Popular in Economics

Central Monetary Authorities

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Central Monetary Authorities

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Economic History

Golden Fetters

UC Berkeley Barry Eichengreen

Economic Inequalities

Visions of Inequality

City University of New York Branko Milanovic

Economics

School Finance Equalization Increases Intergenerational Mobility

Yale University Barbara Biasi

Economics

Olivier Blanchard on U.S. Pandemic-Era Inflation

Peterson Institute for International Economics Olivier Blanchard

The Price of Housing in the United States, 1890-2006

View more insights in Social Sciences - Economics - Economic History

View more figures from University of California, Merced

View the page of Rowena Gray

Read the Study

Next in Economic History

Morts Pour la France: A database of French fatalities of the Great War

Pauline Grosjean | UNSW Sydney

Planting the Seeds of Research

Louis A. Ferleger | Boston University

How do newly constructed housing price indices for 1890-2006 reveal trends in U.S. housing markets and explain uneven price growth over time?

Image courtesy of interviewee. December 3, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video presents an overview of a study on housing prices and rental trends in the United States, highlighting a gap in existing literature regarding systematic data collection. The authors identified that previous research relied on limited data sources and methodologies, often leading to unreliable outcomes.

The study utilizes a new dataset derived from newspaper listings covering 30 mid-sized and large U.S. cities from 1890 to 2006. This video dataset aims to provide annual, city-level data that reflects contemporary market conditions, allowing for a more accurate analysis of long-term trends in housing prices and rents. The authors incorporated data from the Bureau of Labor Statistics and the U.S. Census, but addressed their limitations, such as infrequent surveys and geographic restrictions.

Key findings indicate that real rents have increased by about 20% over the century, challenging previous interpretations that suggested a decline. The study also reveals increased inflation in housing prices dating back to the late 1960s, earlier than previously documented. These findings have implications for the Consumer Price Index (CPI) as housing constitutes a significant portion of consumer budgets.

The authors discuss the broader economic significance of housing prices on living standards and wealth accumulation, noting substantial variations across different regions and cities. They suggest that housing investments, while generally more stable, yield lower returns than stocks.

The study was supported by the National Science Foundation and included collaborations with several researchers, allowing a large-scale data collection effort. The data is made freely available for future research, encouraging further exploration on housing cycles and affordability.

Lastly, the authors advocate for policymakers to consider the long-term trends in housing prices when crafting policies to address housing supply and affordability issues, stressing the importance of this data in understanding economic conditions affecting consumers.

Language:

Export as PDF

Export as PDF
Here are five key points from the video:

1. **Gap in Existing Research**: The authors identified a gap in the literature concerning the long-term trends in house prices and rental prices in the United States. Previous studies utilized limited data sources and methodologies, which constrained their findings.

2. **New Data Sources**: To address this gap, the authors used new newspaper listings to gather more comprehensive data on house prices and rentals across 30 cities from 1890 to 2006. This new dataset offers more reliable and extensive information than previously available sources, particularly from the Bureau of Labor Statistics, which surveyed fewer cities and had limitations in response rates.

3. **Findings on Price Trends**: The study confirmed significant trends in housing prices, including a notable increase during the 1920s followed by a decline during the Great Depression. They also found that real rents increased by about 20% over the 20th century, contradicting previous assertions of declining rents.

4. **Impact on Economic Indicators**: The new dataset resulted in different calculations for rental price inflation (3.5% per year) compared to existing BLS measures (2.6%). This disparity affects the Consumer Price Index (CPI), with potential implications for understanding living standards and the cost of living.

5. **Future Research Directions**: The authors plan to explore new measures of housing affordability by integrating the data on prices and rents with income information. They encourage further academic exploration and exploitation of the dataset to better understand housing market dynamics and their implications for policy and economic modeling.

Export as PDF
Here are some key quotes extracted from the video that highlight the main ideas and findings discussed:

1. **Gap in Existing Literature**:
- "We wrote this study because we identified a gap in the literature... there are existing studies of house prices and rental prices over the long run in the United States."

2. **Limitations of Previous Studies**:
- "But they, you know, people had, researchers had done their best... but we identified that there's other sources that could be used and that would allow for a larger sample size."

3. **Focus on Rental Prices**:
- "We particularly noticed this for rental prices... since 1914, the Bureau of Labor Statistics has done surveys... but they do this for a limited range of cities."

