Does gentrification reduce crime?

David Autor, Christopher Palmer, Parag Pathak 16 November 2017



Separating cause from effect is notoriously difficult when it comes to gentrification. Do coffee shops and yoga studios enter neighbourhoods because the new residents demand them, or do new residents turn up in response to these upscale amenities?

One of the most important neighbourhood amenities is public safety. Rising urban crime in the 1980s accelerated suburbanisation (Cullen and Levitt 1999), and then falling urban crime in the 2000s contributed to gentrification (Ellen et al. 2017). These trends motivate asking: if changes in local crime affect neighbourhood turnover, does a change in neighbourhood composition affect crime? Theoretically, gentrification could increase or decrease crime. On one hand, crime may increase as an influx of relatively affluent residents could make for more attractive targets. Neighbourhood turnover might also increase crime by reducing social cohesion (e.g. Wilson 1987, Sampson et al. 1997). Several other forces could cause gentrification to reduce crime, however. The ‘broken windows’ theory holds that mitigating overt signs of decay (e.g. fixing broken windows) – as generally occurs during gentrification – may deter criminal activity (Wilson and Kelling 1982). Affluent residents may indirectly cause the local property tax base to grow, shift municipal priorities towards increased crime-fighting, and invest in private security measures that deter crime, such as alarm systems (Farrell et al. 2011). Finally, gentrification may reduce crime because rising rents may force local criminals to relocate, while increased local economic activity may induce incumbent criminals to partake in legitimate employment.

Mirroring these ambiguous theoretical predictions, and owing to the difficulty of isolating causality in such settings, the existing empirical evidence on the relationship between crime and neighbourhood change is mixed (McDonald 1986, Taylor and Covington 1988, Covington and Taylor 1989, Lee 2010, Van Wilsem et al. 2006, Papachristos et al. 2011, Aliprantis and Hartley 2015).

In a recent paper, we study whether and by how much gentrification affects crime (Autor et al. 2017). Our approach analyses an unintentional policy experiment in neighbourhood change that occurred when a stringent rent control regime that kept rental prices well below market for approximately one-third of Cambridge, MA residential rental properties rapidly and unexpectedly ended. Before 1995, much of Cambridge’s rental housing was rent controlled, with rents capped at unit-specific maximums. A state-wide ballot initiative narrowly ended rent control in Massachusetts on 1 January 1995. This initiative allowed rents to rise to market levels and set off a wave of neighborhood change in Cambridge. In an earlier paper, we show that in the years immediately following deregulation, neighbourhoods with the highest rent-control shares experienced a 20% increase in residential turnover and a substantial increase in property values compared to neighbourhoods with less rent control (Autor et al. 2014).

This sudden change in the policy regime allows us to examine how neighbourhood change affects crime (as opposed to the other way around), holding constant city-wide factors affecting overall criminal activity. We assemble geocoded detailed criminal incident-level data from 1992-2005 from the Cambridge Police Department archives. We then track the evolution of criminal activity through tight geographic comparisons across narrow slices of the city to contrast the experience of more versus less rent control-intensive neighbourhoods in the post-control era. Figure 1 summarises our main finding. In the years immediately following the end of rent control, crime fell significantly more in neighbourhoods that were heavily rent controlled. By 1996, overall crime in gentrifying neighbourhoods (blocks with one standard deviation higher rent control exposure initially) fell by 16% more than in neighbourhoods relatively less exposed to rent control. Table 1 shows how effects differ across categories of reported crimes. Property crimes account for the largest decrease in the number of reported criminal incidents, although violent crimes account for the bulk of the value in averted crimes.

Figure 1 Effect of exposure to gentrification on reported crime by year

Notes: Figure plots event study coefficients for the effect of a standard deviation increase in Rent Control Intensity (RCI) on total crime per area, 1992–2005. RCI measures the intensity with which a block was exposed to rent controlled properties prior to 1995. Plotted estimates are coefficients on RCI x Year variables from an event-study regression in which the dependent variable is total crime per 1,000m2. The specification includes year and block fixed effects. The year 1994 is the omitted RCI x Year category. Robust standard errors are clustered at the block level. The vertical line 1994 indicates the year preceding rent decontrol.

Table 1 Effect of gentrification on reported crimes by crime category

Notes: Table reports estimated decrease in each category of crime from 1995-2005 for a one standard deviation higher than average exposure to rent deregulation in Cambridge, Massachusetts, after allowing for neighborhood block fixed effects, year fixed effects, and linear census tract trends. The average cost of crime is a frequency-weighted average of Cohen and Piquero’s (2009) estimates of total direct cost of various types of crime and monetizes effect of crime on the victim, criminal justice costs, and offender productivity. The value of averted crime is an estimate of the present value of the annual reduction in Cambridge crime for each category using a 5% discount rate.

Might crime have fallen in these neighbourhoods anyway, simply due to the urban renaissance occurring throughout the US in the 1990s? Two key facts convince us otherwise. First, in the three years prior to deregulation, crime did not fall differentially in those neighbourhoods that were about to experience significant turnover after deregulation (Figure 1). Second, the effect of deregulation persists even while accounting for other geographic trends in crime that were occurring prior to rent decontrol, such as improvements in public safety in neighbourhoods with high initial levels of crime.

