Analysis of gender bias in sentencing data

All Data on this Page relate to England & Wales only

Where can discrimination arise?

  • Being arrested
  • Being prosecuted
  • Being found guilty
  • Sentencing

On this page the analysis addresses only the last of these: bias in sentencing. It should be borne in mind that biases in the likelihood of being arrested, being prosecuted once arrested, and being found guilty if prosecuted may also contribute to the overall bias. Hence the disparity based on sentencing alone, as considered here, may be an underestimate.

Two different analyses are presented. The first is based on all offences (summary and indictable offences). The second is based on indictable offences only. The second analysis, using indictable offence data only, is probably the more reliable and indicates a somewhat smaller gender disparity (though still large).

Analysis Using All Offence Data

The falling crime data indicated by the crime surveys (see Key crime and prison data) is reflected in the reducing number of people being sentenced. The numbers of men and women being sentenced between 2004 and 2014 are given below. Note that only a small fraction of these people will be sentenced to prison, the great majority receiving lesser sentences such as fines, community service, compulsory rehabilitation or suspended sentences.

Table 1: Numbers of Males and Females Sentenced
(All Courts, All Offences, All Ages)
from: Criminal Justice Statistics, Quarterly Update to March 2014, Ministry of Justice Statistics Bulletin

year males females ratio
2004 1,276,900 280,000 4.6
2006 1,166,200 282,000 4.1
2009 1,031,800 300,800 3.4
2012 921,200 298,100 3.1
2014 811,400 282,700 2.9

Note that it is men alone who account for the substantial reduction in the number of people being sentenced (despite it being specifically the male prison population which is increasing). In round terms, men now commit about 3 times more crimes than women.

If men commit 3 times more crime than women, why are there 21 times more men in prison? (see Key crime and prison data)

The reasons are easily identified. They are,

  • A greater percentage of convicted men are sentenced to prison;
  • Men are given longer sentences on average than women;
  • Women are paroled earlier than men

These statements are true, and supported clearly by the data presented below. However, these facts alone do not constitute unfair discrimination. The fact that convicted men are more likely to be sent to prison than women, and for longer, might simply be because men’s crimes tend to be more serious. Evidence that men are, in fact, treated far more harshly as regards prison terms follows, considering the three bullet points above in turn.

Disparity in Being Sentenced to Prison

Ministry of Justice (MoJ) data for 2009 shows that, of all men convicted, 8.8% were sentenced to immediate custody, compared with 2.7% of convicted women. This is referred to as an “imprisonment disparity” of 8.8/2.7 = 3.4. In other words, in 2009 a convicted man was 3.4 times more likely to be sent to prison than a convicted woman. In years 2010 and 2011 this imprisonment disparity was 3.4 and 3.6 respectively. This disparity has been increasing. In 1999 it was 2.5 (see Figure 1 below).

This disparity still does not necessarily imply unfair discrimination when crime severity is taken into account. However, Figure 1 below shows the imprisonment disparity broken down by crime category (click to enlarge).

Figure 1: Imprisonment Disparities by Crime Category, 1999-2009 From supplementary Table 2i of these government crime data

Imprionment disparity by crime category to 2009

With the exception of drug offences, the disparity is substantially greater than 1 in all crime categories. This is the first indication that the disparity is due, at least in part, to a harsher treatment of men. Many people would be willing to believe that men’s violent crimes are more serious on average than women’s violent crimes (rightly or wrongly). But why should men’s fraud, forgery, theft, handling stolen goods and motoring offences be worse too? Indeed, the assumption that they are is itself prejudice.

Disparity in Sentence Length

In this section the data sources were this for 2009 and for 2010 & 2011 the MoJ Sentencing Datasets Sheets S4.46, S4.47, S4.12, S4.13.

The 2009 data has been used here because the 2009 report includes a gender break-down against a fine-graded set of around 120 offence categories. To ease analysis, categories where <10 people of either sex were imprisoned have been ignored. Also life sentences or indeterminate sentences are not used here (because the sentence length is unknown). Some similar categories of offence were pooled together to improve the statistics. The result was a reduced set of 42 offence categories (listed here).

