, 261 <271>; 50, 177 <193>). The case groups recognised but not conclusively defined by the case-law of the Federal Constitutional Court are standardised cases of an exceptional lack of confidence in the existing legal situation (cf. BVerfGE 72, 200 <258>; 97, 67 <80>). To answer the question whether a retroactive amendment of the legal situation was reasonably foreseeable it is of significance whether the previous regulation would, objectively speaking, result in the concerned persons’ legitimate expectations in its continued applicability (cf. BVerfGE 32, 111 <123>).

One exception to the general inadmissibility of “true” retroactivity is made when the parties concerned, already at the time the retroactivity refers to, could not have reasonably relied on a legal provision’s continued application but in fact had to expect its amendment (cf. BVerfGE 13, 261 <272>; 30, 367 <387>; 95, 64 <86 and 87>; 122, 374 <394>). There is no room for a protection of expectations where the legal situation was so unclear and confusing that a clarification had to be expected (cf. BVerfGE 13, 261 <272>; 18, 429 <439>; 30, 367 <388>; 50, 177 <193 and 194>; 88, 384 <404>; 122, 374 <394>; 126, 369 <393 and 394>) or where the previous law was so incompatible with the system and unfair that serious doubts as to its constitutionality had existed (cf. BVerfGE 13, 215 <224>; 30, 367 <388>). Furthermore, the protection of existing expectations has to step back where overriding concerns of general interest prevailing over the principle of legal certainty require a retroactive elimination (cf. BVerfGE 13, 261 <272>; 18, 429 <439>; 88, 384 <404>; 95, 64 <87>; 101, 239 <263 and 264>; 122, 374 <394 and 395>), where the citizen could not reasonably rely on the existence of a right that did not in fact exist, given that the provision providing for this ostensible right was invalid (cf. BVerfGE 13, 261 <272>; 18, 429 <439>; 50, 177 <193 and 194>; 101, 239 <263 and 264>; 122, 374 <394 and 395>) or where the retroactive amendment of the law which was objectively justified causes no or only insignificant damage (so-called minimum threshold - Bagatellvorbehalt , cf. BVerfGE 30, 367 <389>; 72, 200 <258>).

bb) Of the case groups of permissible “true” retroactive laws recognised by the case-law, the only ones relevant here are that of an unclear and confusing previous legal situation or that of its incompatibility with the system and unfairness. Neither of them can justify the retroactive effect of § 43 sec. 18 KAGG on the 2001 and 2002 assessment periods.

(1) (a) A provision’s need for interpretation alone does not justify its retroactive amendment; only if the openness to interpretation has reached a degree which leads to a confusing legal situation, the legislature may enact a new clarifying regulation which has effect for the past.

The case groups recognised in the case-law of the Federal Constitutional Court determining exceptions to the prohibition of “true” retroactive laws all have in common that particular circumstances prevent the development of legitimate expectations in the applicable law in the first place or that they destroy existing legitimate expectations. The mere openness of, and need for, interpretation of a provision and the resulting insecurity as regards its content is not such a particular feature that could destroy legitimate expectations. Otherwise, especially in the early years of a statutory provision, one could generally never develop the legitimate expectation that there will be no retroactive amendments as long as there is no relevant established case-law on this matter yet.

If every apparent problem of interpretation were deemed to prevent the development of constitutionally protected legitimate expectations, the legislature would, to a large extent, be free to amend the law with retroactive effect as soon as this appears to be opportune – for instance because the jurisdiction interprets the law in a way which does not correspond to the legislature’s ideas and expectations. In that case, the legislature may at any time take the initiative and amend the applicable law with effects for the future, as long as it adheres to the requirements of the Basic Law. However, a legal situation that is in need of interpretation and to this extent thus unclear does not imply a “carte blanche” for retroactive amendments of the law. Such a far-reaching competence of the legislature to enact laws with “true” retroactive effect would largely debase the confidence in the applicable law as protected by Art. 20 sec. 3 GG.

In addition, a competence of the legislature to retroactively clarify provisions that turn out to be in need of interpretation that is not limited to particular exceptional cases would undermine the competence of the judiciary under the Basic Law to interpret the law with binding effect (cf. BVerfGE 126, 369 <392>).

