单一罪, 即行为人实施一个犯罪行为或几个具有内在联系的犯罪行为而具备一个独立的犯罪构成。

数罪，即行为人在判决宣告以前，实行几个具有不同犯罪构成的犯罪行为;或者是实行几个同种犯罪构成的犯罪行为，但是这些行为之间缺乏内在联系。

如果在前一个犯罪行为受到宣判以后，刑罚还没有执行完毕又犯新罪的，也是数罪。

区分单一罪与数罪是涉及量刑和计算刑法时效的重要问题。

在刑法理论上,大致有4种主张:

①以行为的单复数作为罪数的标准。认为犯罪是一种行为，因此罪数应决定于行为数。这种主张是从刑法的客观主义理论出发的。

②以结果的单复数作为罪数的标准。认为法律使犯罪人受到刑罚制裁，是因为犯罪人侵犯了法律保护的客体,因此,罪数应决定于犯罪的结果数。这种主张也是从刑法的客观主义理论出发的。

③以犯意的单复数作为罪数的标准。认为犯罪行为是表明犯罪人恶性的手段，犯罪结果是证明犯罪人恶性的条件，二者都只是犯意的具体体现，不是犯罪的本质，不足以决定罪数，而犯意的单复数，才是决定罪数的本质。这种主张是从主观主义理论出发的。

④以犯罪构成的单复数作为罪数的标准。

认为上述3种理论都是片面的,纵观犯罪的本质，必须既考虑到犯罪客观方面的要件，也要考虑到犯罪主观方面的要件，因此，主观、客观相结合的犯罪构成才是罪数的标准。

________________________________________________________

区分一罪与数罪的标准，首先依刑法的特殊规定，即法定标准。刑法规定数行为为一罪的按一罪，规定按数罪的依照规定；通常采用犯罪构成说为区分标准；继续犯、想象竞合犯、结果加重犯按实质的一罪论处；结合犯、惯犯按法定的一罪论处；连续犯、吸收犯、牵连犯按处断的一罪论处。

区分一罪与数罪的意义

数罪是指一个人所犯之罪的个数是两罪以上（包括两罪）的犯罪。区分一罪与数罪就是区分罪数。行为人的行为究竟构成一罪还是成立数罪，是司法实践中经常遇到的问题，正确区分一罪与数罪具有重要意义。

正确区分罪数，有利于准确定罪。准确定罪的含义，除了包括准确地认定行为是否构成犯罪，构成何种犯罪之外还包括准确地认定行为是构成一罪还是数罪。这三者又是密切联系的。一方面如果没有正确区分罪数，定罪就不准确；另一方面如果没有正确区分罪数，就会影响罪名的确定。例如，行为人以抢劫的故意，持刀杀死被害人后，立即取走其财物。如果认定为一罪，就是抢劫罪；如果认定为数罪，就可能是故意杀人罪与盗窃罪。

正确区分罪数，有利于适当量刑，有利于实现我国刑法罪行相适应的原则。刑罚以犯罪为前提，刑罚应与犯罪相适应，故对一罪只能一罚，对数罪应当并罚。将一罪定为数罪，常常会导致无根据的加重行为人的刑事责任；将数罪定为一罪时往往会导致无根据的减轻行为人的刑事责任。只有正确区分罪数，才能为适当量刑提供前提条件。

区分一罪与数罪的标准一:通常采用犯罪构成说。

根据我国刑法理论的通说,犯罪构成是刑法规定的,决定某一行为的社会危害性及起程度。而为该行为成立犯罪所必需具备的一切客观要件与主观要件的有机整体.因此，犯罪构成是认定犯罪的法律标准。犯罪构成有四个共同要件:即犯罪客体,犯罪客观要件,犯罪主体与犯罪主观要件.任何行为凡是符合某种犯罪构成的就成立犯罪,否则就不成立犯罪。行为符合一个犯罪构成的就是一罪,行为符合数个犯罪构成的就是数罪,行为数次符合同一个犯罪构成的也是数罪。但在适用这一标准时,涉及对犯罪构成本身与以及符合犯罪构成的认识.