4. **Methodology and Data Sources**:
- "Our main goal was to have annual data city level over this long 12-decade time period."

5. **Historical Housing Project**:
- "The hip hop dataset is now housed at the Philadelphia Federal Reserve Bank... and it's now called the Historical Housing Project."

6. **Utilization of Newspaper Listings**:
- "We identified a new source of data that could help us... using new newspaper listings."

7. **Implications on Economic Measures**:
- "We actually take that seriously in the paper... our measures of living standards today should be revised downwards somewhat due to this result."

8. **Findings on Housing Prices and Rents**:
- "We find that real rents have increased about 20% over this long century... more inflation in the price series compared to the previous literature."

9. **Return on Housing Investment**:
- "We find that rental yields were very important in the early period... but over time, it's really shifted toward the capital gain component being really important."

10. **Broader Applications of the Data**:
- "We're very excited to put this data set out there to have it freely available for other researchers to use... macro economists might be very interested in this data set."

11. **Future Research Directions**:
- "We're excited to do... measures of affordability... understanding housing markets more deeply."

12. **Call to Action for Policymakers**:
- "As an economist... who’s think started thinking about housing... policymakers coming up with new measures to increase housing supply."

These quotes encapsulate the core intentions, findings, and implications of the study, as well as the importance of the new dataset being released for future research and economic policy considerations.

X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:90000

1
00:00:00.065 --> 00:00:02.685
We wrote this study because we identified a,

2
00:00:03.005 --> 00:00:04.125
a gap in the literature.

3
00:00:04.665 --> 00:00:08.565
So, there are existing studies of house prices

4
00:00:08.745 --> 00:00:12.045
and rental prices over the long run in the United States.

5
00:00:12.625 --> 00:00:14.765
Um, but they, you know, people had,

6
00:00:14.965 --> 00:00:16.285
researchers had done their best.

7
00:00:16.285 --> 00:00:18.525
They had cobbled together various different sources,

8
00:00:19.085 --> 00:00:20.365
whatever was available at the time,

9
00:00:20.735 --> 00:00:23.485
using the best methodologies available at the time.

10
00:00:24.025 --> 00:00:27.685
Um, but we identified that, there's, other sources

11
00:00:27.795 --> 00:00:30.645
that could be used and that would allow us

12
00:00:30.665 --> 00:00:33.005
to have a more systematic data collection

13
00:00:33.465 --> 00:00:36.005
and would allow for a larger sample size

14
00:00:36.005 --> 00:00:37.845
that we could be more confident with the results

15
00:00:37.945 --> 00:00:41.445
and with the long run trends that we would, draw out

16
00:00:41.445 --> 00:00:42.645
of those data sets.

17
00:00:43.025 --> 00:00:45.925
We, particularly noticed this for rental prices.

18
00:00:46.545 --> 00:00:51.245
So since 1914, the Bureau of Labor Statistics has,

19
00:00:51.955 --> 00:00:56.205
done surveys, conducted surveys, where they,

20
00:00:56.205 --> 00:00:57.565
asked people about the rents

21
00:00:57.565 --> 00:00:59.525
that they pay on the units that they're in.

22
00:00:59.905 --> 00:01:02.645
Uh, but they do this for a limited range of cities,

23
00:01:02.895 --> 00:01:04.405
which has increased over time,

24
00:01:04.585 --> 00:01:07.885
but makes things difficult to compare across a very long,

25
00:01:08.195 --> 00:01:09.605
over a century time period.

26
00:01:09.655 --> 00:01:12.805
There were also some, some issues in terms of tenants

27
00:01:12.905 --> 00:01:15.125
who moved or didn't respond

28
00:01:15.205 --> 00:01:17.485
to the survey in a particular year,

29
00:01:17.495 --> 00:01:19.485
which made the data less reliable.

30
00:01:20.105 --> 00:01:24.045
Um, we also, wanted to construct a series, both

31
00:01:24.065 --> 00:01:25.205
for sales prices

32
00:01:25.345 --> 00:01:28.605
and rental prices that was very relevant for market rates.

33
00:01:29.145 --> 00:01:31.805
So we wanted to sort of ask the question, if you wanted

34
00:01:31.805 --> 00:01:35.125
to move to the average US city, you know,

35
00:01:35.255 --> 00:01:36.405
could you afford to do

36
00:01:36.405 --> 00:01:38.565
so at these various different time periods?