Sentence length disparity is defined as the ratio of the average sentence length for a man convicted of a given category of offence to the average sentence length for a woman convicted of the same offence. Figure 2 shows these disparities for the 42 offence categories. (NB: The histogram excludes those few offences where disparities were within +/-20% of 1).

Figure 2 shows that, with 3 minor exceptions, men receive substantially longer prison sentences than women for almost all offence categories (in 39 out of 42 categories). The main evidence that disparities result, at least in part, from unfair harsher treatment of men lies in the persistence of the disparity across almost all offences.

Year 2009 is not exceptional. Data for 2010 and 2011, for a smaller set of 12 offence categories, confirms similar levels of disparity for 11 of the 12 categories.

Figure 2: Sentence Length Disparities (2009)

Disparity in sentence length

In 2009 the average sentence length disparity across all offences was 1.55. In years 2010, 2011 and 2013 the average sentence length disparity across all offences was 1.51, 1.45 and 1.82 respectively.

Disparity in Early Release from Prison

Prison discharge tables show that women are, on average, discharged from prison after serving 48% of their sentence. For men this is 53%, a disparity of 53/48 = 1.1. This difference is not explained by better behaviour of women in prison. Women commit larger numbers of disciplinary offences per capita in prison than men.

Overall Disparity based on All Offences

We can now understand why, in 2009, there were 20 times more men than women in prison. It is because of the combination of the four factors shown in Table 2.

Table 2: Why there were 20 times more men in prison (2009): data based on all offences

Male:Female Convictions 3.4
Imprisonment disparity 3.4
Sentence length disparity 1.6
Parole disparity 1.1
Product 20 times more men

That is: 3.4 x 3.4 x 1.6 x 1.1 = 20

How much of this is unfair discrimination? This is hard to decide with confidence. As an illustration, if we assume that the greater number of men convicted is fair (i.e., that men commit 3.4 times as many crimes), then the unfairness resides in the other three factors. The analysis above does support there being a substantial element of unfairness inherent in these factors. If the entirety of these three factors is due to unfairness then the overall unfairness factor is 3.4 x 1.6 x 1.1 = 6.

This would lead to the conclusion that 5 out of every 6 men in prison would not be there if they were treated like women. However, there are several problems with such a sweeping conclusion,

  • Whilst it is clear than unfairness exists within the imprisonment and sentence length disparities, it has not been established that the entirety of these disparities are due to unfairness;
  • The treatment based on all offences may be misleading because, as we shall see below, summary and indictable offences have very different statistics and the latter dominate as regards imprisonment;
  • There are exceptional offence categories which are grossly misrepresented by the broad-brush treatment used above, the most important being sexual offences. A statement on sexual offences is included below.

To offset these criticisms, recall that any gender disparity arising from, (i)the probability of arrest, (ii)the likelihood of prosecution, and, (iii)the probability of conviction, has not been addressed here, but is likely to exist (see, for example, Evidence of similar gender bias in the USA).

Analysis Using Indictable Offence Data Only

Summary offences are less serious than indictable offences. Summary offences are tried in magistrates courts. This restricts the maximum possible prison sentence. Magistrates cannot sentence to more than 12 months in prison, and a maximum of 6 months for a single offence. Confining attention to those cases sentenced to prison by a magistrate, the average sentence length is only 2 or 3 months. But you are far less likely to be sent to prison for a summary offence than for an indicable offence. Consequently, although summary offences are three times more common than indictable offences, the overwhelming majority of prisoners are there for indictable offences (perhaps 95% or so). Consequently, it is reasonable to repeat the above analysis using data for indictable offences only. This is desirable because Figure 1 shows that the imprisonment disparity is substantially smaller for indictable offences than for summary offences. Table 3 re-emphasises the point.

Table 3: Imprisonment Disparity: Indicable and Summary Offences Compared Sources: 2009 and 2010 & 2011 MoJ Sentencing Datasets, Sheets S4.46, S4.47, S4.12, S4.13

year Indictable Summary Both
2009 1.9 5.7 3.4
2010 1.9 5.3 3.4
2011 1.9 5.7 3.6

Moreover, whilst the number of summary offences committed by men has been falling steeply, the decline in the number of indictable offences has been less marked – see Figure 3.

