# RECORD cc4e9f217dbac07b4b73c8b633af8dfec0f75eb1
# TITLE: Property price and anti-social behaviour: An empirical study of Northern Ireland
# YEAR: 2025.0  VENUE: International Journal of Strategic Property Management  DOI: 10.3846/ijspm.2025.23983
# AUTHORS: D. Lo; Yao-Xuan Huang; Michael McCord; Y. Yau; John McCord; Tariq Aziz; M. Haran; Lara Al-haddad
# PAGES: 17  CANDIDATE TABLE PAGES (1-based): [7, 8, 9, 10, 11, 12]
# ABSTRACT: This study examines the impact of anti-social behaviour (ASB) on property prices. By analysing over 14,500 market transactions in Northern Ireland, we find that the prevalence of ASB within a neighbourhood exerts a direct and negative influence on house prices, albeit with diminishing effect at the margin. Furthermore, the dampening effects of ASB are more pronounced in districts characterised by higher population density, proximity to the capital city (Belfast), and lower property values. A thorough analysis of district-level data across a wide range of statistical indicators further indicates that the adverse impact of ASB on property prices is most acute in areas marked by social and economic deprivation, including factors such as income, employment, education, and access to services. Lastly, our submarket analysis suggests that the apartment sector and public housing are disproportionately affected by ASB in terms of price depreciation compared to other property types.


===== PAGE 1 (score 13, numeric-paren 0, table-words 0, var-words 13) =====

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), 
which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Copyright © 2025 The Author(s). Published by Vilnius Gediminas Technical University
ISSN: 1648-715X / eISSN: 1648-9179
INTERNATIONAL JOURNAL 
OF STRATEGIC PROPERTY 
MANAGEMENT
2025
Volume 29
Issue 6
Pages 469–485
https://doi.org/10.3846/ijspm.2025.23983
PROPERTY PRICE AND ANTI-SOCIAL BEHAVIOUR: AN EMPIRICAL STUDY OF 
NORTHERN IRELAND
Daniel LO   1*, Yaoxuan HUANG   2, Michael McCORD   1, Yung YAU   3, John McCORD   4,  
Tariq AZIZ   5, Martin HARAN   1, Lara AL-HADDAD   6
1 Faculty of Computing, Engineering and the Built Environment, Ulster University, Belfast, the United Kingdom
2 School of Graduate Studies, Lingnan University, Hong Kong, China
3 School of Graduate Studies and Department of Sociology and Social Policy, Lingnan University, Hong Kong, China
4 Faculty of Law, Ulster University, Belfast, the United Kingdom
5 College of Business Administration, Prince Muhammad Bin Fahd University, Dhahran, Saudi Arabia
6 Faculty of Business, Yarmouk University, Irbid, Hashemite Kingdom of Jordan
Article History:  
Abstract. This study examines the impact of anti-social behaviour (ASB) on property prices. By analysing over 
14,500 market transactions in Northern Ireland, we find that the prevalence of ASB within a neighbourhood 
exerts a direct and negative influence on house prices, albeit with diminishing effect at the margin. Further-
more, the dampening effects of ASB are more pronounced in districts characterised by higher population 
density, proximity to the capital city (Belfast), and lower property values. A thorough analysis of district-level 
data across a wide range of statistical indicators further indicates that the adverse impact of ASB on property 
prices is most acute in areas marked by social and economic deprivation, including factors such as income, 
employment, education, and access to services. Lastly, our submarket analysis suggests that the apartment 
sector and public housing are disproportionately affected by ASB in terms of price depreciation compared to 
other property types.
	■received 23 February 2025
	■accepted 3 April 2025
Keywords: property price, anti-social behaviour, hedonic modelling, Northern Ireland, housing market.
 * Corresponding author. E-mail: d.lo@ulster.ac.uk
1.	Introduction
Anti-social behaviour (ASB) poses a significant challenge 
in built environments, particularly within urban residential 
property markets, where factors such as high population 
density, the close proximity of dwellings, diverse socioeco-
nomic demographics, and frequent tenant turnover inten-
sify its visibility and impact. Broadly defined as actions that 
cause harassment, alarm, or distress to others, ASB encom-
passes a wide array of activities, from noise disturbances 
to vandalism and violence (Anti-social Behaviour, Crime 
and Policing Act 2014, the U.K.). The presence of ASB in a 
neighbourhood can lead to declining property values, cre-
ate an inhospitable living environment, and deter poten-
tial buyers and traders, thereby exerting a negative impact 
on the real estate market. Studies have shown that areas 
with elevated levels of ASB generally experience stunted 
community development as prospective homebuyers and 
investors are dissuaded by the perceived risks associated 
with such neighborhoods, including diminished quality of 
life and safety concerns (Bannister & Kearns, 2013). Con-
sequently, understanding the dynamics of ASB is essential 
for real estate stakeholders, including homebuyers, prop-
erty developers, and urban planners, as it plays a pivotal 
role in shaping housing supply and demand as well as 
market performance.
In the U.K., ASB has developed into a notable soci-
etal issue that cannot be disregarded, particularly within 
the urban centres of major cities with high population 
density. Although the Anti-social Behaviour Act 2003 and 
subsequent legislative measures have been implemented 
to mitigate the detrimental effects of ASB on communi-
ties, its persistence continues to challenge policymakers 
(Home Office, 2014). Despite these legislative efforts, the 
enduring prevalence of ASB underscores the difficulties in 
effectively curbing its impact, rendering it a pervasive and 
complex problem for those tasked with developing and 
enforcing policy (Home Office, 2014). The linkage between 
ASB and real estate markets can indeed be understood 
through the concept of “stigma” associated with certain 
neighbourhoods. As Gibbons (2004) observed, areas bur-
dened with high levels of crime and ASB often develop 
a negative reputation, resulting in reduced demand for 
housing and, consequently, shrinking property values. For 