37
00:01:38.575 --> 00:01:41.285
Maybe you're looking to move to opt opportunity

38
00:01:41.425 --> 00:01:44.205
to access high wage, higher wages and good jobs,

39
00:01:44.865 --> 00:01:46.405
and could you actually do that?

40
00:01:46.405 --> 00:01:48.885
Could you afford, to live in these various places?

41
00:01:49.585 --> 00:01:52.445
And some of these previous data sets were not quite

42
00:01:52.475 --> 00:01:54.885
capturing current market conditions.

43
00:01:54.995 --> 00:01:57.245
They might capture them with a lag,

44
00:01:57.385 --> 00:02:00.805
but they, they weren't capturing exactly what we wanted

45
00:02:00.805 --> 00:02:04.005
to capture with our methodology, the existing data sources.

46
00:02:04.345 --> 00:02:06.045
Um, on the, on the sales side,

47
00:02:06.045 --> 00:02:07.965
on the house price side mostly,

48
00:02:07.965 --> 00:02:10.685
most notably is the case Schiller Vice Index

49
00:02:10.685 --> 00:02:12.725
that people have used quite extensively,

50
00:02:13.305 --> 00:02:16.445
and which again, goes also back to 1890.

51
00:02:16.945 --> 00:02:18.765
Uh, it did a great job with the data

52
00:02:18.765 --> 00:02:22.405
that was available at the time in terms of, computing,

53
00:02:22.465 --> 00:02:24.125
house prices and coming up

54
00:02:24.125 --> 00:02:26.525
with a national index over time.

55
00:02:27.465 --> 00:02:30.645
Uh, but that's the one that we are trying to improve on,

56
00:02:30.665 --> 00:02:33.005
of having annual data at the city level.

57
00:02:33.465 --> 00:02:35.805
Uh, that goes back also to 1890.

58
00:02:36.425 --> 00:02:39.365
The other main source that has been used is the US Census,

59
00:02:39.945 --> 00:02:42.685
which began asking about house prices

60
00:02:42.825 --> 00:02:46.565
and house rents, back in 1930.

61
00:02:47.145 --> 00:02:48.845
Uh, but that's of course only every decade.

62
00:02:49.185 --> 00:02:52.605
So it has limitations in how we can use that dataset.

63
00:02:52.785 --> 00:02:56.045
Uh, if we want to know things at a higher frequency,

64
00:02:56.175 --> 00:02:58.685
we're, we're going to need, need to have a different

65
00:02:58.885 --> 00:02:59.885
Approach. And

66
00:02:59.885 --> 00:03:03.765
we don't have any city level, annual data set

67
00:03:03.825 --> 00:03:06.965
for sales prices until about 1975 in the us.

68
00:03:07.225 --> 00:03:09.965
And so that was sort of the key, difference

69
00:03:09.965 --> 00:03:11.365
between the previous literature

70
00:03:11.365 --> 00:03:13.525
and what we've tried to do with our current paper.

71
00:03:14.385 --> 00:03:16.925
On the rental side, as I said there, the Bureau

72
00:03:16.925 --> 00:03:20.565
of Labor Statistics has put together data since 1914,

73
00:03:21.105 --> 00:03:22.445
but with these limitations

74
00:03:22.625 --> 00:03:26.605
and limited to a smaller range of cities, we are able

75
00:03:26.605 --> 00:03:30.125
to do cover, 30 cities in, in our, in our study,

76
00:03:30.135 --> 00:03:32.725
which has improved upon data availability.

77
00:03:32.985 --> 00:03:36.245
Our main goal was to have annual data city level

78
00:03:36.795 --> 00:03:40.565
over this long 12 decade time period that was,

79
00:03:40.565 --> 00:03:43.565
systematically collected with a large sample size

80
00:03:43.745 --> 00:03:47.325
and where we used the same consistent methodology across

81
00:03:47.395 --> 00:03:49.805
that whole, whole 12 decades.

82
00:03:50.105 --> 00:03:52.605
The hip hop dataset is now housed at the Philadelphia

83
00:03:52.605 --> 00:03:53.845
Federal Reserve Bank.

84
00:03:53.855 --> 00:03:55.525
We're very happy to partner with them

85
00:03:55.545 --> 00:03:57.285
and make this data freely available.