Figure 3: Trend of Convictions for Indictable and Summary Offences MOJ Sentencing Data: Quarterly Supplementary Tables (Table Q4a, June 2-14)

Convictions for Indictable and Summary Offences

Moreover, whilst men commit roughly 3 times more crimes than women overall, confining attention to the more serious (indictable) crimes reveals that men commit roughly 6 times more crimes than women – see Figure 4.

Figure 4: Ratio of Number of Men and Women Convicted

Ratio o men to women convicted

The imprisonment disparity factor and the sentence length disparity factor for indictable offences alone are shown in Figure 5, below. This Figure also shows the combined (product) factor, including the 10% increase for parole disparity.

Figure 5: Disparity Factors for Indictable Offences Alone

Disparity factors for indictable offences alone

Table 4 puts these all the disparities together to provide an explanation for there now being ~21 times more men than women in prison, and why this has increased from ~15 in 2003. Figure 6 shows this in graphical form.

Table 4: Disparities based on indictable offences alone: how they increase from 2003 (black) to 2013 (red) and the explanation of the increasingly large ratio of men to women in prison

Male:Female Convictions 5.7 increased to 6.2
Imprisonment disparity 1.7 increased to 1.9
Sentence length disparity 1.4 increased to 1.7
Parole / HDC disparity 1.1 (assumed constant)
Product (resulting ratio of men:women in prison) 15 increased to 22

Figure 6: Why there are 21 times more men in prison, based on analysis of indictable offences alone, and why this has increased from 15 in 2003 to 21 in 2013

Overall disparity for indictable offences alone

Conclusions from Analysis of Indictable Offence Data

  1. It is clear that the increasing absolute number of men in prison is not due to an increasing number of convictions, which has change relatively little. It is due to harsher sentencing now than previously (in 2003). In view of the Characteristics of prisoners one may question why our society has found it necessary to treat men more harshly than previously. What has been gained?
  2. It is clear that the increasing ratio of male to female prisoners is due to the imprisonment disparity and the sentence length disparity both having increased since 2003.
  3. Whilst a quantification of the degree of gender discrimination suffers from the same uncertainties as the analysis based on all offence data, the increases in the gender disparities for indictable offences since 2003 provide additional evidence that there is gender bias and the gender bias is increasing.
  4. If it is assumed that the factor of 6 times more men convicted of indictable offences is a fair reflection of men’s greater criminality than women’s, but that the other disparities result from bias, then the degree of unfairness has increased from 2.6 in 2003 to 3.5 in 2013. (NB: 1.9 x 1.7 x 1.1 = 3.5).
  5. The fact that the discrimination against men in the criminal justice system is not only large but also increasing should be of serious concern to society at large, because such a state of affairs is not sustainable.
  6. The reduction in the overall unfairness factor from 6, based on all offences, to 3.5, based on indictable offences only, might be interpreted as addressing the tendency of men’s crimes to be more serious. If so, the unfairness factor of 3.5 is a better estimate – though this is not entirely clear.
  7. Recall that this factor refers to bias in sentencing only. Biases in arrest, prosecution and conviction are likely also to exist. Consequently the overall gender bias is likely to be greater than the derived factor of 3.5, perhaps substantially greater.

In The institutionalised nature of the bias we will discover that the origin of the bias against men in the criminal justice system lies in government policy. The bias is systemic.

The Implications of Sexual Offences

Sexual offences are exceptional because there are well over 100 men in prison for sexual offences for every one woman. At 31st Dec 2015 there were 11,975 sex offenders in prison. Note that whilst this is 14% of the prison population, sex offences account for only ~3% of men convicted annually – but they receive particularly long sentences. The rate of imprisonment of men for sex offences is currently increasing at a rate of 9% per year. There are eight men’s prisons in England which hold only sex offenders. So sex offences do not fit the general trend of the above analysis.

However, sex offences do not provide a counter-argument to the claim that men are treated more harshly in the criminal justice system. On the contrary, sex offences are the most extreme example. In this case the discrimination is more profound. Sexual offences against males are not regarded as being so serious as sexual offences against females – especially when the offender is female. And society can barely recognise the possibility of an adult woman sexually abusing an adult man. In sex offences we face a discrimination too ingrained to be approached by mere data analysis.

Related links (open in this tab)