I am looking for other countries that allow reducing punishment for a crime through government (executive) ordinance.
That's surely possible in the US through presidential/executive commutation. Generally, in the US it's granted on a case by case basis. Obama granted more commutations than any prior president.
Ford and Carter outright pardoned Vietnam-era draft dodgers, thus completely absolving them of the crime before they were even convicted. (That's possible in the US.)
In the countries of the Council of Europe, commutation must be an option for life sentences. If a foreign country doesn't provide such an option, that even prohibits extradition to such a country.
On extradition and life imprisonment, the Court examines whether there is a risk that
the extradited person might be exposed to an irreducible life sentence, which as such, as we
have seen, is not compatible with the Convention (Vinter v. United Kingdom, cited above). In
the case of Trabelsi v. Belgium, 4 September 2014, the Court considered that the life sentence
which the applicant faced in the United States was irreducible inasmuch as US law did not
provide for an adequate mechanism for reviewing that type of sentence, which meant that his
extradition to the United States had amounted to a violation of Article 3 of the Convention.
It looks like what's happened in Romania was (in part) for the benefit of some corrupt politicians. The specter of something like that has been raised in the US recently with Trump mentioning he might pardon himself. As a high-level historical precedent, Ford pardoned Nixon, but this decision proved very damaging for Ford's subsequent presidential campaign. So yeah, there might be political consequences for actions that are nevertheless legal. (I'm curious how that turned out in Romania, I'll have to check out the recent election news.)
The 2016 affair in which some Romanian politicians got reduced sentences supposedly for writing books [that sometimes turned out to be ghostwritten or plagiarized] is perhaps also worth recalling here. I think that was rather more original and rather obviously an avenue for corruption itself.
As for the main question on the application of the more favorable sentence... there are some partial (and complicated) examples from the US, e.g. the Fair Sentencing Act of 2010. It
reduces the disparity between the amount of crack cocaine and powder cocaine needed to trigger certain federal criminal penalties from a 100:1 weight ratio to an 18:1 weight ratio and eliminated the five-year mandatory minimum sentence for simple possession of crack cocaine, among other provisions.
This law was not entirely retroactive as it was initially passed,
Effective November 1, 2011, the Fair Sentencing Act of 2010 applies retroactively to reduce the sentences of certain offenders already sentenced for federal crack cocaine offenses before the passage of the bill. However, the nonprofit organization Families Against Mandatory Minimums, a major advocate of the Fair Sentencing Act, has lobbied Congress to make the entire act retroactive.
The FIRST STEP Act, passed in December 2018, retroactively applied the Fair Sentencing Act, aiding around 2,600 imprisoned people.
Amusingly perhaps, the latter act also has non-retroactive sentence reduction provisions, although apparently the non-retroactivity of these was a mistake rather than intentional, this time:
The legislation increases the number of good conduct time credits that prisoners receive from 47 days per year to 54 days. Due to a legislative drafting error, this change is not being applied retroactively.
So as far as the US goes, there are examples and (intentional) counterexamples, i.e. it's not a universally applied principle that laws stipulating sentencing reductions apply to those already convicted.
And the fact that Congress needs to spell it out each time was actually mentioned in a Supreme Court case:
In Dorsey v. United States, the Supreme Court made clear that Congress
possesses sole responsibility for drawing the line between justice and finality in
sentencing. That is, when Congress passes a law reducing sentences, it must also
choose whether to apply those reductions retroactively or to accept the sentence
disparities between pre- and post-reform offenders that will otherwise inevitably
The case is rather recent (2012) and was one of those controversial 5-4 decisions. It involved exactly the application of the Fair Sentencing Act.
