Home, the door and the doorstep

Burglary sits in an odd place among the crimes people fear. It is far from the worst harm this course deals with, yet lesson 1 showed it topping a city's fear rankings above murder, because it seems serious enough and likely enough at once. It is also where the strongest evidence in this whole course is. A long run of survey data from England and Wales lets researchers compare households that had particular locks and lights with households that had none, and the difference is large. This lesson teaches that evidence and its limits, sets out the one piece of home security that is really disputed, says what the data show about burglaries with someone at home, and ends at the front door, where the con artist and the burglar turn out to use the same move.

That responsibility sentence matters more than usual here, because home security is the place where "you should have locked it" comes most easily to people's lips. Nothing in this lesson says a household that was burgled did something wrong. A lock changes how hard a house is to get into. It has nothing to do with who chose to break in.

What made burglary fall

In the 1990s burglary in England and Wales began a long fall, and criminologists argued about why. Tseloni, Farrell, Tilley and colleagues tested one explanation, the security hypothesis: that households fitted better locks and lights, and burglars found fewer houses they could get into. Their 2017 paper in Crime Science used the Crime Survey for England and Wales from 1992 to 2011/12, which asks households both whether they were burgled and which security devices they had.¹

The paper draws the same conclusion itself: the decline was "solely in forced rather than unforced entries to households", and the main drop came when the house was empty, which is when locks are locked and timer lights are on.¹ Put with the rest of their evidence, the authors wrote: "The study concludes that there is strong evidence that security caused the decline in burglary in England and Wales in the 1990s."¹

Notice what kind of evidence that is. Nobody randomly assigned locks to houses. The authors built a case from several survey patterns pointing the same way: security spread just before burglary fell, households with no security became far riskier than everyone else, and the fall came only where security had to be overcome. It's a strong comparison rather than a trial, which is why lesson 1 used it as the example of the label tested: a strong comparison that measured the outcome that matters.

WIDE, and what "almost 50 times" means

The paper also reports which devices did the work, from the same team's analysis of the 2008/09 to 2011/12 surveys. The measure is the security protection factor (SPF). It compares the burglary risk of households with a given device or set of devices against households with "no security", meaning none of the devices the survey asked about. An SPF of 13 means households with those devices had about a thirteenth of the risk of burglary with entry that no-security households had.¹

Single devices did less on their own. Window locks had an SPF of 7; indoor lights on a timer, double or deadlocks on the doors, and external lights on a sensor had 3 each. Combinations did much better:¹

The four together are what the team calls WIDE: "properties fitted with the four 'WIDE' devices: window locks (W), indoor lights on a timer (I), door double or deadlocks (D) and external lights on a sensor (E) have an SPF of 49."¹ The authors compare that with the four single SPFs added together, 16, and treat the difference as what combining them adds. In the discussion they put it plainly: WIDE produces "a very substantial (almost 50 times) lower risk of burglary with entry than 'no security'".¹

That number gets quoted more than any other in this course, and it's easy to misread in three ways.

It's against no security at all. The comparison group is households with none of the listed devices, and by 2008/09 to 2011/12 that was under one household in twenty (4.9%), down from about one in six in 1992 to 1996 (on a shorter list of devices, which the authors say may partly explain the higher earlier figure). Those households had become nearly eight times as likely as the average household to be burgled.¹ Most homes already have some of the devices. So "almost 50 times" isn't what adding WIDE to a typical home does, and "WIDE cuts your risk by 98%" is the wrong sentence: it takes a comparison against the rarest and riskiest group and turns it into a promise about yours.

It's a comparison between households, not a trial. Households that fit four devices may differ in other ways from households that fit none. The team's wider case is what makes them confident the devices themselves matter, but no one put WIDE on some homes and not others and waited.

It counts devices, not how good they are or whether they were used. In the authors' words, the survey "has asked only about the presence or absence of security devices. It has not been designed to capture their quality, nor whether the devices were in use at the time of the victimisation."¹ A window lock left unlocked every night is still a window lock in the survey. They add that the survey could not test measures aimed at whole areas (planning standards for new housing, street lighting, gates across the alleys behind houses), so security's full role may be "understated".¹

Callum and Sian ask the landlord

Callum and Sian are moving into a rented ground-floor flat in Lincoln. The front door has a latch but no deadlock. The windows close with catches, not locks. There's no outside light, and neither of them owns a plug-in timer. They want to know what to ask the landlord for and what to do themselves.

