Coffeeshops Vanish From The Streets – Impact On Surrounding Houses
Author: Jeroen Sanderink
Studentnumber:10211411
MSc Business Economics, Real Estate Finance track
University of Amsterdam, Amsterdam Business School
Master Thesis
Name of the first evaluator: Martijn Dröes
Name of the second evaluator: Marc Francke
22 June 2016
-Abstract- This thesis aims to sketch the external effect that coffeeshops may have on house prices in the near
vicinity. The first part gives an overview- and the history- of the coffeeshops in the Netherlands. It discusses the external effects that coffeeshops (could) have on house prices. In the second part, the method procedure is given, that is used to determine -the average effect of coffeeshops on house prices, and -the heterogeneity between coffeeshops. Both, coffeeshops that opened and coffeeshops that are closed are examined. The analysis conducted in this thesis shows both, positive and negative effects of coffeeshops on house prices. Overall, the average price that homeowners are willing to pay to live within 1250 meters of a coffeeshop is a premium of 1.84 percent of the actual transaction price. This positive effect is somewhat counter to expectations, but there are several possibilities that can declare the positive effect. The most plausible reason, is that a coffeeshop creates more vivacity in the neighborhood, which in turn outweigh the nuisance that the coffeeshop produces. A second reason could be that coffeeshops are perceived as amenities, wherefore people are willing to pay a premium to live near it and benefit from the advantages it creates. Note that this reason is somewhat related to the previous one, whereby the vivacity would be the benefit of the amenity. Next, the positive effect could be originated through lower crime rates and thus more safety in the neighborhood due to coffeeshops. Finally, there may be some unobserved effects, like differences in housing quality or local amenities, that might have influenced the results. In any case, further research is needed to draw such conclusions. The estimation is somewhat underestimated nowadays, as the heterogeneity analysis puts forward, since coffeeshops until 1999 had an overall negative price effect of (up to) 10.44 percent on surrounding properties. Although coffeeshops seem to have nowadays on average a positive effect, there are still several coffeeshops that decreased the house prices in the neighborhood. Mainly bigger sized coffeeshops, or coffeeshops in smaller municipalities, have resulted in lower house prices. One possible reason for this, is that the (perceived) nuisance factor is more present for these houses. At the end of this thesis, propositions are given for further research.
2
Statement of Originality
This document is written by Student Jeroen Sanderink (student number: 10211411) who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Index
Statement of Originality ... 2 Index... 3 1. Introduction ... 4 2. Literature Review ... 8 2.1 History of Coffeeshops ... 82.2 Evaluation Reports and Anecdotal Evidence ... 11
2.2.1 Coffeeshops in General ... 12
2.2.2 First Entrant ... 14
2.2.3 Zero Policy ... 15
2.3 Similar Externality Studies ... 17
2.4 Effects of a Coffeeshop on a Neighborhood ... 20
2.4.1 Drug Related Nuisance ... 20
2.4.2 Drug Related Crime ... 21
3. Method and Data ... 25
3.1 Data ... 25
3.1.1 Descriptive Statistics ... 26
3.1.1 Variables ... 33
3.2 Method ... 35
3.2.1 External Effect of Coffeeshops ... 35
3.2.2 Distance ... 36
3.2.3 Anticipation- and Adaptation- Effects ... 37
3.2.4 Concentration and Size ... 37
4. Results and Discussion ... 39
4.1 Results (openings) ... 39
4.1.1 Average Treatment Tests ... 39
4.1.2 Heterogeneity Between Coffeeshops and Over Time ... 45
4.1.3 Validity Checks ... 49 4.2 Results (closings) ... 50 4.3 Discussion ... 52 5. Conclusive Remarks ... 54 References ... 56 Appendix ... 59
4
1. Introduction
Since 1919, Dutch drug policy agreements have been laid down in the Opium Law (Toshkov & Wieldraaijer, 2013). It then prohibited the production; selling; and consumption of cannabis until 1976 (Toshkov & Wieldraaijer, 2013). Because the use of cannabis rose drastically in the second half of the 1960s, and thereby the number of prosecutions, a debate started if the current policy was the appropriate manner to deal with the rising drug problem (Toshkov & Wieldraaijer, 2013). The result of this discussion was that the use of cannabis would be decriminalized (EMCDDA, 2008; Toshkov & Wieldraaijer, 2013). The revised Opium Law further allowed1 for the sale of cannabis by small retailers, the so called coffeeshops, in order to eliminate the black market (Toshkov & Wieldraaijer, 2013). The rules that a coffeeshop must adhered to, were stipulated in national guidelines in 1979, the so-called AHOJ2-criteria (EMCDDA, 2008). In the early stages of the revised law, the market share
of the Dutch coffeeshops, relative to street dealers, increased drastically (EMCDDA, 2008). Additional amendments were made in 1996, and a fifth criterion, and other measures were added to the national guidelines (until then so called AHOJ-criteria) (Toshkov & Wieldraaijer, 2013; Bieleman et al., 2015). The fifth criterion3 forbids the coffeeshops to sell more than five grams of cannabis per transaction, and stipulate a maximum permissible stock of five hundred grams (Staatscourant, 1996). Many coffeeshops closed, due to the more stringent control (Bieleman et al., 2015).
To cope with recent problems, the government appended other criteria to the then called
AHOJG-criteria. First, in 2013, the inhabitant-criterion (Ingezetene-criterium, I) was added, which
states that only inhabitants of the Netherlands may sell and buy cannabis in coffeeshops (Bieleman et al, 2015). This was done to reduce the drug tourism and nuisance problem, and comply to the desires of the neighboring countries (Benschop et al., 2015; Bieleman et al., 2015; Blom, 2006; Staatscourant, 1996). Second, since 2014, a distance-criterion (Afstands-criterium) is in effect, which states that the minimal distance between a coffeeshop and secondary schools or vocational schools for students younger than 18 years old has to be at least 350 meters (Bieleman et al., 2015). First, the plans were to add this criterion in the nationwide tolerance policy, but it was later decided that it demands a more custom approach, the municipalities are therefore free to implement the
distance-criterion when the situation asks for it (Bieleman et al., 2015).
The use of cannabis is not evenly dispersed across the Netherlands (EMCDDA, 2008). The prevalence rate is greater in urban areas relative to rural areas (EMCDDA, 2008). This can be partly
1
The sale of cannabis remains an offence according to the law, but under certain conditions (read: AHOJG-criteria) would not be prosecuted (EMCDDA, 2008).
2
The AHOJ-criteria stands for: no Advertising (Affichering), no Hard drugs, no nuisance (Overlast), no selling to Juveniles (Jeugdigen).
3 The fifth criterion was appended to the AHOJ, since then called AHOJG. The G stands for: no selling of Great
5 explained because a coffeeshop needs a certain threshold of customers, in order to exist (Benschop et al., 2010). This means that it is more likely that a coffeeshop is situated in an urban area, which has a greater population density, than a rural area (Benschop et al., 2010). Since the introduction in 1976, the number of coffeeshops increased steadily until the mid 80s. From then, the licit cannabis outlets expanded heavily, and the amount of coffeeshops in the Netherlands was estimated at 1460 in 1995 (Bieleman et al., 1996). Since the tightening of the national guidelines (AHOJGI criteria) in 1996 and the enforcement thereof by municipalities, resulted in a strong decline of the number of coffeeshops. Throughout the years the number of coffeeshops decreased further to 591 coffeeshops, measured in 2014 (Bieleman et al., 2015). In short, the strong decline from 1996 onwards, is due to the tightening of the national tolerance policy and the enforcement thereof (Bieleman et al., 2015). Most regulations are introduced, since it is not lawful that governments facilitates crime, and they try to keep their integrity by introducing appropriate measures (Staatscourant, 1996; Bieleman et al., 2015). This in turn, provides possibilities for new entrepreneurs to enter the market and open a coffeeshop.