Since questions of interpretation often arise with regard to newly enacted laws, there would be the danger that the rule-exception-ratio would be distorted in the case of “true” retroactivity in the sense that it would then not remain generally impermissible, but – like the “quasi” retroactivity – be generally permissible. Such an outcome would be incompatible with the constitutional principles of the protection of legitimate expectations and legal certainty.

(b) On the contrary, an uncertainty of a legal situation that suffices to justify a “true” retroactive clarification of the law requires additional qualifying circumstances under which the applicable law appears to be so confusing that it cannot provide the basis for a constitutional protection of legitimate expectations. This can be the case in particular if the meaning of the legal provision in question remains completely incomprehensible even bearing in mind its wording, its systematic context and purpose.

(c) Before the insertion of sentence 2 that is subject to review here, § 40a sec. 1 KAGG allowed for different interpretations. The diverging decisions of the finance courts on the interpretation of this provision prove this. The lack of an interpretation by the highest courts as to whether § 8b sec. 3 KStG (f.v.), which was not explicitly mentioned in the original wording, was applicable and the insofar inconsistent case-law of the finance courts, however, do not yet qualify as a confusing legal situation. In respect of its wording and content, the legal provision was not doubtful or even incomprehensible, but had been drafted clearly. Its need of interpretation, in particular in respect of the systematic link to § 8b KStG, has led to diverging, yet as such justifiable point of views. However, this does not justify a “clarification” by a “true” retroactive law.

(2) The original statutory law was also not incompatible with the system or unfair to an extent that would justify the “true” retroactivity prescribed in § 43 sec. 18 KAGG.

Neither the interpretation of the referring Finance Court (no application of § 8b sec. 3 KStG) nor the opposite interpretation of the former § 40a sec. 1 KAGG (application of § 8b sec. 3 KStG) is constitutionally mandated. In the present context, when interpreting the former § 40a sec. 1 KAGG, systematic and teleological aspects may indeed provide good reasons for a result contrary to the literal interpretation on the basis of the wording and the applicability of § 8b sec. 3 KStG (f.v.); […]. Nonetheless, the perspective of the referring Finance Court also does not lead to a result, which is, to an extent raising serious doubts as to its constitutionality, incompatible with the system or unfair (cf. BVerfGE 13, 215 <224>; 30, 367 <388>).

[…]

c) In the present case, there are no other reasons exceptionally justifying a regulation with “true” retroactive effect beyond the case groups recognised in the case-law of the Federal Constitutional Court. […]

[…]

d) There is no reason to deviate from the protection of legitimate expectations anchored in the rule of law and the fundamental rights and the resulting exceptional nature of permissible “true” retroactivity in those cases in which the legislature wishes to clarify the applicable law for the past. It would, however, amount to such a deviation if the legislature’s wish to determine ex post facto the “true” content of a previously enacted law and to correct – also for the past – an interpretation that does not conform with its ideas was only limited in case of individual proceedings that have become final or non-appealable or as regards legal situations that do not leave serious room for interpretation. This would sacrifice both the special protection against “true” retroactive laws, which the Court developed in its established case-law, and the differentiation between generally impermissible “true” retroactivity and generally permissible “quasi” retroactivity.

III.

Insofar as § 43 sec. 18 KAGG results in the application of § 40a sec. 1 sentence 2 KAGG in the 2001 and 2002 corporate tax assessment periods, this provision violates the constitutional principles of the protection of legitimate expectations and is void (§ 78 sentence 1 in conjunction with § 82 sec. 1 BVerfGG).

IV.

The decision was adopted with 5:3 votes regarding its final conclusion, and with 6:2 votes regarding the constitutional principles.

Separate opinion of Justice Masing

on the order of the First Senate of 17 December 2013

– 1 BvL 5/08 –

I cannot agree with the decision. Contrary to the first impression, the decision does not concern specific problems of statutory law, but fundamental questions regarding the reach of the legislature’s authority to regulate unclear legal questions of the past that have remained unresolved – here, the possibility of tax-deductions for losses that financial institutes incurred in particular as a consequence of the attacks of 11 September 2001. Referring to the prohibition of retroactivity, the Senate enjoins the legislature from enacting a clarification to the effect that these losses may not be passed on to the general public.