Step 1: list what the flat has against WIDE. Window locks: no. Indoor lights on a timer: no. Door double or deadlock: no. External light on a sensor: no. It has none of the survey's other devices either (no alarm, door chain or window bars), so by the survey's definition the flat as it stands is a no-security household.

Step 2: sort who does what. The locks and the outside light are fixtures, so they are for the landlord to agree to. A timer for an indoor lamp is a small thing the tenants can buy and plug in themselves. The authors themselves suggest that minimum standards set for landlords are one way to get these devices to the people most at risk, which gives the request some backing, though it is their suggestion, not a rule anywhere this course read.¹

Step 3: price it. They look up, for their own area, what a deadlock, window locks and an outside sensor light would cost to fit. This course names no products and no prices, and it doesn't teach fitting a lock.

Step 4: label it. Each request is tested, in this course's labels, in the sense lesson 1 gave: a strong comparison between households in England and Wales measured burglary with entry, against households with no security at all.

The alarm dispute

Look back at the table. There's no burglar alarm in it. In the same England and Wales data, the 2017 paper reports that "somewhat counterintuitively, alarms were not identified as contributory", and it recommends caution: "It would be prudent to think twice before recommending an alarm as part of a standard package of security devices to reduce the risk of domestic burglary."¹ A 2015 paper by Tilley, Tseloni and colleagues went further. From its abstract: "It finds that alarms have become associated with increased rather than decreased risk of burglary with entry. This counter-intuitive finding needs to be treated cautiously."²

Tilley and colleagues offered several explanations, and they tested none of them. Beside the timing one, they call a second plausible, though they had no data to test it: that an alarm can flag that something worth stealing is inside. The others include that offenders and police have adapted (police no longer always attend an alarm) and that many alarms are poor-quality or do-it-yourself systems.² Each is a hypothesis, and nobody has yet shown which, if any, is right. The authors themselves warn that, given earlier research suggesting that alarms have been effective deterrents, it "would be premature to conclude that domestic burglar alarms are (or have become) counter-productive", and they call for more research.²

The other side of the question comes from the US, where one study asked burglars. A team at the University of North Carolina at Charlotte surveyed 422 people imprisoned for burglary in three states for a 2012 report, Understanding Decisions to Burglarize from the Offender's Perspective. This course read the university's release about it, not the report. From the release: "Approximately 83 percent said they would try to determine if an alarm was present before attempting a burglary, and 60 percent said they would seek an alternative target if there was an alarm on-site." And: "Among those who discovered the presence of an alarm while attempting a burglary, half reported they would discontinue the attempt, while another 31 percent said they would sometimes retreat."³ A Rutgers study of burglar alarms in Newark, New Jersey, by S. Lee, looked at whole neighbourhoods instead of single homes. Again this course read only the university's release, which reports that "Neighborhoods in which burglar alarms were densely installed have fewer incidents of residential burglaries than the neighborhoods with fewer burglar alarms", and calls it the first study of alarms "scientifically ruling out other factors that could have impacted the crime rate".⁴ Its advisory committee included Felson and Clarke, whose work lesson 4 was built on, with Clarke as its chair.⁴

Who paid for each is worth knowing. The two England and Wales papers were funded by the Economic and Social Research Council, a public research funder, and the 2017 paper says the funder had no role in its design, analysis or writing.¹² Both US studies were funded by the Alarm Industry Research and Educational Foundation, which the Charlotte release describes as working under the auspices of the largest trade association for the electronic security industry.³⁴ None of that tells you a result is wrong, and this course makes no such suggestion about either side: the way to judge a study is by its design.

Can both be right?

Yes, because they measured different things. Burglars asked in prison said what they would do on meeting an alarm. The England and Wales survey measured which households, with and without alarms, were actually burgled. On one of the explanations the 2015 paper calls plausible, an alarm might put some burglars off a house while alarmed houses, on average, are still riskier because some of them were targets before the alarm went in. The Newark study counted burglaries, but by neighbourhood rather than by household, so it answers a different question again. The two sets of studies also come from different countries, with different alarms and police response.