86
00:03:57.465 --> 00:04:00.445
The, the indexes are available for, for anybody to use

87
00:04:00.465 --> 00:04:02.165
or to just view on their site,

88
00:04:02.505 --> 00:04:04.405
and they've made it look very nice, I have to say.

89
00:04:04.905 --> 00:04:07.885
Um, and, and it's now called the Historical Housing Project

90
00:04:08.065 --> 00:04:09.405
on that site, I should point out.

91
00:04:10.025 --> 00:04:13.965
Um, and so yes, working with my co-authors, Ronan Lyons,

92
00:04:13.965 --> 00:04:16.805
who's at Trinity College, Dublin, Allison Scherzer,

93
00:04:16.805 --> 00:04:19.045
who's now at the Phil Philadelphia Federal Reserve,

94
00:04:19.105 --> 00:04:21.085
and David Atos, who's,

95
00:04:21.085 --> 00:04:23.325
finishing his PhD at University of Pittsburgh.

96
00:04:23.865 --> 00:04:27.685
Uh, the four of us collaborated on this huge project,

97
00:04:27.905 --> 00:04:31.325
and we identified a new source of data that could help us

98
00:04:31.945 --> 00:04:34.685
do this big project of systematic data collection

99
00:04:34.745 --> 00:04:36.205
for house prices and rents,

100
00:04:36.495 --> 00:04:39.445
which was using new newspaper listings, which is

101
00:04:39.445 --> 00:04:41.725
what people would've used if they are thinking

102
00:04:41.725 --> 00:04:45.285
of buying a house or renting a property from 1890

103
00:04:45.285 --> 00:04:46.605
through 2006.

104
00:04:47.235 --> 00:04:49.925
Once we get into the two thousands, you know,

105
00:04:49.925 --> 00:04:53.765
the internet comes along and competes for newspapers.

106
00:04:53.825 --> 00:04:57.165
So we haven't, yet, added maybe some,

107
00:04:57.195 --> 00:04:58.405
some internet listings.

108
00:04:58.545 --> 00:05:00.645
We may do that in the future, the near future,

109
00:05:01.025 --> 00:05:03.165
and we will certainly update the dataset when we,

110
00:05:03.165 --> 00:05:04.485
when we add more data.

111
00:05:05.705 --> 00:05:08.725
Uh, and so we were able to look at newspaper listings

112
00:05:08.905 --> 00:05:10.365
for 30 sort of mid-size

113
00:05:10.425 --> 00:05:14.165
and large US cities, 1890 to 2006.

114
00:05:14.785 --> 00:05:18.565
Um, and this project wouldn't have been possible without the

115
00:05:18.565 --> 00:05:19.645
four of us collaborating

116
00:05:19.945 --> 00:05:23.525
to get a National Science Foundation grant, which allowed us

117
00:05:23.525 --> 00:05:27.045
to hire a sort of army of maybe about 50,

118
00:05:27.485 --> 00:05:29.925
research assistants who worked on various aspects

119
00:05:30.065 --> 00:05:31.245
of the data collection.

120
00:05:31.625 --> 00:05:33.805
And it also wouldn't have been possible, maybe, you know,

121
00:05:33.805 --> 00:05:38.245
20 years ago before newspapers had been, scanned in

122
00:05:38.425 --> 00:05:41.045
and available online on a wide scale.

123
00:05:41.345 --> 00:05:43.925
We encourage people in the future if they want to expand

124
00:05:43.925 --> 00:05:47.045
what we've done, you know, now that we have AI technology

125
00:05:47.105 --> 00:05:50.005
as well, allowing us to search through all

126
00:05:50.005 --> 00:05:52.725
of these various texts that are available online.

127
00:05:53.225 --> 00:05:54.405
Uh, you know, some people

128
00:05:54.405 --> 00:05:56.845
who have more computer skills will definitely be able to,

129
00:05:56.940 --> 00:05:59.445
to expand upon our initial approach.

130
00:06:00.145 --> 00:06:04.295
Um, and so what we did was we went through these newspapers.

131
00:06:04.515 --> 00:06:08.215
We looked for listings that told us what the price was

132
00:06:08.315 --> 00:06:09.895
or the rent told us

133
00:06:10.005 --> 00:06:12.295
what the basic location within the city was

134
00:06:12.755 --> 00:06:14.615
and the size of the unit.