===== PAGE 7 (score 18, numeric-paren 0, table-words 2, var-words 10) =====

International Journal of Strategic Property Management, 2025, 29(6), 469–485
475
4.	Data and descriptive statistics
The sample data for residential market transactions was 
obtained from various local real estate agencies in North-
ern Ireland, covering the period from the first quarter of 
2018 to the second quarter of 2022. This dataset includes 
sales prices alongside detailed property-level information 
such as building age, property size, address, property type, 
Energy Performance Certificate (EPC) scores and building 
conditions. This data was then merged with a dataset 
from the Northern Ireland Statistical and Research Agency, 
which provides Super Output Area (SOA)-level information 
on ASB and socioeconomic characteristics, as outlined in 
Section 3. In total, approximately 14,500 transaction obser-
vations were utilised for the Hedonic regression modelling.
Table 2. Composition of the sample 
Property type
Detached
Semi-detached
Terraced
Apartment
Townhouse
4018
(29.28%)
4990
(36.36%)
3571
(26.02%)
520
(3.79%)
626
(4.56%)
Belfast
Within Belfast
Outside Belfast
6231
(42.91%)
8290
(57.09%)
Private vs public
Private
Public
12,200
(84.02%)
2321
(15.98%)
Energy Performance 
Certificate
A
B
C
D
E
11
(0.08%)
335
(2.31%)
11,943
(82.26%)
2,226
(15.33%)
4
(0.03%)
Geography 
Rural
Urban
1649
(11.76%)
12377
(88.24%)
Table 3. Descriptive statistics
Aggregate
Variable
Average
Median
Q1
Q3
Stand. Dev.
Property price (£)
193,900
162,000
119,500
230,000
145,124
ASB (per 1000 pop.)
26.51
21.10
14.22
33.10
19.98
Submarket (property type)
Detached
Property price (£)
292,394
250,000
190,000
345,000
194,025
ASB (per 1000 pop.)
18.69
15.00
10.90
22.10
13.07
Semi-detached
Property price (£)
173,908
149,940
131,000
199,950
89,880
ASB (per 1000 pop.)
23.87
19.7
14.00
31.00
15.56
Terraced
Property price (£)
116,392
104,000
81,000
135,000
104,795
ASB (per 1000 pop.)
39.15
32.6
23.2
46
25.95
Apartment
Property price (£)
143,613
129,250
94,950
169,950
92,866
ASB (per 1000 pop.)
31.39
25.4
16
40.3
23.06
Townhouse
Property price (£)
176,391
153,000
125,950
208,000
92,286
ASB (per 1000 pop.)
25.99
20.7
14.8
31.8
17.72
Belfast
Property price (£)
192,864
152,000
110,500
230,000
140,701
ASB (per 1000 pop.)
30.59
25.6
16
39.7
20.87
Non-Belfast
Property price (£)
196,433
167,000
125,000
231,500
148,3666
ASB (per 1000 pop.)
24.07
18.6
13
30.4
19.02
Private
Property price (£)
205,757
167,500
125,000
247,500
143,315
ASB (per 1000 pop.)
24.66
20.2
12.2
30.1
20.85
Public
Property price (£)
185,144
159,950
115,000
220,000
148,490
ASB (per 1000 pop.)
26.18
22.1
13.7
32.6
18.01