Given the nuisance that a coffeeshop could cause, like parking problems and noisy, loitering customers around a coffeeshop (Beelen et al., 2009), it is expected that a coffeeshop has a negative impact on the neighborhood, and house prices in particular. Meanwhile, it is somewhat ambiguous what the effect of a coffeeshop has on the safety in a neighborhood. At the one hand, it could decrease crime, and in particular illegal criminal organization. But it could also be a safe haven for the organized crime to launder their money or obtain new customers for their cannabis marketing (Bieleman & Snippe, 2006). Note that it became more difficult for the organized crime to infiltrate coffeeshops since the BIBOB act. Therefore it is more likely that a coffeeshop nowadays creates more safety in the neighborhood. At last, a coffeeshop could be seen as an amenity that creates more vivacity in the neighborhood (Beke et al., 2012). This in turn has positive impacts on neighborhoods and house prices in particular. Anecdotal evidence of local residents, reveals ambiguous attitudes towards coffeeshops (Benschop et al., 2013; Benschop et al., 2014; Bieleman et al., 2015). This study is the first that examines the external effect of coffeeshops on nearby houses. The research question that it endeavor to answer is: what is the effect on house prices when the
coffeeshop on the corner closes its doors? To answer this question, two analyses are conducted. The
first analysis examines the effect of the arrival of coffeeshops in various cities of the Netherlands over the period from 1980 to 2016. The second analysis examines the closure of coffeeshops over the period from 2000 until 2016. This analysis further tries to distinct the external effects, that are due to various reasons for closure. Both analyses make use of a detailed dataset of transacted houses in the near vicinity of the coffeeshops. With the use of a difference-in-differences method, it
6 is feasible to identify the effect that a coffeeshop has on the neighborhood in terms of house prices in that neighborhood (Dröes & Koster, 2014).
This thesis conduct a research in the field of externalities and house prices. There is done much research to the effect of externalities on house prices, but there is no study that the author is aware of, that examine the effect of coffeeshops on house prices. Therefore this thesis reviewed literature, that did research to externalities, that are somewhat comparable with coffeeshops. For example, Colwell et al. (2000) found that group homes have a negative effect on house prices, that are in sight of- or within 60 meters, from the group home. Quang Do et al. (1994) found a similar negative effect for houses that are situated within a 260 meter boundary from a church. Des Rosiers et al. (2001) examined the proximity effects of school on house prices, and concluded that easy access to school outweighs the negative proximity effects. This means that up to a certain distance, the effects of a school are positive for surrounding house prices (Des Rosiers et al., 2001). They found that houses experienced the biggest positive effect from a 400 meters distance (2001).
A reason why this has never been investigated before, is because coffeeshops are only present in the Netherlands. In other countries it is more difficult, and not legal, to buy cannabis for recreational use, except for Uruguay. Uruguay has passed a bill that fully legalize the production; selling; and consumption of cannabis, and is therefore more liberal towards cannabis than the Netherlands (Volkskrant, 2013). Also the Czech Republic is since 2010 more liberal towards drug use than the Netherlands, but selling drugs is still illegal there (Wall Street Journal, 2009; Benschop et al., 2010). This study improves the understanding of the effect of coffeeshops on the neighborhood, and house prices in particular. It reveals useful information for governments, municipalities, and other involved institutions to shape their policies towards drugs in the best interest of the majority. Also, this study states more about externalities in general. It examines the willingness of homeowners to pay, to live near a coffeeshop. In addition, home-buyers and –sellers can anticipate more efficiently when there is a coffeeshop located near the house.
According to the opening analysis conducted in this thesis, the general effect of coffeeshops on house prices is 1.84 percent. This effect is statistically significant at the 1 percent significance level. That the effect of coffeeshops is positive, is somewhat counter to expectations. However, the premium homeowners are willing to pay to live near a coffeeshop, could be explained by various reasons. First, the most plausible explanation is that coffeeshops contributes to the vivacity in the neighborhood (Beke et al., 2012). This means that a coffeeshop is perceived as amenity instead of an externality. A second explanation, could be that coffeeshops dampen crime levels in the near vicinity as mentioned by Beke et al. (2012); Benschop et al. (2013) and Bieleman et al. (2016). Lower crime
7 levels would in turn result in safer neighborhoods, but further research is needed to draw such conclusions. Finally, there may be some unobserved effects that might have influenced the results.
By examining the effect over time, it becomes clear that coffeeshops until 1999 had overall negative price effects for houses. But nowadays, house prices responds overall positive on nearby coffeeshops. This would suggest that the amenity factor of a coffeeshop is more positive than the nuisance factor is negative. This is in line with Beelen et al. (2009), who stated that nuisance due to coffeeshops sharply declined due to a more stringent coffeeshop policy. There are still coffeeshops that have depressing impacts on neighborhoods, and thus house prices. Mostly larger coffeeshops or coffeeshops situated in smaller municipalities have negative impacts on house prices. A logical explanation for this would be that nuisance levels originating from larger coffeeshops, reach certain thresholds, so that the nuisance factor outweigh the amenity factor. Next, in case of coffeeshops in smaller municipalities, the perceived nuisance levels are presumably relatively higher, and enough to outweigh the amenity factor. At last, the effect between established- and closed- coffeeshops differ from each other, presumably due to a selection bias within the data of the closed coffeeshops. Most of these coffeeshops closed due to a clearly negative reason, or were part of bigger renewal projects to positively transform neighborhoods.
Based on the results of this thesis, a cautious conclusion can be drawn that the more stringent policies and projects adopted by government and municipalities, reduced a lot of the negative effects of coffeeshops. This would also indicate that a coffeeshop does not have to facilitate inconveniences, provided that, it will be actively managed by municipalities, police, and the coffeeshop owner himself. Next to that, when the sale of soft drugs would be illegal, it would probably not lower prevalence rates, and the illicit drug market would take over (Miron & Zwiebel, 1995; Beke et al., 2012). Participants of this illegal drug market are more likely to use violence to solve their disputes, since they cannot use the judicial system (Miron & Zwiebel, 1995). This in turn would result, that residents feel less safe in their neighborhood (Beke et al., 2012; Benschop et al., 2013). In a broad sense, coffeeshops are an useful tool to regulate the soft drugs market, and exclude to some extent the illicit drug market. When a coffeeshops produce inconveniences, then there are most likely ways to reduce these.
This thesis proceeds as follows. The next part contains a brief literary composition about: the history and developments of coffeeshops; evaluation reports and anecdotal evidence; externality studies; and it concludes with a discussion to the effects a coffeeshop has on a neighborhood. In the third part of this thesis the research models are developed and a description of the data is given. The results that are obtained are presented and discussed in the fourth part. And in the last part some concluding remarks are given, as well as recommendations for further empirical research.