And shamelessly stealing your discovery of lex mitior as applying to European law... it does seem that reducing sentences that way (post-conviction) is actually the norm in Europe:
In 2009, New Mexico prospectively repealed the death penalty. Three years later in 2012, New Mexico prosecuted a defendant for a capital murder that was committed before repeal, and it sought to subject him to the death penalty. If state prosecutors had prevailed with the jury, they would have secured the very kind of sentence—death—that state officials had been lauded in Europe for outlawing three years earlier. A prosecution like New Mexico’s could never occur in Europe, and not merely because Europe has long outlawed the death penalty. It could never occur because, in contrast to the law of most American jurisdictions, European states embrace a doctrine known as “ lex mitior ” (“the milder law”). The latter doctrine is a counterpart to the ex post facto prohibition. Both doctrines concern retroactivity in criminal law, but they are the converse of one another. The ex post facto doctrine prohibits retroactivity by prohibiting the state from prosecuting persons under criminal statutes that either retroactively criminalize conduct that was hitherto lawful or retroactively increase penalties for conduct that, while unlawful all along, was hitherto punishable less severely. In contrast, lex mitior mandates retroactivity by mandating that criminal defendants receive the retroactive benefits of repealing statutes that either decriminalize conduct altogether or reduce punishment for it. After surveying laws in the United States regarding the retroactive effect of ameliorative repeals, the author addresses whether punishing offenders under harsher laws that obtained at the time of their conduct can serve consequentialist and/or retributive purposes of punishment. He concludes that, although doing so can be morally justified under limited circumstances, typically it is not—a conclusion that bears upon lex mitior’s proper scope, whether it consists of a binding norm (as it is among European nations), a nonconstitutional norm (as it presently is within the United States), or, when legislative intent is uncertain, a function of the rule of lenity.
Apparently, the expansion of lex mitior in Europe was made by the CJEU:
The principle of applying the more lenient sanction – also called principle of lex mitior - constitutes a general principle of national criminal laws as well as a general principle of Union law, as confirmed by the Court of Justice of the European Union over time.
The same article notes that Romania was even sued at CJEU for not applying the principle to some cases decided before it joined the EU.
Actually the more widely applicable ECHR doctrine includes it as well:
Scoppola v Italy (No 2) (2010) 51 EHRR 12 where the European Court
of Human Rights affirmed that the ‘fundamental principle’ of lex mitior was embodied within art. 7 (at ).
Also of interest are international conventions in this regard, the International Covenant on Civil and Political Rights in particular:
The international treaty law governing retroactive amelioration is explicit and clear. The major governing human rights treaties allow an offender to benefit from a change in law that imposes a lighter penalty than the one in existence at the time the offense was committed. Article 15 of the ICCPR, to which the United States is a party, contains a provision that prohibits
criminal ex post facto laws noting however, “[i]f, subsequent to the commission of the offence, provision is made by law for
the imposition of a lighter penalty, the offender shall benefit thereby.” The United States is one of 167 countries that are
party to the ICCPR. The ICCPR sets the main legal framework for the lex mitior principle under which the countries of
the world have fashioned their constitutions and penal codes. There has been little dispute from the international community
about this section of Article 15. Germany attached a reservation indicating that a lighter penalty would not be provided
in certain instances. Nonetheless, Germany enshrines the right to retroactive application of ameliorative law prior to final
judgment in its criminal code. Italy and Trinidad and Tobago specified that this right would only apply to cases still in
progress or prior to a final judgment. The United States is the only country that has attached a reservation indicating that
this section of the Article would not apply under any circumstance. The reservation states “[t]hat because U.S. law generally applies to an offender the penalty in force at the time the offence was committed, the United States does not adhere to
the third clause of paragraph 1 of article 15.” However, Article 4(2) of the ICCPR lists this Article as non-derogable, which
means that it may not be reduced in any way.
Text of ICCPR article 15:
No one shall be held guilty of any criminal offence on account of any act or omission which did not constitute a criminal offence, under national or international law, at the time when it was committed. Nor shall a heavier penalty be imposed than the one that was applicable at the time when the criminal offence was committed. If, subsequent to the commission of the offence, provision is made by law for the imposition of the lighter penalty, the offender shall benefit thereby.
Nothing in this article shall prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of law recognized by the community of nations.