===== PAGE 8 (score 41, numeric-paren 0, table-words 10, var-words 17) =====

476
D. Lo et al. Property price and anti-social behaviour: An empirical study of Northern Ireland
Table 2 summarises the composition of the dataset. 
The majority of the observations are either detached 
(29.28%), semi-detached (36.36%), or terraced (26.02%) 
properties, with apartments and townhouses accounting 
for only 3.79% and 4.56% of the sample, respectively. In 
addition, 88.24% of the properties are located in urban 
areas, with 42.91% situated within the Belfast local gov-
ernment district. Notably, a significant proportion of the 
properties (84.02%) were privately built. Regarding EPC 
scores, 82.26% of the observations achieved an average 
energy efficiency rating of C, with an additional 15.33% 
receiving a D rating.
Table 3 provides descriptive statistics of the sample 
data at both aggregate and submarket levels, focusing on 
ASB intensity and property prices. The average and me-
dian house prices in Northern Ireland are approximately 
£193,900 and £162,000, respectively, indicating a market 
skewed towards lower-priced housing. The average ASB 
intensity is 26.51 incidents per 1,000 population per year, 
with the top 25% most ASB-affected SOAs exceeding 
33.10 incidents.
Submarket-level analysis reveals a strong negative cor-
relation between property prices and ASB. For instance, 
terraced housing exhibits the highest average ASB intensi-
ty (39.15 incidents) and the lowest average property price 
(£116,392). Conversely, detached houses, which command 
the highest average price (£292,394), are the least affected 
by ASB, with an average intensity of 18.69 incidents. Other 
property types fall between these two extremes in terms 
of both ASB intensity and property prices.
Furthermore, the data shows a notable difference be-
tween SOAs in Belfast and those in other regions, with 
Belfast averaging 30.59 ASB incidents compared to 24.07 
in other areas, despite generally higher house prices in 
Belfast. Publicly-built residential properties also tend to 
experience higher ASB rates compared to privately-built 
ones. The average and median ASB figures for publicly-
built properties are 26.18 and 22.1, respectively, whereas 
the corresponding figures for privately-built properties are 
24.66 and 20.2.
Our correlation analysis, presented in Table 4, fur-
ther supports the negative relationship between property 
prices and ASB, as evidenced by a correlation coefficient 
of –0.24. This finding provides initial evidence that ASB 
activity within a neighbourhood may depress property val-
ues. Additionally, the analysis indicates a positive correla-
tion between ASB and population density, suggesting that 
ASB is more prevalent in overcrowded neighbourhoods. 
Finally, all socioeconomic ranks, except for accessibility, 
show a negative correlation with ASB, indicating that ar-
eas affected by ASB generally have poorer socioeconomic 
conditions.
5.	Empirical findings and discussion
The findings from the fixed effects hedonic analysis is 
presented in three parts. First we examine the effect of 
ASB at the general level accounting for housing attributes 
and locational dummy variables and develop a series of 
interaction terms to further account for, isolate and de-
termine whether urban population density, property value 
and public housing have an impact upon the nature and 
level of the effect (Models 1–6). Secondly, we develop a 
series of property type models (Models 7–11), to examine 
ASB and scrutinise and distill whether there is a differen-
tial effect on property prices. Lastly, we specify a number 
socioeconomic ranking models which are based on the 
ranking of multiple deprivation measure (Model 12) and 
the underpinning domains of deprivation such as income, 
employment, education, access and crime (Models 13–17) 
to establish whether deprivation impacts on house prices 
with respect to ASB. Indeed, this approach aligns with Mc-
Cann (2013), who suggested that socio-economic factors, 
such as income and employment levels, influence residen-
tial preferences and the value placed on neighbourhoods 
affected by ASB.
Overall, the initial hedonic base models (Models 1 and 2) 
and property attribute models (Models  3–6) in Table  5 
show adjusted R2 values ranging between 63.1% and 
82.3% indicating the coefficients to explain between 63% 
and 82% in the variation of sales price, with the coefficients 
displaying the expected sign, statistical significance and 
reasonable level of magnitudes. For instance, an increase 
in property size increases the price of the dwelling with 
one additional square metre of living space increases the 
price by between 0.51% and 0.84% (Models 1–6). Similarly, 
the age of the property comprises a value significant effect 
ranging between 0.009% and 0.057%, and if the property 
is privately built this exhibits a positive effect ranging be-
tween 5.9% and 16.8% across the models (Models 1–6). 
Inspection of the rural coefficient also reveals a negative 
and statistically significant effect ranging between 3.99% 
and 8.75%, inferring that rural properties are lower in value 
relative to their urban counterparts. This finding resonates 
with the work of Braakmann (2012), who observed that 
urban areas tend to experience a greater depreciation in 
property values due to factors like social disorder, a con-
clusion that is particularly relevant in the context of ASB.
Turning to the parameter of interest, namely ASB, the 
base models (Models 1 and 2) reveal that ASB tends to de-
press house prices, demonstrating that when the number 
Table 4. Correlation analysis
Price
ASB
ASB
–0.23976
1.00000
Density
–0.23766
0.372488
Income rank
0.269692
–0.33513
Employment rank
0.406457
–0.44976
Education rank
0.482588
–0.46102
Accessibility rank
–0.13940
0.390209
Crime rank
0.271278
–0.59419
MDM rank
0.385611
–0.58832