8
2. Literature Review
This chapter starts with a short history about coffeeshops. To get an idea about the effect of coffeeshops in the neighborhood, evaluation reports and anecdotal evidence are discussed in the subsequent section. Thereafter, relevant existing literature studies, that examined the effect of similar externalities, are reviewed. These studies are consulted for their methods that they used, the problems they encounter, and how to interpret the obtained results. In the last paragraph, the effects that a coffeeshop could have on a neighborhood are summarized and a hypothesis, with its likely outcome, is given.
2.1 History of Coffeeshops
Since 1919, Dutch drug policy agreements have been laid down in the Opium Law (Toshkov & Wieldraaijer, 2013). It then prohibited the production; selling; and consumption of cannabis until 1976 (Toshkov & Wieldraaijer, 2013). Because the use of cannabis rose drastically in the second half of the 1960s, and thereby the number of prosecutions, a debate started if the current policy was the appropriate manner to deal with the rising drug problem (Toshkov & Wieldraaijer, 2013). The result of this discussion was that the use of cannabis would be decriminalized (EMCDDA, 2008; Toshkov & Wieldraaijer, 2013). They also made a clear distinction between soft- and hard- drugs (Toshkov & Wieldraaijer, 2013; Reinarman, 2009; Staatscourant, 1996). This is done since soft drugs has a lower risk to the public health than hard drugs (Staatscourant, 1996). The Staatscourant further argues that the reason behind this separation is, that cannabis users should be kept out of harder and more criminal environments, which is accompanied with hard drugs (1996). In addition, the enforcement of hard drugs increased (Staatscourant, 1996).
The revised Opium Law further allowed for the sale of cannabis by small retailers, the so called coffeeshops, in order to eliminate the illicit market (Toshkov & Wieldraaijer, 2013). The rules that a coffeeshop must adhered to, were stipulated in national guidelines in 1979, the so-called
AHOJ-criteria (EMCDDA, 2008). In the early stages of the revised law, the market share of the Dutch
coffeeshops, relative to street dealers, increased drastically (EMCDDA,2008). Note that the production of cannabis remains illegal and prohibited4 (EMCDDA, 2008).
Additional amendments were made in 1996, and another criterion was added to the national guidelines (until then so called AHOJ-criteria) (Toshkov & Wieldraaijer, 2013; Bieleman et al., 2015). The additional criterion, (G), forbids the coffeeshops to sell more than five grams of cannabis per transaction, and stipulate a maximum permissible stock of five hundred grams (Staatscourant, 1996). Other amendments incorporated in the revised national guidelines of 1996 that applies to
4 Legalizing the cultivation of cannabis seems impossible, due to international treaties (Toshkov & Wieldraaijer,
9 coffeeshop were: prohibition on the admission of juveniles; no alcohol nor slot machines in coffeeshops; and no selling of soft drugs in catering establishments (Beelen et al., 2009; Staatscourant, 2016). Since then, municipalities can also - in consultation with the Public Prosecution (Openbaar Ministerie) and the police, so called the trialogue (het driehoeksoverleg) - adapt the national guidelines to deal with local priorities and needs (Beelen et al., 2009; Benschop et al., 2010; Bieleman et al., 2015). In addition, they are allowed to regulate coffeeshops via a licensing system (Benschop et al., 2010; EMCDDA, 2008; Monshouwer et al., 2011). This means that they have the opportunity to not allow coffeeshops in their communities (zero policy), or not allow new coffeeshops to enter their municipality (extinction policy), or limit the number of coffeeshops (Benschop et al., 2010; EMCDDA, 2008; Monshouwer et al., 2008). All these amendments were motivated by the developments in the prevalence rate of drug users, and due to the issues with regard to criminality and nuisance (Bieleman et al., 2015; Staatscourant, 1996). But also due to the consequences of the Dutch drug policy for neighboring countries (for example the drug tourism problem) and to keep the acceptation of the Dutch policy by foreign countries (Benschop et al., 2010; Blom, 2006; Staatscourant, 1996).
The act Damocles, that is introduced in 1999, empowers the mayor to close down coffeeshops, when they violate the rules that are laid down in the local tolerance policy (Bieleman et al., 2015; Burnt et al., 2012; Nederlands Staatsblad, 1999, 217). This means that municipalities do not have to take legal actions anymore, but can act themselves (Burnt et al., 2012). The Damocles act is also applicable to other public- and not public- premises, provided that they sold soft- or hard- drugs (Burnt et al., 2012; Nederlands Staatsblad, 1999, 217). Further, in order to maintain the integrity of the Dutch government, instruments for municipalities and other governmental institutions became operative through a newly introduced act, the so called BIBOB act, in 2003 (Bieleman et al., 2015; Staatsblad, 2003). Directors of governmental institutions obtain information through these instruments, about employers and employees of public premises (therefore also coffeeshops), in order to evaluate the risk that criminality is facilitated through the acts of them (Bieleman et al., 2015; Staatsblad, 2003). An example of facilitating crime could be that a municipality provide a license for a coffeeshop, when the owner will use the coffeeshop to launder money (Staatsblad, 2003). Therefore, the execution of the instruments of the BIBOB act let municipalities control the integrity of coffeeshop holders, and close malicious coffeeshops (Toshkov & Wieldraaijer, 2013). The
BIBOB- and Damocles- act contribute to the decentralization of control to municipalities, and creates
a better oversight.
To further cope with recent problems, the government appended other criteria to the then called
10 states that only inhabitants of the Netherlands may sell and buy cannabis in coffeeshops (Bieleman et al, 2015). This was done to reduce the drug tourism and nuisance problem, and comply to the desires of the neighboring countries (Benschop et al., 2015; Bieleman et al., 2015). The enforcement of the inhabitant-criterion is not in every area the same, but varies between municipalities (Benschop et al., 2015). For example, the municipality of Amsterdam and some eastern border municipalities still tolerate foreign cannabis users in their coffeeshops (Benschop et al., 2015; Toshkov & Wieldraaijer, 2013). Whereas the municipalities in the southern border areas are more strict, and do not tolerate the selling of cannabis to foreigners (Benschop et al., 2015). The reason for this differences is that the southern border areas experienced a lot of nuisance from the drug tourists (Beke et al., 2012; Benschop et al., 2015). Second, since 2014, a distance-criterion (Afstands-criterium) is in effect, which states that the minimal distance between a coffeeshop and secondary schools or vocational schools for students younger than 18 years old has to be at least 350 meters (Bieleman et al., 2015). First, the plans were to add this criterion in the nationwide tolerance policy, but it was later decided that it demands a more custom approach, therefore municipalities are free to implement the distance-criterion when the situation asks for it5 (Bieleman et al., 2015). Some municipalities had already adopted a distance criterion in their policy (for example Gemeente Arnhem). For other municipalities, the national distance criterion will result in many closures of coffeeshops (AT5, 2013). To summarize, most regulations are introduced since governments (and municipalities) wants to keep nuisance, criminality, and the prevalence rate of drug users (soft- and hard drugs) at a minimum (Bieleman et al., 2015; Staatscourant, 1996). Thereby, it is not lawful that governments facilitates crime, so they try to keep their integrity by introducing appropriate measures (Staatscourant, 1996; Bieleman et al., 2015). Some measures are also taken to solve concerns with neighboring countries (Benschop et al., 2015; Bieleman et al., 2015; Staatscourant, 1996).