So whether an alarm protects a home is a contested question in this course's labels. What would settle it is a study that records when each alarm was fitted relative to any earlier burglary, and follows households afterwards, or better, one that fits alarms to some homes at random and not to others. The consensus on locks and lights is much firmer. If money is short, the evidence says spend it on the combination in the table first. That's the course's reading of the England and Wales evidence; an alarm on top is a reasonable choice, and the evidence points both ways and settles neither.

How burglars get in, and whether anyone's home

The current figures are the Office for National Statistics' tables on the nature of burglary, which run from the year ending March 2015 to the year ending March 2025 (there are no survey data for the two pandemic years).⁵ They count household incidents of domestic burglary in a dwelling where the burglar got in, and each year's figure rests on the incidents the survey's sample picked up, between about 120 and 340 a year depending on the question. So look at the range across the years rather than at any single year, which is lesson 2's rule about one year and a trend.

• Someone was at home in between 51% and 68% of burglaries with entry each year.
• The way in was a door in about two-thirds to four-fifths of them, and the door was not locked in between about a fifth and a third of them (18% to 35%).
• The burglar pushed past whoever opened the door in between 7% and 12%, and got in by false pretences in between about 3% and 8%.

The unlocked door is the row a household can act on. A lock only helps when it's locked, and the step from that row to keeping the outside doors locked while you're in is this course's: sensible, untested, in this course's labels, since nobody has trialled it.

The US picture comes from an older report. The Bureau of Justice Statistics looked at the National Crime Victimization Survey for 2003 to 2007 and found: "In about 28% of these burglaries, a household member was present during the burglary. In 7% of all household burglaries, a household member experienced some form of violent victimization."⁶ Where nobody was home, unlocked ways in were common there too: "In 40% of unlawful entries to unoccupied residences, offenders gained access through an unlocked door or window".⁶ Where someone was at home and the offender got in without force, the report's table shows someone inside letting the offender in in 18% of those burglaries, and someone opening the door and the offender pushing in in 12%.⁶

The two "someone at home" figures are different measures, and this course keeps them apart:

The definitions and periods differ, so the gap between the columns isn't a finding about the two countries. What each column says within itself is the finding.

Two slogans, both too simple

Those figures answer two popular beliefs that point in opposite directions.

"Burglars strike when nobody's home." In England and Wales, more often than not, someone was. You may meet a respectable version of this: lesson 1's Warr wrote in 2000 that "burglaries normally occur when no one is home".⁷ That summarised older US research, and it fits the US report above, where no one was home in most burglaries. It doesn't fit the current England and Wales tables.

"Burglars attack people at home." Being home is common; violence isn't the usual outcome. In the US data, a household member suffered violence in 7% of all household burglaries.⁶ In England and Wales the survey asks a narrower question. Where someone at home knew about the burglary and saw the burglar (between about a fifth and a third of burglaries with entry each year), more than half of those incidents involved some threat or use of force in every year's table. Each year rests on fewer than 90 such incidents, and ONS says estimates based on fewer than 50 cases, as several of these are, should be treated with caution.⁵ Small as the counts are, every year points the same way: where someone at home saw the burglar, threat or force was more common than not. The table doesn't say who went to whom.

If you think someone is inside

The two sources this course read on it agree: a UK charity, and a leaflet from the prosecutors and police chiefs of England and Wales.

The Suzy Lamplugh Trust, the UK personal safety charity lesson 4 met, says that if you are in your home and think there is an intruder, "call the police", and "don't go and investigate by yourself". And: "If you come home and suspect an intruder has been inside, don't enter the property."⁸ For that case, where you come home to it, its page adds that you can use a mobile or a neighbour's phone for the call. The leaflet for householders written by the Crown Prosecution Service and the National Police Chiefs' Council, which is mostly about the law, says: "Wherever possible you should call the police."⁹

That's a crime in progress, the case GOV.UK's page on contacting the police gives for 999: "Call 999 if you are reporting a crime that is in progress or if someone is in immediate danger."¹⁰ In the US, the Office for Victims of Crime says to call 911 if you feel in immediate danger.¹⁰ This course read no US source that speaks to an intruder in the home in particular: the Office for Victims of Crime's line is its general one, and the Lamplugh advice is written for the UK. The callout at the top of this lesson has the silent-call steps for both.