===== PAGE 9 (score 73, numeric-paren 44, table-words 4, var-words 7) =====

International Journal of Strategic Property Management, 2025, 29(6), 469–485
477
action term with ASB (Model 3), the results suggest that 
neighbourhoods of a higher living density tend to have 
a stronger impact of ASB on house prices, i.e., property 
values in more densely populated neighbourhoods are 
more affected by the levels of ASB. In a similar vein, the 
interaction term between price and ASB (Model 4) exhibit 
properties of lower value to be more affected by ASB, with 
a more pronounced negative price effect stemming from 
ASB being observed in public estates/properties than in 
the private counterparts (Model 5). In addition, the results 
show that in terms of property pricing, the capital city of 
Northern Ireland, Belfast, is more affected by ASB as dis-
played by the negative and statistically significant coef-
ficient on ASB × Belfast (Model 6). This finding aligns with 
Besley and Mueller (2012), who highlighted that the effects 
of ASB cases per 1,000 population increases by 10%, prop-
erty prices tend to decrease between 0.39% and 0.95% 
depending on whether marginal effects are accounted for. 
This result is consistent with the findings of Besley and 
Mueller (2012), who found that violence and other socially 
undesirable effects in Belfast resulted in a reduction of 
house prices by 8–14%. As observed in Model 2, a dimin-
ishing marginal effect of ASB on price is also detected as 
indicated by the positive sign of the squared term on ASB. 
This diminishing effect suggests that, as the level of ASB 
rises, the impact on prices becomes less pronounced, simi-
lar to the results of Seo (2018), who found that neighbour-
hood disorder showed decreasing returns in its influence 
on housing prices as disorder increased. When further ac-
counting for urban (neighbourhood) density and its inter-
Table 5. Housing attribute models
Model 1
(Base)
Model 2
(Base 2)
Model 3
(Density)
Model 4
(Property price)
Model 5
(Private)
Model 6
(Belfast)
Constant
10.79874
(0.132176)***
11.18415
(0.130570)***
11.18743
(0.130277)***
11.39189
(0.092940)***
11.31744
(0.128098)***
11.17922
(0.130478)***
ASB
–0.003930
(0.000164)***
–0.009514
(0.000366)***
–0.008774
(0.000376)***
–0.182533
(0.001480)***
–0.009762
(0.000455)***
–0.008565
(0.000417)***
ASB × ASB
4.83×10–5
(2.81×10–6)***
5.22×10–5
(2.84×10–6)***
4.48×10–5
(2.00×10–6)***
4.86×10–5
(2.84×10–6)***
4.61×10–5
(2.85×10–6)***
Age
0.000901
(0.000118)***
0.002997
(0.000507)***
0.002931
(0.000506)***
0.002727
(0.000361)***
0.002858
(0.000507)***
0.002771
(0.000509)***
Age × Age
–1.42×10–5
(3.80×10–6)***
–1.30×10–5
(3.80×10–6)***
–1.47×10–5
(2.71×10–6)***
–1.28 ×10–5
(3.80×10–6)***
–1.28×10–5
(3.81×10–6)***
Size
0.005028
(6.64×10–5)***
0.008369
(0.000173)***
0.008245
(0.000173)***
0.005017
(0.000126)***
0.008363
(0.000173)***
0.008346
(0.000173)***
Size × Size
–8.31×10–6
(3.90×10–7)***
–8.12×10–6
(3.90×10–7)***
–3.92×10–6
(2.80×10–7)***
–8.31×10–6
(3.90×10–7)***
–8.27×10–6
(3.90×10–7)***
Parking onsite
0.085923
(0.010863)***
0.063596
(0.010765)***
0.054154
(0.010803)***
0.027686
(0.007667)***
0.064033
(0.010774)***
0.060777
(0.010773)***
Rural
–0.071657
(0.010610)***
–0.087541
(0.010440)***
–0.088568
(0.010417)***
–0.039933
(0.007440)***
–0.088370
(0.010449)***
–0.079871
(0.010557)***
Private
0.138356
(0.024704)***
0.124989
(0.024144)***
0.134252
(0.024116)***
0.059928
(0.017191)***
0.168379
(0.028833)***
0.124831
(0.024126)***
ASB × Density
–3.74×10–5
(4.60×10–6)***
ASB × Price
0.014814
(0.000125)***
ASB × Private
0.001227
(0.000446)***
ASB × Belfast 
–0.001432
(0.000302)***
Time Dummies
Yes
Yes
Yes
Yes
Yes
Yes
LGD Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Property Type 
Dummies
Yes
Yes
Yes
Yes
Yes
Yes
EPC Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Ext Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Obs
14,521
14,521
14,521
14,521
14,521
14,521
R2	
0.632006
0.651338
0.652927
0.823421
0.651520
0.651877
Adj R2
0.630862
0.650181
0.651752
0.822823
0.650340
0.650698
F
552.4417
563.2336
555.5791
1377.164
552.1440
553.0134
Prob(F)
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
Notes: LGD = Local government district; EPC = Energy Performance Certificate; Ext = Condition of external repair. *** indicates 1% statistical significance. 