It thereby substantially changes the foundation on which the case-law on retroactivity resides. The Senate deprives this case-law – not according to its self-perception, but as a result and through the reasoning – of its foundation which is enrooted in the protection of legitimate expectations and replaces it with abstract and substantially misguided ideas of the separation of powers. The safeguarding of an objective area of competence of the regular courts takes the place of the protection of individual freedom. To the detriment of Parliament: By means of a revaluation of the former case-law, the Parliament is deprived of a means to clarify retroactively unresolved legal questions not only if this would be at odds with the citizen’s trust, but in fact generally. As a matter of principle, the Parliament forfeits a way of taking over political responsibility for old cases. This constitutes a serious disruption of the balance between the principles of democracy and the rule of law.

I.

The Senate repeals the challenged provision on account of a violation of the prohibition of retroactivity even though it also assumes that the original legal situation did not inspire confidence in the complainant, which the amendment of the law could have disappointed. It thereby deprives the prohibition of retroactivity of its foundation that is intended to safeguard individual freedom.

1. The subject matter of the legal provisions at issue is whether capital companies – in practice in particular banks – are entitled to deduct for tax purposes capital losses of their shares in investment funds in the years 2001 and 2002, whereas the law generally exempts profits from taxes. The Senate as such holds that the claimant in the initial proceedings at no time had legitimate expectations in that she could deduct from her profits partial write-downs for this period. It rightly finds that the original state of the law in this respect was unresolved and recognises that it was both subjectively and objectively unclear for the banks concerned. Nonetheless, the Senate holds that the legislature was not entitled to clarify this with retroactive effect for the unclear old cases and states that it was the task of the regular courts to decide on those cases.

This argumentation puts an end to basing the prohibition of retroactivity on the protection of legitimate expectations: the Senate expressly finds that in the cases in question the regular courts could come to the constitutionally unobjectionable conclusion that, also irrespective of the legal clarification in § 43 sec. 18 KAGG, the former § 40a sec. 1 sentence 2 KAGG was to be properly interpreted in that it is not possible to deduct the partial write-downs asserted by the complainant from the profits. According to the Senate, this question could not, however, be clarified retroactively by the legislature; clarification was reserved for the regular courts alone. Hence, the legislature is thus enjoined from adopting a regulation that the courts are readily permitted to establish by means of interpretation. Although, on the basis of the former legal situation, the citizens could not have confidence in the legal situation in that this would induce them to make arrangements in respect of insofar foreseeable legal consequences, the legislature shall nonetheless be barred from providing a clarification under the prohibition of retroactivity derived from the protection of legitimate expectations. The instruments of the protection of legitimate expectations are therefore taken as an argument against the legislature when ordering legal consequences, which the parties concerned had to anticipate under the old law already and still have to anticipate.

2. The only form of expectation one might deem to have been established here is an abstract degree of confidence in the validity of a law with an open content – and therefore confidence that the regular courts will settle a question that has remained politically unresolved. The Senate’s decision protects the concerned parties’ expectation to have a chance of being subject to a favourable case-law. However, this very aspect shows how far the Senate diverges from the original purpose of the case-law on retroactivity. The prohibition of retroactivity no longer safeguards legitimate expectations in a predictable legal order ensuring the individual’s exercise of freedom in view of foreseeable legal consequences, but merely the chance that the judiciary renders a more favourable clarification of the uncertain position than a democratic-political decision of the Parliament. While the case-law on retroactivity initially protected legitimate expectations in order to safeguard the exercise of individual freedom, it now serves as safeguard of an area of competence for the judiciary versus the legislature. The protection of individual freedom is replaced by an enforcement of objective ideas on the separation of powers and thereby an area of competence of the judiciary.

II.

This notion of the relationship between the legislature and the judiciary as asserted by the Senate cannot be derived from the order of the Basic Law. It unjustifiably releases the legislature of its responsibility.

1. By detaching the prohibition of retroactivity from the trust in a particular legal situation, it is turned into an independent a priori principle of separation of powers in cases of “true” retroactivity, the purpose of which is to generally rule out the legislature’s retroactive intervention in open and unresolved legal questions. Instead of providing for the Parliament to adopt a political decision, it is now for the courts to adopt a non-political decision instead.

a) This is not convincing, not even in terms of its general approach. The legislature may deal with all urgent issues of the community on the basis of the Parliament’s legislative right of intervention that follows from the principle of democracy. In a democratic state under the rule of law it is generally for the legislature that is elected for this purpose and is responsible to the public in a political process to decide on what constitutes law. In principle, this also concerns decisions about problems that originated in the past, or covers the clarification of contentious issues that have remained unresolved and are in need of a solution. The assumption that this democratic responsibility is from the outset limited to the future is not duly justified and in particular cannot be derived from the previous case-law. In particular, neither the idea of a limited historical context of legitimisation nor the dignity of the legislature elected for a certain period of time can be invoked to justify this assumption. For also such retroactive legislation is about addressing problems that have not been dealt with appropriately in the past and which now – being unresolved and in need of solution – affect the present and the future.