The use of cannabis is not evenly dispersed across the Netherlands (EMCDDA, 2008). The prevalence rate is greater in urban areas relative to rural areas (Benschop et al., 2010; EMCDDA, 2008). This can be partly explained because a coffeeshop needs a certain threshold of customers, in order to exist (Benschop et al., 2010). This means that it is more likely that a coffeeshop is situated in an urban area, which has a greater population density, than a rural area (Benschop et al., 2010). Benschop et al. (2010) also stated that coffeeshops were initially found in back streets, but later relocated to busier-, more expensive-, locations6. Despite changes in policies by municipalities, the
5 The municipality first have to go in consultation with the Public Prosecution and the police (trialogue)
(Bieleman et al., 2015).
6 This was mentioned by an article from 1991, which means that this thesis deals mostly with coffeeshops that
11 number of municipalities that allow for coffeeshops remained constant at 103 (Bieleman et al., 2015). Since the introduction of coffeeshop in 1976, the number of coffeeshops increased steadily until the mid 80s. From then, the licit cannabis outlets expanded heavily, and the amount of coffeeshops in the Netherlands was estimated at 1460 in 1995 (Bieleman et al., 1996). Since the tightening of the national guidelines (AHOJGI criteria) in 1996 and the enforcement thereof by municipalities, resulted in a strong decline of the number of coffeeshops. The first real census in 1999, resulted in 846 coffeeshops (Bieleman et al., 2015). This heavy swing can be partly explained by the rule which is included in the national guidelines that forbids the sale of cannabis in combination with alcohol (Staatscourant, 1996). At first, hospitality companies saw potent in the sale of cannabis, and it was easy for them to adopt cannabis in their assortment. This could let to a significant increase in the number of cannabis outlets. After the introduction of the addition to the national guidelines, it was no longer permitted to sell cannabis and alcohol in the same local, whereby hospitality companies returned to their core business. Additional declines in coffeeshops can be explained by the decentralization of control to municipalities, which resulted in a better monitoring and controlling of coffeeshops (Toshkov & Wieldraaijer, 2013). Throughout the years the number of coffeeshops decreased further to 591 coffeeshops, measured in 2014 (Bieleman et al., 2015). The decrease of the number of coffeeshops is particularly found in municipalities, that have relative many coffeeshops (Bieleman & Snippe, 2006). In short, the strong declines from 1996 onwards, are due to the tightening of the national tolerance policy (and enforcement thereof), and the shift of control to municipalities (Bieleman et al., 2015). This in turn, provides possibilities for new entrepreneurs to enter the market and open a coffeeshop.
2.2 Evaluation Reports and Anecdotal Evidence
The effect of coffeeshops on house values, are probably predictable by the (prejudicial) attitude or the perspective of homeowners towards coffeeshops. Therefore evaluation reports are discussed in this chapter. These reports are provided with information from local experts like police officers, municipal officials, and prosecutors. But in some cases, these evaluation reports measures also the attitude of local residents regarding a relocated coffeeshop or regarding the consequences of a change in policy (named: ethnographic research). This is done by interviewing the- and taking surveys among these- local residents in certain points in time. According to Benschop et al. (2015), there is much similarity between the statistical figures and information from local experts. The first part of this chapter discusses the (perceived) nuisance and crime from coffeeshops in the Netherlands and its trends, followed by a sketch of the situation in 2014. Thereafter two special situations (first entrant and zero policy) are discussed.
12
2.2.1 Coffeeshops in General
Beelen et al. (2009) conducted a literature review about drug related nuisance, casu qou, the subjective nuisance perceived by residents of the Netherlands. They analyzed articles from 1995 until 2008, and found seventy suitable articles that they used in their review (Beelen et al., 2009).
In 1995, the residents of most municipalities in the Netherlands complained of multiple types of nuisance that originates from coffeeshops (Beelen et al., 2009). Note that the number of coffeeshops peaked in this period (Bieleman et al., 1996). The local residents faced hindered parked vehicles, and taken parking places (Beelen et al., 2009). Further, they complained about customers that were lingering in the area (Beelen et al., 2009). Sometimes these customers harassed bystanders, vandalize property, or left trash in the area (Beelen et al., 2009). Coffeeshops that were situated in residential areas received more complaints than coffeeshops in areas where there was less housing (Beelen et al., 2009). On the other hand, the areas with less housing had relative more catering, which made it sometimes difficult to distinguish the nuisance from catering establishments or coffeeshops, which in turn could result in less complaints (Beelen et al., 2009).
From 1996 the national guidelines towards coffeeshops became more stringent and the national budget to reduce drug related nuisance increased from 37 million guilder in 1995, to 60 million guilder in 19977 (Beelen et al., 2009). Multiple stakeholders, such as addiction treatment (aid and shelter), probation service, police, justice, and municipalities, helped and cooperated to decrease the nuisance (Beelen et al., 2009). The active stance of these stakeholders, in combination with the more stringent policy since 1996, resulted in a decrease in drug nuisance in the period between 1997 and 2008 (Beelen et al., 2009). Especially in the periods from 1997 until 1999 and from 2005 until 2008, did undergo coffeeshop related nuisance a sharp decrease (Beelen et al., 2009). But there was also an increase between 2000 and 2002 (Beelen et al., 2009). Hard drug nuisance sharply decreased from 1998 until 2000, mostly due to launched projects by municipalities (Beelen et al., 2009). These projects, that are partly financed by the government, entailed closings of deal premises, and a personalized approach for arrests or area denial of (repeat) offenders (Beelen et al., 2009). But also setting up usage premises, to keep drug addicts from the streets (Beelen et al., 2009). A result of these usage premises was that nuisance from annexation and pollution decreased, as well as the number of threats toward local residents made by drug addicts (Beelen et al., 2009).
Articles that are published shortly after the stricter policy measures (from 1999), mostly mentioned only nuisance from groups of youngsters and noise disturbance (Beelen et al., 2009). Some residents, especially southern border residents, still reported traffic nuisance (Beelen et al.,
7
The investments made to decrease the drugs related nuisance is difficult to trace after 1997 (Beelen et al., 2009). These investment figures are only from the government, besides several municipalities invested also in certain projects to reduce the drugs related nuisance (Beelen et al., 2009).
13 2009). Beelen et al. (2009) state that other forms of nuisance, not related to coffeeshops or drugs, were more present in most neighborhoods. Nevertheless, nearby residents of former coffeeshops mentioned that a great part of the nuisance disappeared, since the coffeeshop closed (Beelen et al., 2009).
The coffeeshop situation in 2014 as sketched, via a random sampling, by Benschop et al. (2015), shows a noticeably lower level of nuisance. The authors analyzed the soft drug tourism, soft drug related nuisance, and the illicit market (Benschop et al., 2015). This is done by looking at police figures and registered nuisance incidents about tourism, nuisance, and illicit drug trafficking (Benschop et al., 2015). In addition, they took interviews with local experts, like municipal- and police- officers, whom have insights in the local soft drug market (Benschop et al., 2015). Note that this analysis is performed after the inhabitant-criterion is in effect (Benschop et al., 2015).