===== PAGE 10 (score 77, numeric-paren 34, table-words 6, var-words 19) =====

478
D. Lo et al. Property price and anti-social behaviour: An empirical study of Northern Ireland
of ASB are more pronounced in urban centres such as 
Belfast, where population density and social heterogeneity 
likely exacerbate the effects of ASB.
Table 6 exhibits the role of ASB by property type (Mod-
els 7–11). The findings reveal that amongst all property 
types, apartments seemingly are the most affected by ASB 
(–1.307, p < .001), followed by semi-detached, terraced, 
townhouse and detached with an ASB-induced property 
price discount ranging between 0.66% to 1.13%. This find-
ing suggests that this may be largely due to the effect 
of density, the concentration of younger people gather-
ing in more urbanized areas or “hot spots” of highly con-
centrated space and the potential inter-heterogeneity in 
areas or surrounding areas and how the neighbourhood 
structure is defined. This is concomitant with the findings 
of Ceccato and Wilhelmsson (2011) who observed that ad-
jacency to public places can impact upon, and reveal signs 
of physical and social disorder in which environmental 
cues translate into weakened social ties among residents, 
a negative demand for housing in an area and a nega-
tive effect on property prices. Indeed, of particular note, 
Ceccato and Wilhelmsson (2011) observed vandalism to 
have a significant and independent effect on apartment 
prices in Stockholm municipality indicating that the effects 
of crime vary by housing type and have differential effects 
on house prices. 
Examination of the role of deprivation and its domains 
and the effects of ASB on house prices are presented in 
Table  7. As observed in Model 12 which takes into ac-
count the headline measure of multiple deprivation rank-
ing, the effects of ASB on house price are more significant 
in neighbourhoods of higher multiple deprivation scores. 
This finding is in line with Knox (2011), who argued that 
higher levels of deprivation correlate with increased ASB, 
and suggests that deprived areas are more vulnerable 
to the negative price effects of ASB. Further isolating 
the effects of ASB on house prices applying the sepa-
rate domains of deprivation (Models 13–17) reveals that 
the effects of ASB on house price are more significant 
in lower-income neighbourhoods and in neighbourhoods 
comprising higher unemployment rates. Moreover, the re-
sults show the effects of ASB on house price are more 
significant in neighbourhoods of lower educational attain-
ment, reduced accessibility to services and in neighbour-
hoods which constitute higher crime rates. These results 
align with Haydon and Scraton (2000), who noted that 
socio-economic factors, such as poverty and inequality, 
exacerbate ASB in Northern Ireland, contributing to its 
Table 6. Property type models
Model 7
(Detached)
Model 8
(Semi Detached)
Model 9
(Terraced)
Model 10
(Apartment)
Model 11
(Townhouse)
Constant
11.43231
(0.218770)***
10.48241
(0.191199)***
10.95269
(0.156221)***
11.50655
(0.223572)***
10.53464
(0.201467)***
ASB
–0.006672
(0.000834)***
–0.011330
(0.000732)***
–0.007973
(0.000642)***
–0.013072
(0.002463)***
–0.006937
(0.001467)***
ASB × ASB
2.65×10–5
(7.97×10–6)***
6.88×10–5
(7.21×10–6)***
4.01×10–5
(4.21×10–6)***
0.000104
(2.39×10–5)***
1.80×10–5
(1.28×10–5)
Age
0.006723
(0.000856)***
0.000851
(0.000923)
–0.002881
(0.001315)**
–0.005064
(0.002595)**
0.003915
(0.002399)*
Age × Age
–3.41×10–5
(6.40×10–6)***
9.60×10–6
(7.21×10–6)
1.90×10–5
(9.28×10–6)**
6.06×10–5
(1.95×10–5)***
–2.86×10–5
(1.86×10–5)
Size
0.007221
(0.000264)***
0.010776
(0.000460)***
0.008178
(0.000864)***
0.007011
(0.001226)***
0.011030
(0.000841)***
Size × Size
–6.12×10–6
(5.23×10–7)***
–1.40×10–5
(1.28×10–6)***
–5.60×10–6
(3.44×10–6)*
–1.96×10–5
(3.97×10–6)***
–1.50×10–5
(2.39×10–6)*** 
Parking onsite
–0.011208
(0.043728)
0.063166
(0.030378)**
0.027108
(0.014338)*
–
0.091669
(0.029170)***
Rural
–0.079585
(0.015479)***
–0.107333
(0.021478)***
–0.050162
(0.027596)*
0.044664
(0.083693)
–0.019899
(0.038885)
Private
–0.008144
(0.089321)
0.231046
(0.046031)***
0.070744
(0.036865)**
0.275627
(0.076130)***
0.277295
(0.067088)***
Time Dummies
Yes
Yes
Yes
Yes
Yes
LGD Dummies
Yes
Yes
Yes
Yes
Yes
EPC Dummies
Yes
Yes
Yes
Yes
Yes
Ext Dummies
Yes
Yes
Yes
Yes
Yes
Obs
4,018
4,990
3,571
520
626
R2
0.523332
0.471448
0.379724
0.492908
0.790389
Adj R2
0.518053
0.466960
0.372340
0.461798
0.777955
F
99.13489
105.0602
51.42368
15.84406
63.56403
Prob(F)
0.000000
0.000000
0.000000
0.000000
0.000000
Notes: LGD = Local government district; EPC = Energy Performance Certificate; Ext = Condition of external repair. *, **, *** indicate 10%, 5% and 1% statistical 
significance respectively.