Admittedly, Parliament’s legislative powers are indeed limited by the rule of law and these limits may be reached more quickly with laws that have effect for the past than with others. Hence, the legislature cannot, of course, simply interfere ex post facto in individual proceedings that are final or reassess conduct of the past and impose sanctions on that conduct in a way which the parties did not have to anticipate. In particular, the fundamental rights and the derived presumption of freedom often set limits in this respect. This in fact is the correct core of the case-law on retroactivity. Such limitations of the legislature must, however, be based on a specific need for protection of those affected. They do not already follow generally from an abstract limitation of the legislature’s competence. There is no reason why the legislature should not be permitted to regulate disputed and unclear legal questions arising in unresolved old cases, as long as legitimate expectations are not frustrated. The fact that the legislature might better discern conflicts of interest resulting in the context of laws that have consequences for the past than in the context of those that have consequences for the future does not render the legislature’s clarification impermissible. In a modern state, legislation is generally not confined to the adoption of future-oriented regulations that lack awareness of the specifics of a particular situation, but almost always intends to reach a specific reconciliation of interests.

b) On the contrary, the separation of powers clearly speaks in favour of the general permissibility of a retroactive clarification of open and disputed questions in unresolved legal situations by the legislature. If the practical application of a law shows that important questions of general relevance remained unresolved or that legal provisions are formulated in an unclear or misleading manner, the Parliament is called upon to clarify, in the exercise of its political-democratic responsibility, how these questions are to be answered in the ongoing proceedings. The assumption that the legislature is only entitled to one attempt and is then barred from clarifying problems arising in the course of time up until the revision of the law for the future is not supported by the legitimising foundations of our Constitution. In particular, such an assumption cannot be based on ideas of temporally segmented legitimisation contexts, for the former legislature did not in fact clarify the respective questions. This becomes particularly obvious if the legislature, as in the present case, only wants to enshrine its – more than understandable (see below IV.) – understanding of what the former regulation already intended to achieve anyway.

2. Also functionally, the distinction between legislation and jurisdiction created by the Senate is not convincing.

a) In view of the increasingly complex requirements that the legislature must meet in a highly differentiated and sectoral as well as internationally manifold interconnected world, one cannot reasonably expect that it will always be possible to foresee all consequences of a legislative endeavour. In the network of interests of the numerous users and the persons concerned, legal provisions can cause unforeseeable misunderstandings, ambiguities or unintended practices. It also needs to be taken into account that the legislature might make mistakes or that inaccuracies occur. A legal reform as it underlies the disputed provisions in the present case makes this particularly clear. The legislature took on the Herculean task to fundamentally revise the entire corporate tax law – from the credit procedure (Anrechnungsverfahren ) to the half-income procedure (Halbeinkünfteverfahren ) and hence the taxation of almost all important companies – with far-reaching impacts on the corporate structure as well as on international contexts. In that context, the legal provisions at issue here were only one small, subordinate aspect. It is obvious that such an undertaking cannot ensure that all questions are resolved immediately and in a clear, well-grounded and unambiguous manner – and all the parties concerned had to be aware of this.

b) According to the view of the Senate, all respective problems generally have to be clarified exclusively by the courts until a revocation for the future. The Senate assumes that although the legislature may amend evolving uncertainties pro futuro , statutory incompatibilities and disputed issues, which arise under a given legal situation, are –extreme exceptions aside (see IV. 3 below) –to be dealt with by the courts exclusively.

This is not convincing with regard to the task conferred upon the courts in a constitutional state upholding the separation of powers: while the courts should seek to identify the legislative intention in view of unclear legal situations and, if this is not possible, could feel compelled to establish their own democratically unguided ideas of justice, the legislature is deprived of the possibility of realising such a clarification in order to relieve the courts.