According to Benschop et al. (2015), there are great differences between municipalities in terms of soft drug nuisance, and soft drug tourism. Most municipalities barely faced (nuisance from) drug tourism, except for some southern- and eastern- border municipalities8 (Benschop et al., 2015). Also interesting to mention is that most municipalities barely faced coffeeshop nuisance (Benschop et al., 2015). The municipalities that did face coffeeshop nuisance are mostly medium and large municipalities (Benschop et al., 2015). In general: the bigger the municipality (per capita) the greater the number of nuisance incidents per capita (Benschop et al., 2015). The reported nuisance consists mainly of traffic- and parking- problems in narrow streets, and lingering youngsters in busy neighborhoods (Benschop et al., 2015). In addition, municipalities reported that nuisance is not always coming from coffeeshop (customers), but also from drug –runners and –dealers who serve local- and foreign- customers (that are refused to enter the coffeeshop) (Benschop et al., 2015). Benschop et al. (2015) stated that in some municipalities the illicit soft drug market is moving to areas away from the coffeeshop as a result of taken policy measures, which resulted in a decrease in nuisance around the coffeeshops. Half of the municipalities had to face little or no drug -runners or – dealers (Benschop et al., 2015). The authors further mentioned that separate municipalities face a seasonal effect in terms of nuisance incidents (Benschop et al., 2015). The nuisance is the lowest in the winter months, with a peak in the summer months (Benschop et al., 2015). The model presented in this thesis does not control for the seasonal effects, but distinguishes the time of sale per month. Thereby, a small part of the seasonal effect in nuisance incidents, is partly controlled for.
In summary, the level of nuisance from coffeeshops and the illegal market varies –geographically, and –over time (Beelen et al., 2009; Benschop et al., 2015). Some municipalities experience great nuisance from coffeeshops, others have succeeded to keep the nuisance to a minimum (Beelen et
8 19 out of 31 municipalities did not face nuisance from soft drugs tourists, and 16 municipalities had not soft
14 al., 2009; Benschop et al., 2015). The southern border municipalities faced the greatest nuisance from coffeeshops and the illegal market (Beelen et al., 2009; Benschop et al., 2015). This is mainly due to the large influx of foreigners who want to buy (legal) cannabis (Beelen et al., 2008; Benschop et al., 2015). In general, the larger the municipality, the greater the nuisance (Benschop et al., 2015). Therefore, the size of the municipality is controlled by neighborhood dummies in the model, which is explained in greater detail in the Method and Data-section. Further, since nuisance levels had declined significantly from 1999 onwards (Beelen et al., 2009), the effect of coffeeshops on house prices is also examined over time.
2.2.2 First Entrant
The municipality of Lelystad had to face for some time an illicit drug market (Benschop et al., 2013). The drug -dealers and -runners were causing nuisance by telephoning and lingering in the neighborhood (Benschop et al., 2013). In the past, the police had already tried to combat these illicit street dealers, but without success, it just opened the way for new entrants (Beke et al., 2013). Therefore the municipality decided that it had a need for a coffeeshop (Benschop et al., 2013). So, in mid August 2011, the first coffeeshop in Lelystad opened its doors (Benschop et al., 2013). Benschop et al. (2013) surveyed the local residents, police officers and municipalities about the consequences of the advent, and discussed it in their article. They measured the overall consensus at three stages: before-; shortly after-; and a year after- the advent of the coffeeshop (Benschop et al., 2013).
The outcomes of Benschop et al. (2013) states that the coffeeshop does not attract beginning users. Also there is less nuisance of cannabis users that are smoking in public places like parks, in squares, or on the street, then before the advent of the coffeeshop (Benschop et al, 2013). Presumably, this can be partly explained by the proactive stance of the municipality, the police and the coffeeshop owner himself, and their solutions for known inconveniences (Benschop et al., 2013). For example, a porter of the coffeeshop ensures that customers do not linger in the neighborhood, or that they do not make excessive use of the available parking places (Benschop et al., 2013). In addition to that, more police officers were deployed (Benschop et al., 2013).
The residents in some neighborhoods signaled less openly soft drug use and soft- and hard drug deals after the advent of the coffeeshop (Benschop et al., 2013). But this effect only became visible at the last metering (Benschop et al., 2013). The attitude towards the coffeeshop became more positive, although, it remained divided (Benschop et al., 2013). The stance towards cannabis use in general became somewhat more positive then before (Benschop et al., 2013). In addition, the local residents felt more safe after the arrival of the coffeeshop (Benschop et al., 2013). Some property owners had complained about an expected depreciation of their house due to a coffeeshop in their neighborhood, but a year after the advent they were somewhat more positive (Benschop et al.,
15 2013). All forms of traffic, such as traffic congestions, and cumbersome bikers, were high before- and increased in- the beginning of the advent, but quickly decreased (Benschop et al., 2013). Except for parking problems, they remained the same (Benschop et al., 2013). Illegal drug dealing increased in some neighborhoods, but overall there was a declining trend (Benschop et al., 2013). Also, the illicit drug market significantly decreased in general (Benschop et al., 2013). The personalized approach for arrests of offenders, that is adopted by the police a year before the advent, probably contributes to these decreases (Benschop et al., 2013).
To summarize, people were somewhat averse towards the coffeeshop beforehand, but the expected nuisance that it would bring, held off (Benschop et al., 2013). Instead, the agitation in the neighborhood declined and the number of street dealers decreased (Benschop et al., 2013), probably, partly due to the active stance of the police (more police officers deployed), municipality, and the coffeeshop owner. To account for the transition, and the changing attitude towards coffeeshops associated with it, there is made use of adaptation effects in the empirical research. Presumably, the effect of coffeeshops on house prices is at the time of the advent greater in magnitude than a year-, or two years after.
2.2.3 Zero Policy
Bergen op Zoom and Roosendaal faced a somewhat different situation. The coffeeshops in these cities attracted many Belgium- and French drug tourists, that wanted good quality cannabis from a legal source9 (Beke et al., 2012). The local residents were hindered by the major influx of the drug tourism, to which the municipality saw no other option than to adopt a zero policy (Beke et al., 2012). This means that, since 16 September 2009, no coffeeshops are allowed anymore (Beke et al., 2012). This resulted in the closure of eight coffeeshops in total (Beke et al., 2012). The researchers studied the developments that were induced by the closures of the coffeeshops (Beke et al., 2012). They examined, inter alia, the drug tourism-, the illicit drug market-, and the drugs criminality (Beke et al., 2012). It combines police figures with surveys among local residents, police officers, and store owners (Beke et al., 2012).
Before the newly adopted policy, the municipalities had invested in, camera surveillance, stringent enforcement, and more intensive detection, to deal with illicit drug outlets, drug dealers, and cannabis cultivators (Beke et al., 2012). Although they booked some progress, the drug nuisance persisted (Beke et al., 2012). To comply with the drug problems, they adopted the zero policy. Some immediate effects after the closures were the sharp decline of drug tourists in both municipalities10 (Beke et al., 2012). Many tourist looked for alternatives, and now use one of the eight coffeeshops in
9 Weekly 25000 drug tourist, mainly by car (Beke et al., 2012). 10 The number of drugs tourist decline with 90% (Beke et al., 2012).
16 Breda11 (Beke et al., 2012). Local residents that live in the near vicinity of the coffeeshops in Breda, have not complained about increasing nuisance (Beke et al., 2012). However, according to the authors, the increase of the illicit drug market in Breda will highly likely change this in the near future (Beke et al., 2012). Further, the drug trafficking and drug production is partly relocated in Belgium, this because the officials in Belgium has less experience in tackling drug related activities, whereby criminals have a smaller chance to get caught12 (Beke et al., 2012).