135
00:06:15.195 --> 00:06:18.775
And it also, most advertisements also give you a bunch

136
00:06:18.775 --> 00:06:21.615
of amenities and characteristics that,

137
00:06:22.155 --> 00:06:23.615
to sell the, the property.

138
00:06:24.115 --> 00:06:25.535
And so we record all of that.

139
00:06:26.395 --> 00:06:30.455
We then, run an analysis that controls

140
00:06:30.475 --> 00:06:32.935
for these different characteristics, the location

141
00:06:32.995 --> 00:06:35.135
and size being the most important ones,

142
00:06:35.595 --> 00:06:39.135
and allows us just to strip out the sort of annual variation

143
00:06:39.135 --> 00:06:41.855
that gives us the trend in house prices.

144
00:06:42.355 --> 00:06:45.335
And, housing rents the trend that's not driven just

145
00:06:45.335 --> 00:06:47.935
by the fact that, you know, house size changes over time,

146
00:06:48.285 --> 00:06:50.615
certain neighborhoods are more popular,

147
00:06:50.635 --> 00:06:53.295
and have more amenities and so cost more.

148
00:06:53.555 --> 00:06:56.055
So we've kind of stripped out that, component

149
00:06:56.355 --> 00:06:59.015
to construct these trends across cities.

150
00:06:59.595 --> 00:07:02.295
And so, you know, folks could work

151
00:07:02.295 --> 00:07:04.535
with the city level trends over time,

152
00:07:04.835 --> 00:07:08.135
or they could, use our national level in indexes

153
00:07:08.135 --> 00:07:09.975
where we sort of aggregate these things up

154
00:07:10.475 --> 00:07:13.375
to the national level, using these cities.

155
00:07:13.475 --> 00:07:14.775
And because we have mid-size

156
00:07:14.775 --> 00:07:16.735
and large cities, you know, we're representing a,

157
00:07:16.815 --> 00:07:20.455
a large fraction of the US population with this data set.

158
00:07:20.515 --> 00:07:21.735
We do a lot in this project,

159
00:07:22.075 --> 00:07:24.375
and not least the data product that comes out of it,

160
00:07:24.875 --> 00:07:26.055
but also with our findings.

161
00:07:26.555 --> 00:07:31.335
Um, we were very happy to be able to, update,

162
00:07:31.595 --> 00:07:33.935
two stylized facts in the literature

163
00:07:33.935 --> 00:07:36.415
that people had already kind of questioned.

164
00:07:36.915 --> 00:07:40.295
Uh, and so one of those was that, we do find

165
00:07:40.525 --> 00:07:43.895
that there's a ramp up in housing, prices

166
00:07:44.235 --> 00:07:47.455
during the 1920s, the interwar period.

167
00:07:47.715 --> 00:07:51.255
So we see a 1920s boom that's discussed in the literature,

168
00:07:51.635 --> 00:07:54.975
and we also see a bust during the Great Depression when it,

169
00:07:55.035 --> 00:07:56.775
of course, we would expect to see one.

170
00:07:57.035 --> 00:08:00.735
And so the previous data sets that had been put together due

171
00:08:00.735 --> 00:08:03.215
to various data limitations had actually not, they had very

172
00:08:03.805 --> 00:08:06.455
flat series during that time period, which was surprising

173
00:08:06.955 --> 00:08:09.495
to, economic historians of, of that period.

174
00:08:09.715 --> 00:08:13.135
So we were happy that our series seems to match,

175
00:08:13.135 --> 00:08:15.255
some other smaller studies, some case studies

176
00:08:15.555 --> 00:08:18.375
and the historical narrative for that time period.

177
00:08:19.435 --> 00:08:22.535
Um, another fact that we were able to sort of update on was,

178
00:08:22.875 --> 00:08:24.695
thinking about real rans.

179
00:08:24.835 --> 00:08:26.695
So this is rents where we've adjusted

180
00:08:26.795 --> 00:08:28.295
for general price inflation.

181
00:08:29.385 --> 00:08:33.245
Um, in the existing, bureau of Labor Statistics data,

182
00:08:33.425 --> 00:08:36.605
it had looked like real rents were actually declining over

183
00:08:36.665 --> 00:08:38.325
the 20th century,

184
00:08:38.325 --> 00:08:40.445
which I think certainly anecdotally is not sort of what,

185
00:08:40.445 --> 00:08:44.005
what people imagine, although of, of course, we're very,

186
00:08:44.105 --> 00:08:46.685
you know, in this time period of high sales, prices

187
00:08:46.785 --> 00:08:49.485
and rents, it's, it's hard to look back very far and,

188
00:08:49.485 --> 00:08:51.405
and imagine that they were ever any different.