===== PAGE 11 (score 81, numeric-paren 41, table-words 8, var-words 9) =====

International Journal of Strategic Property Management, 2025, 29(6), 469–485
479
more severe effects on property values. Viewed from an-
other perspective, high-income households appear will-
ing to pay a price premium to distance themselves from 
areas affected by anti-social behaviour and its associated 
activities. This observation aligns with the conceptual ar-
gument put forth by McCann (2013), who suggests that 
income-based residential segregation emerges as affluent 
individuals increasingly cluster within certain neighbour-
hoods, driven not only by considerations such as com-
muting costs and proximity to employment centres but 
also by the desire to secure a safer and more desirable 
living environment.
Robustness check3
As a robustness check for potential spatial autocorre-
lation in property prices, as well as omitted variables that 
could bias our results, we developed three spatial lagged 
3	 We thank an anonymous reviewer for their valuable feedback 
on the robustness of our empirical models, particularly regard-
ing potential spatial autocorrelation/dependence in property 
prices, endogeneity, and omitted variable bias.
models4 (SLMs) (Models 18–20) to incorporate a spatial 
lag (SL) term, 
(
)
1
ln
m
j
j
Price
=∑
, in order to account for spatial 
dependence in housing prices (Equation (2)). The spatial 
lag (SL) captures the price effects of neighbouring prop-
erties, j, sold within two months of the subject property’s 
transaction. m is the number of prior transactions. w is 
the spatial weight that governs the structure of the spa-
tial dependence, defined as 1/d, where d represents the 
4	 According to Anselin (1988), spatial lag models address omitted 
variable bias by incorporating the spatial dependence between 
observations, where the value of the dependent variable in one 
location may be influenced by values in neighbouring locations. 
By including a spatially lagged dependent variable, the model 
controls for unobserved factors that are shared across proper-
ties, thus reducing the potential for omitted variable bias that 
arises from ignoring spatial correlations (LeSage & Pace, 2009). 
This approach ensures that the effects of local unobserved fac-
tors, which might affect both the dependent variable and inde-
pendent variables, are captured, leading to more accurate and 
unbiased estimates.
Table 7. Socioeconomic ranking models
Model 12
(MDM)
Model 13
(Income)
Model 14
(Employment)
Model 15
(Education)
Model 16
(Access)
Model 17
(Crime)
Constant
11.36294
(0.126073)***
11.30127
(0.128444)***
11.31977
(0.126192)***
11.34182
(0.125968)***
10.81866
(0.132222)***
10.97523
(0.083030)***
ASB
–0.013723
(0.000376)***
–0.013102
(0.000393)***
–0.012840
(0.000368)***
–0.012497
(0.000364)***
–0.005718
(0.000502)***
–0.010956
(0.000384)***
ASB × ASB
5.20×10–5
(2.71×10–6)***
5.20×10–5
(2.71×10–6)***
4.49×10–5
(2.72×10–6)***
4.14×10–5
(2.72×10–6)***
4.82×10–5
(2.85×10–6)***
5.57×10–5
(2.86×10–6)***
Age
0.001584
(0.000491)***
0.002053
(0.000500)***
0.002412
(0.000490)***
0.001954
(0.000490)***
0.000785
(0.000122)***
0.002597
(0.000505)***
Age × Age
–3.94×10–6
(3.68×10–6)
–3.94×10–6
(3.68×10–6)
–1.05×10–5
(3.68×10–6)***
–8.68×10–6
(3.67×10–6)***
–1.42×10–5
(3.81×10–6)***
–1.04×10–5
(3.80×10–6)***
Size
0.007935
(0.000167)***
0.008168
(0.000170)***
0.008002
(0.000168)***
0.007704
(0.000168)***
0.005029
(6.64×10–5)***
0.008321
(0.000170)***
Size × Size
–7.59×10–6
(3.77×10–7)***
–8.00×10–6
(3.84×10–7)***
–7.71×10–6
(3.77×10–7)***
–7.26×10–6
(3.77×10–7)***
–8.30×10–6
(3.90×10–7)***
–8.15×10–6
(3.76×10–7)***
Parking onsite
0.037509
(0.010415)***
0.037509
(0.010415)***
0.055876
(0.010401)***
0.036991
(0.010409)***
0.086069
(0.010858)***
0.054918
(0.010735)***
Rural
–0.057876
(0.010111)***
–0.057876
(0.010111)***
–0.067325
(0.010103)***
–0.061424
(0.010095)***
–0.064312
(0.010782)***
–0.082464
(0.010363)***
Private
0.030205
(0.023468)
0.030205
(0.023468)
0.035640
(0.023485)
0.042281
(0.023409)***
0.133855
(0.024722)***
0.107201
(0.024071)***
ASB × Z
2.47×10–6
(7.50×10–8)***
2.47×10–6
(7.50×10–8)***
2.24×10–6
(6.94×10–8)***
2.62×10–6
(7.90×10–8)***
4.79×10–7
(1.27×10–8)***
1.24×10–6
(1.03×10–8)***
Time Dummies
Yes
Yes
Yes
Yes
Yes
Yes
LGD Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Property Type 
Dummies
Yes
Yes
Yes
Yes
Yes
Yes
EPC Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Ext Dummies
Yes
Yes
Yes
Yes
Yes
Yes
Obs
14521
14521
14521
14521
14521
14521
R2
0.675570
0.675570
0.674713
0.675966
0.651339
0.654687
Adj R2
0.674471
0.674471
0.673611
0.674868
0.650158
0.653566
F
614.9659
614.9659
612.5683
616.0784
551.7033
583.8237
Prob(F)
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
Notes: LGD = Local government district; EPC = Energy Performance Certificate; Ext = Condition of external repair. *** indicates 1% statistical significance. 