The major influx of foreign drug customers, was also an income source for illicit cannabis cultivators13 and drug traffickers, who has connection with- or are part of- a criminal organization (Beke et al., 2012; Bieleman & Snippe, 2006). Drug runners and drug dealers were active along the routes and around the coffeeshops, in order to increase their market share (Beke et al., 2012). After the end of the tolerance policy, the illegal drug market experienced a peak, but decreased slightly overall14 (Beke et al., 2012). The simple street dealers lost customers due to the decrease in tourists, some of them focus now more on younger people, others quit (Beke et al., 2012).The authors also examined if there was a change in criminal activity (Beke et al., 2012). They only found an increase in the number of burglaries, but whether there is a causal link is unknown (Beke et al., 2012). Further, the more professional street dealers are still active, and their market became more important (Beke et al., 2012). The visibility of this illicit drug market also became more visible for some local residents (Beke et al., 2012).
Surveys among residents of Roosendaal, has shown that the nuisance from illicit drug dealers and drug runners significant increased (Beke et al., 2012). Residents in some neighborhood feel sometimes very unsafe (Beke et al., 2012). Some areas have a curfew, to nipping these problems (Beke et al., 2012). The situation in the center of Roosendaal is different, there the nuisance from drug dealers and runners decreased significantly (Beke et al., 2012). But also the vivacity in the streets declined, partly due to vacant retail properties (Beke et al., 2012). Further, it is interesting to mention that the drug problems continued to increase in bad neighborhoods, despite the fact that municipalities, police, and housing associations had taking measures there (Beke et al., 2012). People in these neighborhoods are feeling more unsafe, especially in the evenings15 (Beke et al., 2012).
11 The number of customers of coffeeshops in Breda increased with 30% (Beke et al., 2012). 12
This is also because the municipalities of Bergen op Zoom and Roosendaal manage a more stringent and active policy (Beke et al., 2012).
13 Note that the production of cannabis is still illegal and prohibited in the Netherlands (EMCDDA, 2008) 14 Note that it became more difficult to ascertain drug dealers, then before the zero policy (Beke et al., 2012). 15
Unfortunately, this thesis does not examine the effect of coffeeshops on house prices between neighborhoods with different quality levels, since the dataset does not specify the quality of the neighborhood where the house is located. But note that this could give interesting results.
17 Meanwhile, Bergen op Zoom experienced an increase of drug related nuisance more in the center, than in the outskirts (Beke et al., 2012). The development of drug related nuisance in the areas where the coffeeshops were located are dispersed (Beke et al., 2012). In some of these neighborhoods the nuisance has decreased significantly, and local residents feel more safe, even in the evenings (Beke et al., 2012). Residents in other areas has signaled more drug trafficking and experienced more drug related nuisance (Beke et al., 2012). However, there was one local resident that mentioned that there was already an increase in their neighborhood, before the policy changed (Beke et al., 2012).
To summarize, the decline of illicit drug trafficking in the neighborhoods are dispersed (Beke et al., 2012). The sharpest drop of drug trafficking is observed in the areas where the coffeeshops were located (Beke et al., 2012). On the contrary, some neighborhoods experienced an increase in drug dealers (Beke et al., 2012). The overall number of signaled illicit drug trafficking declined, and the perception of safety increased (Beke et al., 2012). The overall decline is partly due to the vanished drug tourists, and the actions of the police. Interesting to mention is that the authors think that the developments of safety, presumably, also could have been reached without the closure of the coffeeshops (Beke et al., 2012). Probably the inhabitant-criterion, that was nationwide adopted two years later, could be a possible solution.
Since the cases (first entrant, and a zero policy) in the three above mentioned cities (Lelystad, Bergen op Zoom, and Roosendaal) are unique, the effect of the policy changes on the house prices are therefore more elaborate studied in the empirical research. Further, since there are significant differences (in quality, and reaction) between neighborhoods, each neighborhood gets its own dummy in order to control for these differences. This is explained in greater detail in the Method and
Data-section.
2.3 Similar Externality Studies
This thesis conduct a research in the field of externalities and house prices. There is done much research to the effect of externalities on house prices (Colwell et al., 2000; Des Rosiers et al., 2001; Dröes & Koster, 2014; Quang Do et al.,1994), but in case of coffeeshops there is, up to this point, none that the author is aware of. The existing literature on externalities is reviewed in the first paragraph of this chapter. The relevance of these articles is based on their effect on the neighborhood in terms of noise pollution (Colwell et al., 2000; Des Rosiers et al., 2001; Quang Do et al., 1994), nuisance (Colwell et al., 2000; Des Rosiers et al., 2001; Quang Do et al., 1994), and safety (Colwell et al., 2000). Except for the article by Dröes & Koster (2014). This article is consulted for the
18 method that they used, since the method used in this paper is most suitable (and complete) to examine the effect of coffeeshops on house prices (2014).
The first article that is reviewed, did research to the proximity effect of schools on nearby house prices (Des Rosiers et al., 2001). A school could produce potentially similar negative effects as a coffeeshop, in terms of noise, increased traffic and parking nuisance, and in some cases vandalism (Des Rosiers et al., 2001). These negative effects only applies for close proximity houses, which means, within 50 meters of a school (Des Rosiers et al., 2001). The positive effect that a school could have for homeowners are reduced traveling costs (Des Rosiers et al., 2001). Their dataset contains 4300 transacted houses in Quebec (Canada), over the period from January 1990 through December 1991 (2001). The opted method in this thesis is a little different than the method that is used by Des Rosiers et al. (2001). The method used by Des Rosiers et al. (2001), is a hedonic pricing model, whereas this thesis uses a hedonic pricing model in combination with a difference in difference model. This is done, in order to check the difference before and after the advent (or closing) of a coffeeshop. The results of this study suggest that a distance of 400 meters from a school is most optimal, in terms of house prices (Des Rosiers et al., 2001). They further mention that the size of the school, measured in total pupils, is also related to house prices (Des Rosiers et al., 2001). Therefore, the size of the coffeeshop, measured in number of employees, and its effect on house prices will be examined in greater detail. This is further explained in the Method and Data-section. The strength of this research is that it test various functional forms and specifications for the size of a school and the distance between a school and a property (Des Rosiers et al., 2001). They conclude that the school as externality with respect to both school size and distance to properties, is best captured by the gamma transformation16 (Des Rosiers et al., 2001). The best suited functional form and specification is discussed in more detail in the Data and Method-section.
Next, Quang Do et al. (1994) examined the effect of churches on housing values. The effect of a church is also similar to that of a coffeeshop. It has for example the potential to create more noise in the neighborhood from arriving and departing visitors, and to increase traffic and parking problems (Quang Do et al., 1994). A church, as a coffeeshop, differs in terms of operating hours from a school, since it operates outside regular working hours (Quang Do et al., 1994). The study of Quang Do et al. (1994) used a hedonic pricing model in order to investigate the effect of churches on house prices. This is somewhat different as the method used in this paper, as mentioned above. Their study covered 469 transacted houses in the Chula Vista (United States), over the period January 1991
19 through September 1992 (Quang Do et al., 2014). Their results indicate that a church has a negative effect on the transaction price (1994). This means that a church can be seen as a negative externality. Based on a straight-line distance from the property to the church, they found a convex relationship between the price of a house and the distance (1994). This means that the negative effect diminished the further a house is located from a church (1994). This diminishing effect was greater for more expensive houses (1994). The negative effect of a church was affecting the house price up to approximately 260 meters (Quang Do et al., 2014). A disadvantage of the method used by Quang Do et al. (1994) is that it does not indicate at which distance there are significant measurable effects. It only states that the opted functional form is significant for all the transacted houses in the dataset. To circumvent this problem, multiple distance dummies are included in the model of this thesis, as further explained in the Data and Method-section.