189
00:08:52.025 --> 00:08:54.965
Um, but what we find is actually that real rents

190
00:08:55.245 --> 00:08:58.885
to have increased about 20% over this long century

191
00:08:58.985 --> 00:09:00.445
period that we look at.

192
00:09:00.985 --> 00:09:04.725
We think that conforms more to, again, the other narratives

193
00:09:04.725 --> 00:09:06.005
that we have in the literature

194
00:09:06.345 --> 00:09:09.245
and what economists had had previously, believed.

195
00:09:09.865 --> 00:09:13.635
Um, and more generally, you know, again,

196
00:09:13.635 --> 00:09:14.795
people have a sense that hes,

197
00:09:14.795 --> 00:09:18.835
prices have risen quite a bit in the recent time periods.

198
00:09:19.015 --> 00:09:23.915
Um, and, we do actually find more inflation in the,

199
00:09:24.095 --> 00:09:26.795
price series compared to the previous literature.

200
00:09:26.895 --> 00:09:30.275
We find that sort of the ramp up in sales prices,

201
00:09:30.295 --> 00:09:33.475
didn't just start with the sort of late eighties, nineties.

202
00:09:33.645 --> 00:09:36.835
We're actually seeing that from the late 1960s onwards.

203
00:09:37.335 --> 00:09:39.435
Uh, and so that's an update to the literature

204
00:09:39.435 --> 00:09:42.915
where we're seeing this uptick already, a couple

205
00:09:42.915 --> 00:09:45.675
of decades earlier than previously thought.

206
00:09:46.815 --> 00:09:48.835
Um, and in terms of thinking about, you know,

207
00:09:48.835 --> 00:09:51.875
what are the implications of finding this,

208
00:09:52.275 --> 00:09:55.955
somewhat higher, price inflation in sales prices and,

209
00:09:56.015 --> 00:09:59.115
and rental prices, you know, we, these are things

210
00:09:59.115 --> 00:10:01.395
that factor into the CPI, right?

211
00:10:01.575 --> 00:10:04.475
So, of course, housing is a large portion

212
00:10:04.735 --> 00:10:06.275
of the average consumer's budget,

213
00:10:06.975 --> 00:10:10.915
and, we need good accurate measures of, of

214
00:10:10.915 --> 00:10:12.075
that housing, right?

215
00:10:12.255 --> 00:10:16.275
Uh, and so having a different measure of that price is going

216
00:10:16.275 --> 00:10:19.915
to make some difference in, in the CPI measure.

217
00:10:20.375 --> 00:10:22.395
So we actually take that seriously in the paper.

218
00:10:22.495 --> 00:10:26.795
We do that calculation where we, put in our,

219
00:10:26.795 --> 00:10:30.435
rental price, inflation, which we find

220
00:10:30.435 --> 00:10:32.395
to be 3.5% per year.

221
00:10:32.495 --> 00:10:35.755
The nominal, the index is increasing in nominal terms,

222
00:10:36.415 --> 00:10:39.715
the existing, component of the CPI,

223
00:10:39.865 --> 00:10:41.235
they take the BLS measure,

224
00:10:41.325 --> 00:10:44.035
which increases about 2.6% per year.

225
00:10:44.615 --> 00:10:46.835
And so this actually makes a difference if we

226
00:10:47.075 --> 00:10:48.115
recompute the CPI.

227
00:10:48.335 --> 00:10:52.995
So we have prices increasing by about 3.3% annually.

228
00:10:53.575 --> 00:10:56.515
Uh, the existing measure would be about 3.1%.

229
00:10:57.175 --> 00:10:58.315
And that's important

230
00:10:58.315 --> 00:11:02.275
because, you know, we're basing our, our measures of cost

231
00:11:02.275 --> 00:11:04.915
of living and also our ideas of

232
00:11:04.915 --> 00:11:07.955
how much living standards have increased over the 20th

233
00:11:07.955 --> 00:11:10.635
century on these, these price measures.