Gentrification lowered Cambridge, MA crime in aggregate, a fact we confirm using FBI Uniform Crime Statistics to compare Cambridge’s experience to similarly sized cities. Our earlier work shows that deregulation caused a $2 billion appreciation in the market value of Cambridge residential property in the decade following rent decontrol. How much of that increase in value was caused by the contemporaneous decrease in crime? We use estimates of the monetary and non-monetary cost of crime frequently used in the criminology literature to quantify the economic value of reductions in crime (Cohen and Piquero 2009). By this measure, an average crime classified as a public disturbance costs roughly $5,000 in terms of its disamenity to residents, whereas the average violent crime has an associated disamenity value of over $60,000 (see Table 1). Adding up the crime reductions and valuing them using the estimates of Cohen and Piquero (2009), we find that the annual reduction in Cambridge crime was worth roughly $10-15 million per year to city residents. Since the improvement in public safety appears stable and durable (see Figure 1), we can estimate how much this shift in Cambridge crime could have increased property values. Using a 5% discount rate, we estimate the value of the decontrol-induced public safety improvements at approximately $200 million, or 10-15% of the total increase in property values generated by rent decontrol.

Gentrification raises property values, but is likely to produce both winners and losers. Cambridge’s experience was no exception. We find that rent de-control caused overall crime to fall by 16% - approximately 1,200 reported crimes annually – with the majority of the effect accruing through reduced property crime. This overall reduction in crime created $200 million in economic value for the city of Cambridge. These estimates provide rigorous evidence that improvements in public safety are an important part of the gentrification process. At the same time, they suggest that the process need not benefit all residents. The neighbourhoods that had the most rent control prior to 1995 saw the most turnover after de-control, suggesting that incumbent renters of controlled units were priced out of existing units. It’s likely that many of these residents left Cambridge altogether and moved to less-expensive municipalities. While the ensuing influx of new renters probably increased city tax collections and landlords’ bottom-lines, those displaced by gentrification did not necessarily benefit from improvements in neighbourhood safety and other amenities.


Aliprantis, D and D Hartley (2015), “Blowing it up and knocking it down: The local and city-wide effects of demolishing high concentration public housing on crime”, Journal of Urban Economics 88: 67–81.

Autor, D H, C J Palmer, and P A Pathak (2014), “Housing Market Spillovers: Evidence for the End of Rent Control in Cambridge, Massachusetts”, Journal of Political Economy 122(3): 661–717.

Autor, D H, C J Palmer, and P A Pathak (2017), “Gentrification and the Amenity Value of Crime Reductions: Evidence from Rent Deregulation”, NBER Working Paper No. 23914.

Cohen, M A and A R Piquero (2009), “New Evidence on the Monetary Value of Saving a High-Risk Youth”, Journal of Quantitative Criminology 25(1): 25–49.

Covington, J and R B Taylor (1989), “Gentrification and Crime: Robbery and Larceny Changes in Appreciating Baltimore Neighborhoods during the 1970s”, Urban Affairs Review 25(1): 142–172.

Cullen, J B and S D Levitt (1999), “Crime, urban flight, and the consequences for cities”, Review of Economics and Statistics 81(2): 159–169.

Ellen, I G, K Horn, and D Reed (2017), “Has Falling Crime Invited Gentrification?”, Census Bureau Center for Economic Studies Paper No. CES-WP-17-27.

Farrell, G, N Tilley, A Tseloni, and J Mailley (2011), “The crime drop and the security hypothesis”, Journal of Research in Crime and Delinquency 48(2): 147–175.

McDonald, S C (1986), “Does gentrification affect crime rates,” Crime & Justice 8: 163–201.

Papachristos, A V, C M Smith, M L Scherer, and M A Fugiero (2011), “More Coffee, Less Crime? The Relationship between Gentrification and Neighborhood Crime Rates in Chicago, 1991 to 2005”, City & Community 10(3): 215–240.

Sampson, Robert J, Stephen W. Raudenbush, and Felton Earls, “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy”, Science 277(5328): 918–924.

Taylor, R B and J Covington (1988), “Neighborhood Changes in Ecology and Violence”, Criminology 26(4): 553–590.

Van Wilsem, J, K Wittebrood and N D De Graaf (2006), “Socioeconomic dynamics of neighborhoods and the risk of crime victimization: A multilevel study of improving, declining, and stable areas in the Netherlands.” Social Problems 53(2): 226-247.

Wilson, J Q and G L Kelling (1982), “Broken Windows,” The Atlantic Monthly, March: 29–38.

Wilson, W J (1987), The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy, University of Chicago Press.



Topics:  Frontiers of economic research

Tags:  gentrification, crime, house prices, broken windows

Ford Professor and Associate Head, MIT Department of Economics

Assistant Professor, Massachusetts Institute of Technology Sloan School of Management

Jane Berkowitz Carlton and Dennis William Carlton Professor of Microeconomics, MIT