Such an approach is not convincing with regard to the practical consequences either. A retroactive clarification by the legislature can resolve uniformly and all at once all pending disputes for the future and the past and establish legal certainty. Yet as a consequence of the decision of the Senate, all cases that have arisen before the amendment of the law must proceed through the entire appeals process in courts. This might take years, burden the courts with many proceedings, involve high costs for the parties concerned and result in a long period of fragmentation of the law und uncertainty. The chance for the citizen, as created by the Senate, to have a favourable judicial decision is therefore not only an opportunity, but also a significant burden – not only for the general public, but also for the parties concerned themselves.

III.

Thus the decision furthermore incorporates a far-reaching change in the case-law on retroactivity and a breach with the previous respective assessments.

It is true that the Senate ties in with some general statements of the Federal Constitutional Court’s prior case-law: the general impermissibility of “true” retroactivity corresponds to the established – an in its previous context correct – case-law. As the numerous citations of case-law show, the Senate merely intends to further develop them consistently. However, this is not done in a convincing manner. For it separates the general statements from their prior link to the principles of the protection of legitimate expectations and further develops them, turning them into independent, separate abstract rules. This accords them a new meaning that is much stricter and breaks with the assessments of the Court’s previous decisions.

1. a) The Federal Constitutional Court so far has been reluctant when it comes to declaring a law void on account of a violation of the prohibition of “true” retroactivity. In the history of the Federal Constitutional Court there are only two such cases and they date back a long time (cf. BVerfGE 18, 429; 30, 272). Naturally, the legal doctrine has been further developed since and the reasoning would perhaps be more differentiated nowadays. The crucial aspect, however, is unambiguous: in both cases the Court expressly referred to a specifically confidence-inspiring legal situation.

[…]

b) The Federal Constitutional Court especially referred to the frustration of specific legitimate expectations under the former legal situation in those cases in which it considered that the laws with “quasi” retroactive effect were unconstitutional. Since a “quasi” retroactive effect is generally permissible and deemed unconstitutional only in the case of exceptional legitimate expectations, specific legitimate expectations had to be established as a starting point already (see, with regard to the legitimate expectation in the lawfulness of an issue of an ID document for a refugee, BVerfGE 59, 128 <164 et seq.>; in the hitherto permissible revocability of pension payments granted voluntarily BVerfGE 74, 129 <155 et seq.>; in the continued application of tax regulations concerning compensation agreements BVerfGE 127, 31 <49 et seq.>). Nothing else applies in the insofar special cases that approach the cases of “true” retroactivity, in which retroactive changes were in question for an ongoing tax assessment period and were declared unconstitutional (cf. BVerfGE 72, 200; 127, 1; 127, 61; BVerfG, order of the First Senate of 10 October 2012 - 1 BvL 6/07 -, Neue Juristische Wochenschrift - NJW 2013, p. 145 et seq.). There, one might, on a formal view, discern a certain relativisation of the criterion of legitimate expectations, since the individual should always expect retroactive amendments of the provisions for the taxable period; by finding that it is partly not tolerable, the judiciary specifically connects the doctrine on retroactivity with the separate protection aspect of legal certainty in these cases. However, and without analysing these decisions under all aspects, also those cases have in common that their underlying legal situation was at least always clear initially which as such could have given rise to legitimate expectations to make arrangements and which was then amended retroactively by the legislature. Also in these cases, the protection of specific legitimate expectations is the very essence of the case-law.

There are no further cases in which the Federal Constitutional Court repealed a law on account of a violation of the prohibition of retroactivity. In particular, there is no case in which the clarification of an uncertain legal situation, which could not have induced legitimate expectations, has been declared unconstitutional.

2. The fact that valuations of the previous case-law are given up is even more obvious if one conversely looks at those cases, in which the Federal Constitutional Court considered retroactive laws for the clarification of open legal questions to be constitutionally unobjectionable. It is sufficient to consider those cases in which the legislature merely intended to reinforce the previous legal situation in response to unforeseen developments in the application of the law, and therefore assumed that the clarification is a means of “authentic interpretation”. They reveal that in the present case the Senate applied much stricter standards than in the case-law so far, also with regard to its substantive assessment of the case,

a) In fact, however, the Federal Constitutional Court has repeatedly held that a competence of the legislature to adopt an “authentic interpretation” is not recognised. The interpretation of unclear legal provisions is generally incumbent upon the courts (cf. BVerfGE 126, 369 <392>; 131, 20 <37 et seq.>; similarly already BVerfGE 111, 54 <107>). So far, also this statement was, however, always integrated in the context of the protection of legitimate interests. It merely opposes a view that “authentic interpretation” is recognised as a distinct title justifying retroactive laws. Its purpose was to refer back to the general – and thus to the protection of confidence-related – principles of retroactivity. The Senate therefore expressly stated that “legislation resulting from a conflict of interpretation that arose between the legislature and the judiciary is not to be assessed differently than a retroactive amendment of the law based on other reasons” (BVerfGE 126, 369 <392>).