How to label it: none of the sources read tested the advice, so it isn't tested. It's sensible, untested. The survey figures above are the evidence nearby, and the step from them is this course's: if seeing the burglar is when threat or force tends to come in, keeping your distance and calling is the option that avoids that moment. People who did go and look, or who found the burglar in front of them, did nothing wrong. The burglar made that moment, and this is about the options open to you if it ever happens.

What the law says if it comes to force in your own home is lesson 14's subject, "The home and property". This lesson states none of it.

The weeks after a burglary

Tseloni's paper repeats, citing earlier research, that a previous burglary is the strongest single predictor of the next one, and it suggests that the WIDE devices be targeted on the households most at risk, which includes those burgled already.¹

The doorstep

Pushing past whoever opened the door and false pretences turn up in the England and Wales tables every year. In the US report, among burglaries of occupied homes where no force was needed, nearly a fifth began with someone inside letting the offender in; the table doesn't say whether those let in were strangers with a story or people the household knew. Beside the forced entries whose fall opened this lesson, Tseloni's paper puts entries by a key, pushing past, or false pretences in a third group, "other", and its text gives no trend for it. The point that follows is this course's, not the paper's: a lock only works on someone who has to get past it, and a con at the door gets you to open it.

So the door itself is where the second half of home security lives. The Suzy Lamplugh Trust's advice on callers is plain, and it's practitioner opinion, in this course's labels.⁸ Use the spy hole, chain, intercom or outside light if you have them, to see who is there. If someone says they represent a company or organisation, ask for ID. And never let the ID check itself: "Use the telephone number given in the phone book or on your utility bill, rather than the one on the identity card." The same page explains why the back of the house matters too: "a bogus caller will try and distract you whilst an accomplice robs you, often by entering through a back door." That's what's called a distraction burglary: one person keeps you at the front while another gets in. So lock the other doors before you answer.

The US National Crime Prevention Council's page on home and neighbourhood safety adds two habits that fit the same picture, also practitioner opinion: keep a spare key with a trusted neighbour, "not under a doormat or planter", and "Illuminate or eliminate places an intruder might hide".¹²

If a caller has ever talked their way past you, the con worked as it was built to. The person who ran it is responsible, and what follows is about the options open next time.

Harold and the man from the water company

Harold is 79 and lives alone. Mid-morning, a man rings the bell. He says he's from the water company, that there has been a problem with the mains in the street, and he needs to check the kitchen tap straight away. Harold can see a second man standing by his back gate.

Nothing has gone wrong yet. Here's how it can go, using the Lamplugh steps, in words he could say.

First, the door stays between them. If Harold has a chain, he puts it on. He can talk through the gap or through the door. The Lamplugh advice is to see who is there before opening at all.

Second, the back. A second man by the back gate is the pattern the Lamplugh page warns about: one at the front to distract, one at the back. So before anything else, the back door gets locked. Harold can say, "Wait there a moment," and do that first.

Third, the ID, and the call-back. "Can I see your ID? Thank you. I'm going to ring the water company on the number on my bill. Please wait outside." Then he closes the door and phones the number printed on his own bill, not any number on the card or any number the man gives him.

Fourth, what the answer tells him. If the company confirms a visit, the man has waited and lost a few minutes. If it doesn't, the door stays shut. If either man tries to get in, that's a crime in progress and a 999 call. If they simply leave, GOV.UK's route for a crime that is not an emergency is to report it online or on 101.¹⁰

How to label the steps: the call-back rule is sensible, untested. The Lamplugh Trust gives it for the door and the FTC for messages, but nobody has run a trial of it. The accomplice warning and the ID steps are practitioner opinion.

One pattern: the door and the phone

The man at Harold's door and the voice on the phone saying your bank account is compromised are running the same play, and two official bodies describe it in nearly the same terms.