===== PAGE 12 (score 48, numeric-paren 23, table-words 8, var-words 2) =====

480
D. Lo et al. Property price and anti-social behaviour: An empirical study of Northern Ireland
Euclidean distance between property i and property j. Ω is 
a coefficient to be estimated, representing the weighted 
average effect arising from the spatial dependence of 
neighbouring properties. Models 18 and 19 are reformu-
lations of Models 1 and 2, respectively, which account for 
the spatial dependence of property prices.
(
)
(
)
1
1
1
1
1
1
Ln
 ln
   
  ,
m
m
j
j
i
j
j
j
n
r
s
v
k
k
p
p
q q
u
u
k
p
q
u
Price
c
w
Price
ASB
S
LGD
PT
T
Z
e
=
=
=
=
=
=
γ
+
=
+
+ α ×
+
β
+
+
γ
µ
+
µ
+
∑
∑
∑
∑
∑
∑
Ω
	 (2)
We further develop Model 20, which follows a two-
stage least squares (2SLS) regression approach to address 
potential endogeneity in the relationship between ASB and 
property prices. In the first stage, we regress ASB (the en-
dogenous variable) on an instrumental variable– previous 
ASB values in 2010 – and a set of control variables above-
mentioned. The instrumental variable, previous ASB, is as-
sumed to be correlated with current ASB but not directly 
with property prices, making it a valid instrument. The pre-
dicted values of ASB from this first-stage regression,  
ASB , 
are used as the instrumented variable in the second-stage 
regression. In the second stage, we estimate the effect of 
ASB on property prices by regressing the natural logarithm 
of property prices on the predicted values of ASB (from 
the first stage) and other control variables. This two-stage 
approach allows us to isolate the exogenous variation in 
ASB, removing any bias that may arise from simultane-
ity or omitted variable bias in the relationship between 
ASB and property prices. Table 8 presents the results for 
the three robustness models. For Models 18 and 19, we 
observe that the estimates for the main variables, includ-
ing ASB, ASB × ASB, and other control variables, remain 
largely consistent with those of the non-spatial models 
(Models  1 and 2) in terms of both the signs and mag-
nitudes of the coefficients. Additionally, both the R2 and 
adjusted R2 values show a moderate improvement with 
Table 8. Results of robustness models
Model 18
(SLM 1)
Model 19
(SLM 2)
Model 20
(2SLS)
Constant
9.234866
(0.098423)***
10.03455
(0.122770)***
11.65343
(1.53242)**
ASB
–0.002611
(0.000755)***
–0.004122
(0.000978)***
ASB × ASB
3.42×10–5
(1.61×10–6)***

ASB
–0.004934
(0.000723)**


×    
ASB
ASB
3.72×10–5
(1.43×10–6)***
Age
0.001332
(0.000051)***
0.002344
(0.000007)***
0.002423
(0.000126)***
Age × Age
–1.11×10–5
(3.01×10–6)***
–1.24×10–5
(1.72×10–6)***
Size
0.004137
(6.23×10–5)***
0.006478
(0.000154)***
0.005343
(0.000243)***
Size × Size
–5.21×10–6
(2.82×10–7)***
–2.66×10–6
(1.23×10–7)***
Parking onsite
0.077516
(0.019778)**
0.075456
(0.006855)***
0.043993
(0.009233)***
Rural
–0.043112
(0.009224)***
–0.064521
(0.008455)***
–0.068233
(0.008433)***
Private
0.163219
(0.022415)***
0.138541
(0.012894)***
0.112243
(0.015488)***
SL Term
1.418073
(0.057915)***
1.219452
(0.048232)***
1.218323
(0.048482)***
Time Dummies
Yes
Yes
Yes
LGD Dummies
Yes
Yes
Yes
Property Type Dummies
Yes
Yes
Yes
EPC Dummies
Yes
Yes
Yes
Ext Dummies
Yes
Yes
Yes
Obs
13,421
13,421
13,421
R2
0.693112
0.715510
0.703323
Adj R2
0.690094
0.711445
0.700238
F
651.3461
655.3474
647.5213
Prob(F)
0.000000
0.000000
0.000000
Notes: ** and *** indicate 5% and 1% statistical significance respectively.