b) Accordingly, the previous case-law deemed the retroactive clarification of open legal questions to be constitutionally unobjectionable in all cases in which there was no legal situation that could have conferred legitimate expectations

[…]

c) When comparing these cases with the present case, in which the legal situation is not even clarified by the highest courts, manifoldly disputed between legal literature and the regular courts and can thus be described by all means as open, it becomes obvious that, in light of the criteria of the previous case-law, legitimate expectations could not even nearly arise in the present proceedings. Also the subject matter does not give rise to any reason why the expectations of banks that they could, in parts, pass on their losses to the general public should be subject to a higher degree of protection than the expectations of pensioners or civil servants in a favourable calculation of their pensions. With this decision, the Senate in fact fundamentally reassesses the prior standards.

3. One cannot contradict by arguing that by consistently applying the criterion of legitimate expectations one would render the principle of the prohibition of retroactive laws invalid and hence sacrifice the standards of protection established through case-law.

The examination of the case-law has demonstrated that these standards of protection have at least so far not been applied as strictly as now adhered to by the Senate and that the fundamental prohibition of “true” retroactive laws, including its exceptions simultaneously developed in the case-law, ultimately had the effect that aspects of legitimate expectations were taken into account and were of crucial influence. The distinction between “true” and “quasi” retroactivity was more a heuristic distinction than a categorical – a distinction that is now altered under the present decision.

If one consequently invokes the protection of legitimate expectations, the legislature will not have an opportunity in the future to claim at its discretion the need for clarification in view of the general need to interpret the law and thus to easily reverse legal decisions ex post facto . Although the law mostly needs interpretation in individual cases, it nonetheless does not follow from such an abstract statement that fundamental statutory decisions and the rules on their implementation are in principle open for unlimited interpretation. One hardly has to assume that our legal system is generally unable to establish specific legitimate expectations in certain legal consequences, or to establish principles on which arrangements can be based. In all cases however, in which the law in fact creates such legitimate expectations – and the Federal Constitutional Court is then called upon to decide on this – the principle of the prohibition of “true” retroactivity rightly applies. Insofar, presumptions under the fundamental rights in favour of individual freedom already ensure that the prohibition of retroactivity is not ineffective. Besides, the case-law material of the Federal Constitutional Court shows that the adoption of laws with retroactive effect cannot be justified in general or in a large number of cases as a clarification of unresolved questions of interpretation. The criterion of legitimate expectations does not erode the previous case-law, but rather forms its fundamental basis.

IV.

Also as far as their subject matter is concerned, the challenged provisions do not suggest that the banks that lodged the complaint were confident that their losses would be accounted for under tax laws. It is at least conceivable that, when applying a substantively appropriate interpretation, the former regulation of § 40a sec. 1 KAGG was in fact never guided by the intention of creating such an eligibility. In any event, however, the complainants undoubtedly had to anticipate such an interpretation and could not legitimately base any arrangements on an adverse understanding.

1. In this respect, expectations cannot simply be based on the ordinary meaning of the provision’s wording. The interpretation of such a provision requires a reasonable evaluation of the provision in its general context, having regard also to the historical development of the provision, its systematics as well as its object and purpose.

Although the wording of § 40a sec. 1 KAGG (f.v.) expressly only refers to § 8b sec. 2 KStG (f.v.), which prescribes the non-consideration of profits for tax purposes, but not also to its sec. 3, which prescribes the non-consideration of losses, this nonetheless does not necessarily mean that the reference cannot possibly be understood in a broader sense. Particularly in complex matters such as tax law, in which the legislature cannot envision every single case, it is for the regular courts to ensure that the provisions are not executed selectively and de-contextualised on the basis of a literal interpretation, but instead interpreted in light of their historical development, context and the main legislative concepts, giving them a sense and further developing them. The strict limitations of Art. 103 sec. 2 GG, according to which, in the special field of criminal law, in case of doubt uncertainties speak in favour of the citizen, do not apply here. The judiciary rather has to further develop and shape the balance of private and public interests inherent in the law in a way that does justice to both. […] The understanding which, according to the general rules of legal interpretation, in substance is most convincing and most likely reflective of the presumed intention of the former legislature […] is decisive.