The US Federal Trade Commission's page on avoiding scams names four signs: "Scammers PRETEND to be from an organization you know"; "Scammers say there's a PROBLEM or a PRIZE"; "Scammers PRESSURE you to act immediately"; "Scammers tell you to PAY in a specific way".¹³ Its advice: "Resist the pressure to act immediately. Honest businesses will give you time to make a decision." And for a message that seems to come from a company you deal with, it says to look up their number yourself rather than call a number they gave you or the one on your caller ID. Take Five, a UK campaign led by UK Finance, the banking industry's trade body, says it in two sentences: "It's ok to reject, refuse or ignore any requests. Only criminals will try to rush or panic you."¹⁴

Run Harold's caller through the FTC's four. He pretended to be from an organisation Harold knows. He brought a problem. He pressed for now. The FTC's fourth sign is about how you pay; at the door, on this course's reading, the payment is getting inside. Urgency is the tell, and the call-back rule is the answer, at the door exactly as on the phone: end the contact, and reach the organisation by a number you already had.

Two courses before this one teach the con itself in depth, and this lesson leaves them there: Personal Finance Fundamentals lesson 10, "Defending it", on what nearly every scam has to do, and Digital Literacy lesson 11, "Recognising a con, and the ten minutes after you clicked", on doing the thing by your own route.

Where to report a con

Attempted or successful, a con is worth reporting. In the UK, Take Five advises contacting your bank at once if you think you've been scammed, which is practitioner opinion, in this course's labels.¹⁴

• In England, Wales and Northern Ireland, the police's reporting service is Report Fraud, online or on 0300 123 2040.¹⁵
• In Scotland, Report Fraud's own site says to report via 101.¹⁵
• In the US, the FTC takes reports at ReportFraud.ftc.gov, and says it can't resolve individual reports but shares them with law enforcement.¹⁵

A distraction burglary, where someone gets in, is a burglary, and a crime in progress is a 999 or 911 call. Lesson 16, "Afterwards", covers what reporting does and does not get you.

Ruth and Imran, a month later

Now a case with every thread in it. Ruth and Imran came home from work five weeks ago to find the lock on their back door forced and their laptops gone. The back door still has the old lock, pushed back into place. One evening a salesman knocks. He is selling alarm systems, he has heard about their burglary, and he tells them it proves they need an alarm. The price is lower if they sign tonight.

Nothing in this case says they caused the first burglary. The question is only what the evidence says about the choices in front of them now.

Thread 1: the risk is real, and it is now. Pease's review, and the earlier research Tseloni's paper cites, say that a household burgled once is at higher risk, and soon. The salesman is right about that much.

Thread 2: what closes the opportunity. The first burglary showed someone a back door that could be forced. On Pease's mechanism, that's the change to make first, and it's a lock, not an alarm: this course's step, sensible, untested. Then the rest of WIDE: window locks, a timer for an indoor light, an outside sensor light.

Thread 4: the sale. Tonight-only pricing is pressure, the FTC's third sign, and the FTC's answer applies: an honest business will give them time to decide. They can take a leaflet, close the door, and look up the company themselves, or other companies, whenever they like.

So the evidence gives them a clear first step, a real choice about the second, and one thing to refuse, which is the deadline.

What people get wrong

"An alarm is the best protection." In England and Wales survey data, alarms were not among the devices associated with lower risk, and one study found them associated with higher risk. Two US studies, one asking burglars and one comparing neighbourhoods, point the other way. It's contested. Locks and lights in combination are not.

"Alarms attract burglars." The England and Wales finding is an association that its authors call counter-intuitive and warn against over-reading, and the idea that an alarm flags something worth stealing is one of several explanations they give, none tested.

"Burglars attack people at home." Violence is the minority case (7% of US household burglaries in 2003 to 2007), though in England and Wales threat or force is common when someone at home sees the burglar.

"An ID card proves who they are." The Lamplugh advice is to check the card, not trust it: a call to the organisation on a number you already had, from your bill, not the one on the card.

"If I hear someone downstairs I should go and check." The sources read say call the police instead, and don't go in if you think someone has been inside. That's advice for next time, not a judgement on anyone who went to look.

Practice

Connections

Back. Lesson 4's situational prevention grid is what this lesson applied to a house: WIDE's locks are the grid's "increase the effort", its timer light is the "leave signs of occupancy" lesson 4 quoted under extending guardianship, and "Strengthen formal surveillance" is where alarms sit. Lesson 3's repeat victimisation gave the weeks after a burglary their urgency, and its point that fraud is the commonest crime is why the doorstep section is about a con. Lesson 1 used this comparison to introduce the label "tested"; now you've seen the evidence behind it.