===== PAGE 13 (score 11, numeric-paren 0, table-words 1, var-words 9) =====

International Journal of Strategic Property Management, 2025, 29(6), 469–485
481
the incorporation of the spatial lag terms, which exhibit a 
positive and statistically significant coefficient, indicating 
spatial dependence in property prices. Lastly, Model 20 
presents the results of the 2SLS model. The coefficients 
on the predicted values of ASB (i.e., 
ASB) as well as its 
squared terms exhibit values and statistical significance 
consistent with those of the non-2SLS models, thereby 
confirming the robustness of our results.
6.	Discussion
The findings of this study provide robust empirical evi-
dence that ASB exerts a significant and negative influence 
on residential property prices in Northern Ireland, after 
accounting for property-level and temporal attributes. 
These findings resonate with the established literature, 
which has long identified a negative relationship between 
socially undesirable behaviours and property values (e.g., 
Braakmann, 2012). Through the application of hedonic 
regression models, the analysis demonstrates that areas 
with higher incidences of ASB experience notable depre-
ciation in property values, a trend that largely aligns with 
previous studies conducted in the U.K. and internationally. 
Specifically, the results underscore that the impact of ASB 
on property prices is neither constant nor linear across 
different spatial and socioeconomic settings of the prop-
erty market. Instead, it is highly dependent on a variety 
of property-level and district-specific factors. As Bonakdar 
and Roos (2023) highlighted, the influence of neighbour 
quality on property prices varies significantly depend-
ing on the characteristics of both the property and the 
neighbourhood. This variability suggests that ASB’s effect 
is, contrary to common perceptions, multifaceted, influ-
enced by the characteristics of individual properties and 
the broader social and economic context of the neigh-
bourhoods in which they are situated. For instance, while 
all property types experience depreciation in response to 
ASB, the magnitude of this effect varies significantly, high-
lighting the complex nature of ASB’s impact on housing 
markets. This complexity was also observed by Marsden 
(2018), who noted that common ASB-related neighbour 
disputes, such as noise and public hygiene issues, are 
more likely to reduce property values in urbanised set-
tings with higher density, a point that is central to our 
study’s findings.
One of the pivotal findings of this study is that the ad-
verse impact of ASB on property values is markedly more 
pronounced in urban areas, with Belfast emerging as a 
particularly affected locale. This is in line with the argu-
ments made by Besley and Mueller (2012), who found that 
urban environments, particularly those with high popula-
tion densities and historical socio-political divisions, tend 
to experience greater negative effects from ASB. Urban 
environments, defined by higher population densities and 
pronounced social heterogeneity, exhibit a heightened 
vulnerability to the pernicious effects of ASB. In Belfast, 
the city’s diverse demographic fabric–comprising varied 
religious, cultural, and educational backgrounds–serves 
as a potential catalyst for ASB. The social tensions and 
frictions that naturally arise from these differences may 
exacerbate levels of social disorder, thereby amplifying 
the prevalence and intensity of ASB. This dynamic under-
scores the complex interplay between urban diversity and 
social cohesion, suggesting that the negative externalities 
of ASB in cities like Belfast are deeply intertwined with the 
broader sociocultural landscape. Consequently, addressing 
ASB in such contexts requires a nuanced understanding of 
the underlying social dynamics that fuel these behaviours.
The differential impact of ASB on various property 
types is another noteworthy finding. Apartments, in par-
ticular, are more adversely affected by ASB in terms of 
property valuation, whereas detached houses experience 
the least impact. Several factors may explain this disparity. 
As Ceccato and Wilhelmsson (2011) pointed out, apart-
ment buildings are more susceptible to the negative ef-
fects of social disorders due to shared spaces that fos-
ter communal disturbances, such as noise nuisance, pets’ 
barking, and public hygiene concerns. Apartments typically 
have more communal areas, such as shared hallways, stair-
cases, and outdoor spaces, which can serve as hotspots for 
ASB. Moreover, issues like noise nuisance, pets’ barking, 
and defecation are more common and noticeable in apart-
ment settings due to the closer proximity of housing units. 
The lower tolerance levels for such disturbances among 
apartment dwellers, combined with the higher visibility of 
ASB in these settings, likely contribute to the more sub-
stantial price reductions observed in apartment markets. 
In contrast, detached houses, often located in less dense 
and more socially stable and prestigious areas, are less ex-
posed to these issues and therefore less impacted by ASB.
Privately built houses also display a markedly greater 
resilience against the price-depressing effects of ASB com-
pared to their publicly developed counterparts. This could 
be attributed to better management practices and strat-
egies typically employed in private developments. Such 
developments usually have access to more substantial re-
sources and possess stronger incentives to proactively ad-
dress and mitigate issues related to ASB. Conversely, public 
housing estates, frequently situated in socioeconomically 
disadvantaged areas, are often hindered by insufficient so-
cial infrastructure and underfunded management, which 
impairs their ability to effectively counteract the adverse 
consequences of ASB. This observation underscores the 
pivotal role that governance and management practices 
within housing developments play in determining the ex-
tent to which ASB can influence property values. Therefore, 
the resilience of privately built homes against ASB-induced 
price depreciation may be indicative of superior govern-
ance structures that prioritise the maintenance of property 
value and community well-being.
The study further elucidates the subtle interconnection 
between ASB and the social and economic characteristics 
of neighbourhoods. This is in line with the literature on 
socio-economic segregation and residential preferences. 
As McCann (2013) suggests, high-income households tend 
to avoid areas with higher levels of ASB, driving a form