Yesterday was not only the 5th anniversary of my sister’s death, it was also the day that the Dutch national statistics office (CBS) made public the number of suicides this year. In 2013 1,854 people committed suicide or 11.0 per 100,000 people. In 1970 the suicide rate was 8.0 per 100,000 people or 1,049 suicides. A rather steep increase, at least when looking at the absolute numbers (Figure 1B). In contrast, the absolute homicide numbers in the Netherlands have been decreasing (Figure 1A), at least between 1997 and 2011. Between 1969 and 2011 the homicide has gradually increased (Figure 1B, green).
Figure 1. A) Absolute homicide numbers by sex in the Netherlands as reported by the CBS for the period 1996-2011. Blue=males; Red=females; Black=respective trend lines. B) Absolute homicide and suicide numbers as reported by the CBS for the period 1969-2013. Blue=male suicide; Red=female suicide; Green=total homicide; Black=respective trend lines.
What is surprising is the increase in suicides by males, but not females (Figure 1B, black trend lines). Their suicide numbers are remarkably stable. Sure, males commit more suicide, a fact common among all nations in the world. As I mentioned previously, males tend to commit suicide about twice as often compared to females. Yet, the ratio male::female is on the incline (Figure 2, red) for suicides and there is no indication to argue that the increasing disparity between male vs female suicide is leveling of. Surprisingly, and in contrast to the general assumption, the homicide sex ratio is slowly declining (Figure 2, green).
Figure 2. Sex ratio for suicide (red) and homicide (green). In black the trend lines are shown for both homicide and suicide.
Another factor potentially influencing suicide rate could be age. In contrast to homicide, where the age is roughly constant (fluctuation most likely due to relatively small number), a shift in age is observed for suicide. Now more and more middle aged people commit suicide, whereas in the past younger people committed relatively more suicide. This observation is consistent with data from the US. In this study by Bloomberg found that, despite less men dying from cancer and heart disease, more men between the age of 45-54 die from suicide and drug overdose.
Figure 3 Comparing age groups for suicide vs homicide between years. Each line represents a year between 1996 and 2011. The numbers represent relative size of each age group (grouped by 10 years) Homicide age groups remain fairly stable, whereas a suicide is becoming more common in middle aged people.
Also maritial status of the person committing suicide (Figure 3A1), method of suicide (Figure 3A2), and why (Figure 3A3). The categories for why people commit suicide is a bit doubtful, as I have pointed out various cases where the police simply assume to know why and how people do what they do without doing a forensic psychiatric autopsy. Yet, it is striking that in about 40% of all suicide it is not known why it was committed. This could be a problem with data collection at the local police departments, but it could also be that the police just didn’t bother to start an investigation because a suicide is not a criminal offense and thus does not require a police investigation. In about 25-30% of all suicide is a goodbye-note found. In many cases a suicide case involves a person with known psychiatric illnesses. I can imagine (but not back up with numbers) that known medical history and suicide notes combined account for the data presented in Figure 3A3.
Figure 4 A1-3) General statistics for suicides for the period 1996-2011 in the Netherlands. B1-3) general statistics for homicide for the period 1996-2011 in the Netherlands. A1) Marital status of suicide victim. A2) Method of committing suicide. A3) Perceived reason why someone committed suicide. B1) Where a homicide happened. B2) How a homicide was committed. B3) Why a homicide happened.
Similarly the methods/weapons used to commit homicide (Figure 3B2) is fairly consistent, in particular the use of a knife. The use of a firearm to kill was slightly out of fashion the second half of the first decade of the 21st century. It is surprising to see that more and more homicides are committed on the premise where the victim lived (Figure 3B1). All other locations for homicide is remarkably predictable. The reason for homicide is largely unknown (grey line in Figure 3B3). I can only speculate why this is. I think that the reason for a homicide is of little concern to the police itself, but more so to the public attorneys (Officier van Justitie). As the police departments provide the CBS with data and not the public attorneys office (Openbaar Ministerie). I would hope that if a homicide is successfully prosecuted by the public attorney, the reason (or at least a strong suspicion) for the homicide is established. If for 50%+ homicides the reason unknown remains, I cannot see how policymakers would be able to adjust police practices to combat homicides. Obviously, a certain number of homicides will remain, such as crime-passionel.
What also needs to be taken into account is the quality of the forensic doctor (schouwarts), which is heavily questioned, as well as how many forensic autopsies are performed. In the latter case we know that less and less are performed by the Dutch Forensic Institute (NFI) (340 in 2012), a subsidiary of the Department of Justice, the same department is in charge of the national police and the public attorney’s office (College van Procureurs Generaal). It would therefore not be surprising to learn that less and less homicides happen in the Netherlands. One such explanation would be that homicides that happened are mislabeled as suicides. This explanation is bit suspect, as it puts very little trust in the capabilities of the authorities.