Another study performed by Colwell et al. (2000), did research to the effect of group homes17 on house prices. They studied the establishment of seven group homes, which opened between 1987 and 1994. The total transacted houses in these seven neighborhoods amounted 641. They did not make a distinction on size nor on specialization (which types of peoples occupies the group homes), not to mention that none of the group home residents have a criminal record (Colwell et al., 2000). The mooted effects group homes should have are similar to that of coffeeshops, namely nuisance from unwanted persons, and a treat of safety in the neighborhood (Colwell et al., 2000). They used data from six years prior and six years following the announcement date. The authors made use of an altered hedonic pricing model, videlicet, they arranged the observations so that all the announcement dates occur at time 0, thereby they could perform a sort of event study (Colwell et al., 2000). They found a significant decline in property values for properties in sight of- or within 60 meters from- a group home (Colwell et al., 2000). Their results indicate further that the price of houses already declined before it is announced that the group home enters the neighborhood (Colwell et al., 2000). This suggest that group homes are placed in areas where house values are experience depressions (Colwell et al., 2000). Important to mention is that when the authors used prior methods suggested18 by other authors, that they did not find significant or unambiguous results (Colwell et al., 2000).
17
Group homes are private residents for young people that cannot live with their families or those with developmental, physical, and mental disabilities (Colwell et al., 2000). But also recovering alcoholics and drugs addicts (Colwell et al., 2000). These residents are mostly located in residential areas, to create more humane situations (Colwell et al., 2000).
18 The methods used were, a matched pair (no significant effect), and three hedonic pricing models (two times
20 The research that is most related to the research that will be conducted in this paper is written by Dröes & Koster (2014). These authors examined the effect of wind turbines on house prices in the Netherlands with the use of a difference-in-differences model (2014). They studied the advents of wind turbines and their effect on house prices over the period from 1985 until 2011. Their results indeed suggest that a wind turbine negatively affect the sales price of a house (2014). The negative price effect depends on the distance between the wind turbine and the house, with a significant effect, α=5%, up to 2 kilometers (Dröes & Koster, 2014). They further mention that homebuyers anticipate whether a wind turbine will be build beforehand (2014). This is observed in declining house prices, three years before the completion of the wind turbine (Dröes & Koster, 2014). The authors also examine whether there are any adjustment effects, by analyzing the house price some years after commissioning (2014). Adjustment effects could arise since the housing market slowly incorporate new information (Droës & Koster, 2014). The model described in the Method and Data-section, also incorporates anticipation and adjustment (adaptation) effects.
2.4 Effects of a Coffeeshop on a Neighborhood
The proximity effects of a coffeeshop on its neighborhood, and in particular the house values in the neighborhood, are discussed in this chapter. Two potential sources of proximity effects that originates from coffeeshops are, as briefly mentioned in previous paragraphs, nuisance and criminality. The first part of this chapter discusses drug related nuisance due to coffeeshop. The direct and indirect crime that originates from- or dissolves due to- coffeeshops, are discussed in the second part. It tries to give a conclusion to the net effect that coffeeshops have on crime rates. The last section discussed the expected impact that a coffeeshop has on house prices, and concludes with the main hypothesis.
2.4.1 Drug Related Nuisance
This section summarizes the different forms of nuisance that result from coffeeshops (direct) and the illicit market (indirect). Direct coffeeshop nuisance as perceived by local residents can be divided into three forms. The first form is traffic nuisance, such as hindered parked bikes and other vehicles, traffic congestions, or parking problems (Beelen et al., 2009; Beke et al., 2012; Benschop et al., 2013; Benschop et al., 2015). In particular southern border municipalities experience this type of nuisance, due to drug tourism (Beelen et al., 2009; Beke et al., 2013). The second form of nuisance are visitors that are lingering in the area around the coffeeshop, screaming, telephoning, harassing bystanders, and vandalize belongings of others (Beelen et al., 2009; Beke et al., 2012; Benschop et al., 2013; Benschop et al., 2015). The last form that is mentioned by municipalities are pollutions in the area (Beelen et al., 2009). For example, visitors are leaving trash, urinate or puke in gardens and porches of surrounding houses (Beelen et al., 2009). Beelen et al. (2009) state that after 1999, other forms of
21 nuisance, not related to coffeeshops nor drugs, were more present in most neighborhoods. Also, residents perceive drug tourists as enjoyment to a certain extent, whereby a coffeeshop could produce more vivacity in the neighborhood (Beke et al., 2012). The municipalities that did face coffeeshop nuisance are mostly medium and large municipalities (Benschop et al., 2015).
Since this research is interested in the external effect of coffeeshops, it tries to keep the discussion about the illicit drug market brief. But, because coffeeshops and the illicit drug market are intertwined, it is good to appoint it briefly (Beke et al., 2012; Benschop et al., 2013; Bieleman & Snippe, 2006; Bieleman et al., 2016). The interrelation that is meant here, is that coffeeshops are mainly supplied by criminal organization19 (Bielaman & Snippe, 2006), but also indirectly provide customers for (simple) illicit drug –dealers and -runners via the concentrated demand (Beke et al., 2012), or dissolve (part of) the illicit drug market as also was the case in Lelystad (Benschop et al., 2013; Bieleman, 201620; EMCDDA, 2008).Anyhow, Benschop et al. (2015) mentioned that local experts not always have a representation of the illicit drug market. They argue that this could indicate that the illicit drug in some municipalities do not produce nuisance that is worth mentioning (2015).
Nuisance from the illicit market originates mainly from drug -addicts and –dealers (Beelen et al., 2009; Beke et al., 2012; Benschop et al., 2013). Nuisance from the illicit market can also be divided into three forms, namely criminality, public disorder, and audiovisual nuisance (Beelen et al., 2009). Criminality include violence, property crime, and petty crime (Beelen et al., 2009). The next paragraph discusses drugs related crime in more detail. The second form, public disorder, mainly involves groups that annex areas (Beelen et al., 2009). For example, drug –addicts and -users that gather together on a street, park, or square (Beelen et al., 2009; Benschop et al., 2013). The last form is audiovisual nuisance, like irritating-, cumbersome-, and uncivilized- behavior by drug – addicts and -dealers (Beelen et al., 2009; Beke et al., 2012). Beelen et al. (2009) state that all these forms of nuisance, influence the degree of safety.
2.4.2 Drug Related Crime
This section summarizes the direct- and indirect- crime that is caused- or resolved- by coffeeshops. The first paragraph, reviewed studies that examined the effect of crime rates on house prices. These studies both found a negative correlation between crime rates and house prices (Ihlanfeldt & Mayock, 2010; Pope & Pope, 2012). The subsequent paragraph discusses the possible (indirect) effects of coffeeshops on crime (rates).
19 As explained in the next section. 20
Bieleman et al., (2016) examined the developments of a policy change that stated that every customer needs a membership. They found that multiple customers switched to the illicit market (2016). The membership criterion is never adopted nationwide, but is only tested in some municipalities (2016).