2. Against that background, it had to be anticipated also under the former version of § 40a sec. 1 KAGG that this provision would be subject to the same interpretation in legal practice as in the legislature’s subsequent clarification. Also back then such an interpretation was at least rather likely.

a) […]

b) […]

c) It is not the task of the Federal Constitutional Court to conclusively clarify this question. This will be – as a consequence of the majority opinion in the Senate – the task of the regular courts. Considering the convincing arguments which, from the beginning, spoke in favour of exactly that interpretation which the legislature then also confirmed explicitly, the retroactive extension of this regulation to the open old cases cannot be deemed to constitute a violation of the principle of the protection of legitimate expectations. In this complaint, the banks had to anticipate that they would not be able to deduce for tax purposes their partial write-downs. This is particularly true when taking into account that the questions of interpretation at issue here were known for a long time already and in general concerned companies, in particular banks with competent legal divisions, that are most likely well capable of detecting such ambiguities more quickly than the finance authorities themselves.

3. In this situation it is not persuasive that the legislature is enjoined from adopting a regulation that covers the old cases too. As mentioned, the assumption that the regular courts are in principle entrusted with settling these cases is not convincing. The delimiting criteria for the recognition of exceptions additionally adduced by the Senate are not convincing either.

a) According to the Senate a retroactive regulation shall be excluded because the former legal situation, although unresolved and unclear, was nonetheless accessible by normal means of interpretation. The Senate assumes that a retroactive regulation is only constitutionally permissible if the former regulation would have led to an entirely incomprehensible or unclear legal situation. This is only deemed to be the case if the meaning of the provision in question is totally incomprehensible or if the provision is totally confusing. Thus, the legislature may not clarify itself what it considers an editorial mistake, because this was too marginal. It would have had to commit more serious mistakes causing more confusion – in which case also the Senate assumes that it would have been allowed to enact a retroactive regulation. Such distinctions are not convincing.

b) The Senate’s distinction between an unresolved legal situation, which does not justify retroactive laws, and that of an unclear and confusing legal situation, which can justify such retroactive laws is a normative question that leaves a margin for future cases. One can only hope that through this the wrong path the Senate chose with the present decision can still be corrected and that, in hindsight, it constitutes a one-off case.

However, the present decision must be objected to in particular if it remains a one-off case. In this particular case, there are particularly few reasons for departing from the case-law: the present case concerns the expectation, in particular that of banks, that losses will be tax-deductible - in a business area which is, overall, speculative in nature. Essentially it concerns the losses incurred in the years in question here that resulted from the price drops caused by the attacks of 11 September 2001. It is not intelligible why the legislature is imposed with requirements in this particular case that are stricter than those imposed in cases that deal with employees’ access to insurance, retirement benefits or the amount of civil servant pensions.

V.

Rightly, § 43 sec. 18 KAGG would have had to be declared constitutional. In this respect it is of little importance whether the unresolved interpretation of § 40a sec. 1 KAGG (f.v.) already leads to the conclusion that the given legal situation is not amendable and that there is thus no retroactivity, whether one assumes that the case involves a merely formal retroactivity which is justified by the unclear legal situation (cf. BVerfGE 126, 369 <393 et seq.>), or finds that the case involves a retroactive effect which, beyond the alternatives of “true” and “quasi” retroactivity or declaratory or constitutive amendment of the law, is to be resolved directly by reference to the open legal question (cf. BVerfGE 50, 177 <193 and 194>; 131, 20 <40 et seq.>). What is decisive here, however, is only that the decision is ultimately guided by the question whether and to what extent the given legal situation gave rise to legitimate expectations. In the case at issue here, such legitimate expectations simply do not exist.

The only way to ensure that the case-law on retroactivity retains its character of protecting individual freedom and self-determination is to consequently consider whether a given legal situation gave rise to legitimate expectations. With the present decision, the Senate deforms the case-law on retroactivity, turning it into an instrument to safeguard the prerogative of the judiciary. The legislature is relieved of its responsibility and the realm of political-parliamentary decisions is narrowed down unjustifiably. This does not reflect the idea of democracy under the Basic Law.