Unnatural deaths on the rise
The death rate in the Netherlands is very predictable. Between 1969 and 2013 (blue line in Figure 5) around 0.83% of the population dies annually. It would therefore be natural to assume that the number of suicides per 100,000 inhabitants is an accurate representation and to a certain level this is true. This approach does assume that other factors that could influence the number of suicides remain the same. You have to take into consideration how suicides are determined. When a person is found dead which sparks a police investigation, who require a forensic doctor (schouwarts) to arrive on the scene to help close a case of suicide. In most cases the public attorney will show little to no interest in considering opening a criminal case if the police already suspects a suicide. Suicide is not a criminal act, so it is not their business. Any statistical information will come directly from the police departments, as we have seen above. This also means that specific data will be missing.
One type of data should be readily available and that is how many people die an unnatural death. With an unnatural death I mean any suicide, homicide, and accident/misfortune. Below you can see the red line representing the relative number of unnatural deaths, with in black the trend line (Figure 5). An obvious increase in unnatural death can be observed in the Netherlands between 1996 and 2011. It is even more surprising when taking into account that crime has been on a continuous decline for many years now.
Figure 5. How many people died in the Netherlands between 1969 and 2013 (blue) and how many of those who died died of unnatural causes (homicide, suicide, or accident) between 1996-2011 (for which data is available) in red. In black the trend line is shown for both the blue and red lines.
Maybe the increase in suicide can explain the increase in unnatural deaths? The suicide rate in the Netherlands went from 8.0 per 100,000 in 1970 to 11.0 per 100,000 in 2013. Yet, the relative suicide rate within the unnatural deaths remains very stable (~30%; Figure 6). The increase of unnatural death cannot be explained by the increase in suicide. If anything, the increase in suicide resembles the increase in unnatural deaths.
We do see two categories show a relative increase within the unnatural deaths (accidents at home and unknown). The unknown category is very troublesome. First, the unknown group could be indicative of two things: a) police departments are not reporting certain type of information any longer, as might be the case for homicides (Figure 3B1-3) or b) the cause of unnatural death is indeed unknown. If the latter is the true it is important to learn why the police is unable to determine the cause of ~10% unnatural deaths in the Netherlands.
Figure 6. How people died of unnatural causes between 1996 and 2011 in the Netherlands.
On a positive note, less people die due to traffic accidents. In order to reduce the number of traffic deaths the public attorneys office (Openbaar Ministerie) assigned a specialized public attorney Koos Spee. Whether or not Spee was responsible for the drastic reduction, the reduction also came as car became safer due to better tires, electronic driver aids, improved chassis, and more airbags. Either way, the reduction in traffic deaths is welcomed by many.
Deadly accidents in the private sphere: nursing homes are getting deadlier in the Netherlands.
Accidents in the private sphere (home in Figure 6) exclude all deaths from occupational, traffic, or sports deaths. For instance people too often underestimate how dangerous their homes are. Usually parents do try to reduce the risk of kids falling or drowning or being electrocuted by accident. Yet, for anyone older than 5 years of age the dangers are waved away. In 1996 about 37% (1999/5309) of all unnatural deaths were a result for an accident at home, whereas this number increased to over 48% in 2011 (2821/5844). It is a serious problem and the most common cause of unnatural deaths. A closer look (Figure 7) reveals that especially elderly die a unnatural death due to accidents (Figure 7A). In ~80% of the cases people die as a result of a fall (Figure 7B).
Figure 7. What is known about unnatural causes of death in the Netherlands that happen in the private sphere between 1996-2011. A) Age distribution per year. B) Cause of death. C) If known, what were people doing. D) Where were these people.
In 1996 very few cases were reported to have died from an accident in a nursing home or hospital, but in 2011 over 20% of the accidental deaths in the private sphere happened in nursing homes and hospitals. As about 80% of these accidental deaths result from a fall it can be safely assumed that deaths during a medical procedure, as you might expect to occur in a hospital, are not part of this statistic. This of course ignores the problem that Dutch hospitals are very lacks in keeping statistics about deaths during a medical procedures. Therefore nursing homes become the site of interest to account for the steep increase in accidental deaths in the private sphere. In the past, problems in the quality of care in nursing homes have become apparent. A shortage of people, a shortage of qualified nurses, and a byzantine bureaucratic oversight with all its excess. This makes the quality of care a structural problem that in part manifests itself in elderly falling and dying.
All taken together a murky picture emerges. The absolute numbers of suicide might on the rise, its contribution to all unnatural deaths in the Netherlands remain very predictable: about 30%. I can therefore only conclude that suicide is not on the increase in the Netherlands, but unnatural deaths are on the increase. The most prominent contributor appears to be nursing homes, maybe as a result from ever decreasing nursing quality.
All sources are embedded in hyperlinks and figures were made in excel.