Other courses. Personal Finance Fundamentals lesson 10 and Digital Literacy lesson 11 teach the con in full. Digital Literacy lesson 10 has passwords, which this course leaves alone.

Forward. Lesson 6, "Out and away", takes the same grid onto the street, where the evidence is thinner. Lesson 14 is the law if it ever comes to force at home, and lesson 16 covers reporting and support. Home Repair and Maintenance, the next course on the Core, covers how the rest of a house works and which jobs to leave to a trade.

Go deeper

• A. Tseloni and colleagues, "Domestic burglary drop and the security hypothesis", Crime Science, 2017. Open access, and this course read it in full. The section on burglars' ways of entry is the clearest example I know of testing a cause with survey data.
• S. Catalano, Victimization During Household Burglary, Bureau of Justice Statistics, 2010. Read in full here. Short, and it answers "what happens when someone's home?" for the US with numbers rather than anecdotes.
• ONS, "Nature of crime: burglary" tables. A spreadsheet, updated each year. Open the tables on point of entry and on contact with offenders, and look at the number of incidents each percentage rests on, at the bottom of each table: that's lesson 2 in practice.
• Federal Trade Commission, "How To Avoid a Scam". One page, read in full. The four signs are worth knowing by heart, and they work at the door.

Sources

1. A. Tseloni, G. Farrell, R. Thompson, E. Evans and N. Tilley, "Domestic burglary drop and the security hypothesis", Crime Science 6:3, 2017. Read: full text (PMC). Re-read 25 September 2026 for the forced and unforced entry rates, the SPF definition and table of combinations (reported from the team's 2014 analysis of the 2008/09 to 2011/12 surveys), the no-security households, the authors' suggestions on targeting and landlords, and the funding statement.
2. N. Tilley, R. Thompson, G. Farrell, L. Grove and A. Tseloni, "Do burglar alarms increase burglary risk? A counter-intuitive finding and possible explanations", Crime Prevention and Community Safety 17(1), 2015, pp. 1 to 19. Read: abstract and the hypotheses section in full (open access PDF, UCL Discovery), and the Conclusion and Acknowledgements, 25 September 2026.
3. UNC Charlotte, "Through the eyes of a burglar: study provides insights on habits and motivations", 15 May 2013, on K. R. Blevins, J. B. Kuhns and S. Lee, Understanding Decisions to Burglarize from the Offender's Perspective (2012). Read: the university's press release only. The report itself was not read.
4. Rutgers University, "Rutgers study finds alarm systems are valuable crime fighting tool", on S. Lee's 2008 doctoral study of burglar alarms in Newark. Read: the Rutgers press release only. The study itself was not read.
5. Office for National Statistics, "Nature of crime: burglary", tables for the year ending March 2015 to the year ending March 2025 (Crime Survey for England and Wales). Read: tables 2b, 3b, 10b and 11b and the notes, downloaded 25 September 2026. Domestic burglary in a dwelling with entry; household incidents; percentages from each year's table.
6. S. Catalano, Victimization During Household Burglary, Bureau of Justice Statistics Special Report NCJ 227379, 2010, NCVS 2003 to 2007. Read: full text-only version, and Tables 1 and 7 in the PDF, 25 September 2026.
7. Mark Warr, "Fear of Crime in the United States: Avenues for Research and Policy", Criminal Justice 2000, vol. 4, National Institute of Justice. Read in full.
8. Suzy Lamplugh Trust, "Personal safety at home". Read: full page, re-read 25 September 2026.
9. Crown Prosecution Service and National Police Chiefs' Council, Householders and the use of force against intruders, revised 2018. Read: in full.
10. GOV.UK, "Contact the police", read in full; Office for Victims of Crime, "Help for Victims", read in full. Both re-read 25 September 2026.
11. Ken Pease, Repeat Victimisation: Taking Stock, Home Office, 1998. Read: summary and section 2.
12. National Crime Prevention Council, "Home & Neighborhood Safety". Read: full page, re-read 25 September 2026.
13. Federal Trade Commission, "How To Avoid a Scam". Read: full, re-read 25 September 2026.
14. Take Five to Stop Fraud, takefive-stopfraud.org.uk. Read from Internet Archive capture 20260902060025.
15. Report Fraud, home page and contact page, and the FTC's ReportFraud.ftc.gov, each read on its own page in a browser on 25 September 2026.