犯罪构成是具有实质内容的。其实质内容又决定了它有特定的外延。因此,如果现实发生的事实完全属于某一犯罪构成所预定的内容,就应认为行为符合一个犯罪构成。例如,抢劫是故意以暴利胁迫或者其他方法强取公私财物的行为,因此，行为人故意以暴力方法强取他人财物时就符合抢劫罪的犯罪构成,不能认为其中的暴力行为另符合故意伤害罪或者故意杀人罪的犯罪构成,因为抢劫罪的犯罪构成所预定的内容包含了以暴力方法致人伤亡。

Most people assume that when offenders appear for sentencing they have been convicted of a single crime. In reality, almost half the offenders at sentence are sentenced for multiple crimes: these are multiple crime offenders. The court must now sentence for more than one offence. How should the sentencing court deal with such a case? That is, how severely should such an offender be punished relative to one who has committed only one of the crimes in the series, or relative to another who has committed a single crime which in itself is more serious than any of the crimes committed by the multiple offender?

Two approaches are possible. The first adopts a purely cumulative strategy: the seriousness of each crime and the punishment each crime merits are determined separately, and the overall punishment is simply the sum of all the individual punishments. For example, if residential burglary warrants one year in prison and the offender has committed five burglaries, s/he receives a five-year prison term. Despite its intuitive appeal and simplicity, the cumulative approach has not been adopted in penal practice. The reason for this is that an offender convicted of five burglaries and who receives a prison term of five years is being punished as severely as an offender convicted of a much more serious crime―manslaughter, for example. This violates the principle of proportionality, which requires the severity of the sentence to correspond to the seriousness of the offence for which it is imposed.

The second strategy ―which is enacted in the law of most countries―is to provide a ‘discount’ in cases where an offender has committed more than one crime, that is, to give what has been called a ‘bulk discount’ for multiple offending. For example, in the case of five burglaries, a court may impose a two-year sentence for the first burglary and then four further terms of one years for the other crimes, but these terms will be served concurrently. The total time in prison is therefore two years, a significant discount from the first example. The problem with this approach is that the seriousness of each individual crime is being depreciated―the offender is being ‘rewarded’ for committing all the crimes in one evening. Had s/he committed the burglaries once a month, and been prosecuted for each separately, the outcome may well be very different.

In fact, if you compare the offender who commits five burglaries in one evening with an offender who commits five burglaries over a two-year period, and who is sentenced for each consecutively, the difference between cases becomes even more striking. As a result of a recidivist premium, in many countries the repeat offender may receive sentences of one, one and a half, two, three, and four years for the five burglaries―adding up to over ten years for the same amount of criminality. Is this difference between the two offenders―one convicted of concurrent crimes, the other of consecutive crimes―defensible? What principles justify the different approaches? Can it be justified to give a penal reduction in cases in which a criminal has managed to commit a series of crimes before being apprehended? The purpose of this seminar was to consider these and the above questions and to provide an analysis of the practical and theoretical challenges of how the state should sentence multiple offending. The seminar and its attendees provided a very rich discussion with competing views as to the strongest justificatory basis for a multiple offence sentencing discount.

Turning to more basic penal theoretical considerations, it’s clear that the justificatory challenges arising in relation to the punishment of multiple offenders have been addressed by few theorists. From a consequentialist perspective―such as traditional utilitarian penal theory―a bulk discount may be justified by reference to the fact that there is no extra crime preventive effect related to the longer prison terms that would follow from the use of a purely cumulative scheme. Whether there are such desirable consequences is obviously an empirical question. However, given the dominance of retributivism in the modern era of penal theory, it is not surprising that the justification question has been addressed mainly from a desert-theoretical perspective. Two approaches have been defended.

One idea has been to invoke the concept of overall proportionality. While (ordinal) proportionality considerations usually concern the relative punishments of individual crimes, overall proportionality concerns the comparison of types of crime. More precisely, the idea is that no number of less serious crimes can be as morally wrong―and therefore deserve a more severe punitive response―as a single instance of a crime of a more serious type. For instance, it might be held that no number of offences of stealing cars can be regarded, as a consequence of cumulation, as the moral equivalent of a single rape.

A different proposal has been that a bulk discount for multiple offending is justified by reference to mercy. As an illustration, suppose that an offender, aged 25, has committed 100 separate residential burglaries and that each of these offences might, if committed as a single offence, be deemed to merit a custodial sentence of six month’s duration. One hundred times six months equals 50 years of nominal prison time, which may mean at least 25 years’ real time for the offender. However, this will deprive the offender of an excessive proportion of the prime years of his/her life. Therefore, the court should reduce the punishment out of mercy.

Determining rules to govern the sentencing of offenders convicted of multiple crimes is a therefore very challenging. To date, no volume has brought together scholars to address the complex issues involved―hence the need for this seminar and subsequent volume.

From: https://www.law.ox.ac.uk/centres-institutes/centre-criminology/blog/2016/01/more-one-crime-sentencing-multiple-offences