22 The first article reviewed, Ihlanfeldt & Mayock (2010), argues that the vast majority of the studies that examine the effect of crime on house prices, treated crime as an exogenous variable. While, it is well known that crime rates are higher (lower) in various neighborhoods. For example, expensive homes are more attractive to criminals, since the loot that can be stolen is higher for these homes, relative to cheaper homes (Ihlanfeldt & Mayock, 2010). Ihlanfeldt & Mayock (2010) further argues that reporting rates are different across neighborhoods, and known to be higher in more wealthier areas. Also the self protection is higher in more richer neighborhoods, which would cause the crime rates to be lower (Ihlanfeldt & Mayock, 2010). All the mechanisms described above, points out the endogeneity of crime rates on house prices (Ihlanfeldt & Mayock, 2010). To overcome a major part of this endogeniety bias described above, the authors took the first differences of the data, and regressed the change in the house price index on changes in the crime rate (Ihlanfeldt & Mayock, 2010). This method not only reduce the endogeneity problem, but also mitigate multicollinearity among the various types of crime (Ihlanfeldt & Mayock, 2010). By controlling for entity fixed effects, they lessened the omitted variable bias (Ihlanfeldt & Mayock, 2010). They further make use of changes in particular commercial land uses as an instrument for changes in particular types of crime (Ihlanfeldt & Mayock, 2010). The reasoning behind this according to the authors is that several land uses are excellent targets for particular types of crime, while other land uses favor other types of crime (Ihlanfeldt & Mayock, 2010). The authors investigated eight types of crimes: murder-, burglary-, robbery-, aggravated assault-, motor-, theft-, larceny-, and vandalism- in the neighborhood, over the period from 1999 to 2007 in Florida (2010). Their findings where that only an increase in robberies or aggravated assaults have an influence on the house prices in the neighborhood (Ihlanfeldt & Mayock, 2010). They further found that vandalism has high correlation with these two types of crime, but does not have an impact on the price of a house (Ihlanfeldt & Mayock, 2010). Note that the results could differ among areas, for example an area that is less urban (Ihlanfeldt & Mayock, 2010).
Another study that did investigate the impact of crime on house prices is conducted by Pope & Pope (2012). The nationwide decrease of the crime rates in the United States in the 1990s, was an ideal circumstance (read: partly natural experiment) to examine the impact of the crime drop on house prices (Pope & Pope, 2012). This because the drop in crime was huge and unexpected, and there was significant difference in crime decreases within and across urban areas (Pope & Pope, 2012). The authors further mentioned that the socioeconomic background did not changed dramatically during this period (Pope & Pope, 2012). This study did examine the effect of crime rates on house price over the period from 1990 to 2000 (Pope & Pope, 2012). They found that there were cities where the crime rate increased instead of following the national trend, and thus have the
23 potential to be endogenous (Pope & Pope, 2012). To instrument this type of potential endogeneity out, they used data of similar (nearest) areas for the areas that differed from the general trend (Pope & Pope, 2012). By using a hedonic pricing model, they found a large and statistically significant relation between crime rates and house prices (Pope & Pope, 2012). The authors do not specify the relationship for different types of crime, but they give an underlying reasoning for the general relation between crime rates and house price (2012). They argue that the immediate benefit of a reduction of crime is that people feel safer and properties are less likely to be damaged (2012). But also positive amenities, such as restaurants and stores, are more likely to enter when crime is permanently reduces (Pope & Pope, 2012). To analyze if the crime that (indirectly) originates from coffeeshops have an impact on house prices, other literature is reviewed in the next paragraph.
The chance of intrusion of -criminals, or - criminal organization in the soft drug market are, due to the high earnings, low risk of detection, and low sentences, high21 (Bieleman & Snippe, 2006). Also the opportunity to launder illegally acquired assets, attracts criminal organizations (Bieleman & Snippe, 2006). For example, they finance new entrants that need capital that the bank is not willing to provide (Bieleman & Snippe, 2006). Further, the production of cannabis is often in the hands of criminal organizations (Bieleman & Snippe, 2006). Next to small home growers, who may legally cultivate five cannabis plants, the coffeeshops are mainly supplied by large illicit growers that has connections with- or are part of a- criminal organization (Bieleman & Snippe, 2006). Hence, coffeeshop owners are somewhat forced to do business with drugs criminals (Bieleman & Snippe, 2009). Coffeeshops can also facilitate criminalization in neighborhoods, by making solely (and on a large scale) use of home growers (Bieleman & Snippe, 2009). This is in particularly true for weak social-economic neighborhoods (Bieleman & Snippe, 2009).
The illicit drug market is often accompanied by violence (and harder criminality). As Miron & Zwiebel (1995) argued, participants of the illicit drug market cannot use the judicial system, and are therefore more inclined to use violence to resolve their disputes22. The authors further state that prohibition increase the ease with which criminals could position themselves in the market (1995). The conclusion of Miron & Zwiebel (1995) is that violence and property crime is lower in a free market than in a prohibited market.
With regard to drug users, Pedersen & Skardhamar (2009) did not find any association between cannabis use and criminal involvement. They found no evidence that the use of cannabis, would
21 The sentences for robbery, weapons possession, or hard drugs trafficking, are higher than the sentences for
soft drugs trafficking (Bieleman & Snippe, 2006).
22 They studied the murder rate in the United States and prohibition laws of alcohol and drugs from 1910, and
24 increase the risk of subsequent non drug specific criminal charges, such as criminal gain or violence (Pedersen & Skardhamar, 2009). According to Miron & Zwiebel (1995), there is a positive relation between the prices of heroin and property crime. They argue that drug addicts steal, in order to get money to fulfill their needs (read: addiction) (1995). This is probably not the case for cannabis users, only for some hard drug users. Since the Dutch drug policy made a substantial separation between soft- and hard- drugs (Reinarman, 2009). The policy, and thus coffeeshops, also kept cannabis users out of the more criminal environment that is related to the hard drug scene (Pedersen & Skardhamar, 2009). Note that multiple articles mentioned a harder criminal milieu that involves with hard drugs (Benshop et al., 2015; Miron & Zwiebel, 1995; Staatscourant, 1996).
In short, I expect that the overall direct and indirect effects of a coffeeshops, results in lower house prices, mainly due to nuisance. These effects probably differ across municipalities. First because not every municipality is as active towards resolving nuisance problems. But also since not all municipalities face drug related nuisance (Benschop et al., 2015). The effect would probably differ over time, since drug nuisance was more present before 1999 according to Beelen et al. (2009). Next to that, I expect that coffeeshops situated in larger municipalities (based on capita), have a greater impacts on house prices, since nuisance levels are greater there (Benschop et al., 2015).
Overall crime would probably not be directly affected by coffeeshops, except for vandalism (Beelen et al., 2009). On the other hand, indirect crime (crime of the illicit market), could increase or decrease. The impact a coffeeshop has on the illicit market is not unambiguous. For example, it could partly dissolve the illicit drug market, since customers rather legally obtain their drugs instead of illegal. But this could result in other criminal activities (read: worse activities), because illicit drug dealers search for alternatives in order to be compensated for the earnings they miss out on (Beke et al., 2012). In contrast, illicit production supplies coffeeshops, and benefits from the legal cannabis market (Bieleman & Snippe, 2006). Therefore they are probably less likely to commit other crimes. But on the other hand, they could use these obtained assets for other criminal activities (Bieleman & Snippe, 2006). I think that the illicit drug market negatively contributes to the perceived safety in the neighborhood, which in turn would let to decreasing house prices (Pope & Pope, 2012).
The main hypothesis that follows from the discussion above, and is tested in this thesis is: what
effect coffeeshops have on house prices. The models that are used to test this hypothesis is
explained in the Method and Data-section, along with some other interesting hypotheses that arises from the discussion.