234
00:11:10.855 --> 00:11:12.555
And so, you know, it's,

235
00:11:12.555 --> 00:11:14.155
it's not maybe a dramatic change,

236
00:11:14.255 --> 00:11:16.955
but if we're thinking about a 0.2 difference per year,

237
00:11:17.305 --> 00:11:19.155
that is going to add up over a century.

238
00:11:19.455 --> 00:11:22.595
Uh, and so we would argue our, our measures of,

239
00:11:22.655 --> 00:11:25.275
of living standards today should be revised downwards

240
00:11:25.875 --> 00:11:27.915
somewhat due to this, this result.

241
00:11:28.055 --> 00:11:30.595
We look at the richness of our data across cities,

242
00:11:31.375 --> 00:11:35.155
and we actually think about how house prices have tra

243
00:11:35.155 --> 00:11:38.315
changed, across cities over time,

244
00:11:38.415 --> 00:11:39.795
or within each city over time.

245
00:11:40.495 --> 00:11:44.395
Uh, and we've actually computed measures of our,

246
00:11:44.445 --> 00:11:45.875
total return to housing.

247
00:11:46.255 --> 00:11:47.995
And so this has two components.

248
00:11:48.535 --> 00:11:50.995
It has the capital gain component component,

249
00:11:51.005 --> 00:11:53.845
which is really about price appreciation, has,

250
00:11:53.905 --> 00:11:56.925
has your house appreciated in value over time, which

251
00:11:56.925 --> 00:11:59.005
of course, these days, everyone feels like they're an

252
00:11:59.285 --> 00:12:01.005
investor in their house, not just a consumer.

253
00:12:01.345 --> 00:12:04.285
And so we all want to achieve capital gains for our house.

254
00:12:04.945 --> 00:12:07.645
Um, and we also look at the rental yield.

255
00:12:07.745 --> 00:12:10.645
So how have rental prices changed relative to,

256
00:12:11.065 --> 00:12:12.165
house prices?

257
00:12:12.505 --> 00:12:15.605
So thinking about it as again, like if you're the landlord,

258
00:12:15.865 --> 00:12:18.365
you want some return each year in terms

259
00:12:18.365 --> 00:12:20.525
of your rental yield for your property.

260
00:12:21.105 --> 00:12:24.925
And so over this long time period, we find that,

261
00:12:24.925 --> 00:12:28.765
rental yields were very important in the early period,

262
00:12:28.905 --> 00:12:31.085
in terms of making up the total,

263
00:12:31.225 --> 00:12:32.605
the total return on housing.

264
00:12:33.105 --> 00:12:35.765
And then over time, it's really shifted toward the capital

265
00:12:35.875 --> 00:12:37.605
gain component being really important.

266
00:12:38.065 --> 00:12:40.125
Um, and again, that's not super surprising

267
00:12:40.125 --> 00:12:43.565
because house prices have appreciated so much over, over

268
00:12:43.565 --> 00:12:45.765
that time period, especially since World War ii.

269
00:12:46.385 --> 00:12:48.885
Uh, so that's probably what we would expect.

270
00:12:50.005 --> 00:12:51.665
We do find that, you know,

271
00:12:51.665 --> 00:12:53.465
across cities there's huge variation.

272
00:12:54.005 --> 00:12:57.425
And so in the paper, one nice thing we do is

273
00:12:57.425 --> 00:12:58.905
to split it up into regions

274
00:12:58.905 --> 00:12:59.985
because, you know,

275
00:12:59.985 --> 00:13:01.785
it does look like the east coast, the west coast.

276
00:13:01.795 --> 00:13:05.025
Those housing markets tend to, on average, do fairly well

277
00:13:05.125 --> 00:13:07.905
and have had quite a lot of, appreciation.

278
00:13:08.485 --> 00:13:11.785
Uh, places like Detroit have, you know, sort

279
00:13:11.785 --> 00:13:14.225
of not been a good investment from quite

280
00:13:14.225 --> 00:13:15.545
early in our time period.

281
00:13:15.725 --> 00:13:19.905
So even from sort of the late twenties, early thirties,

282
00:13:20.215 --> 00:13:22.305
when Detroit still seemed to be doing okay, you know,

283
00:13:22.305 --> 00:13:23.345
apart from the Great Depression,

284
00:13:23.345 --> 00:13:26.985
which was affecting everyone, we do see that,

285
00:13:27.325 --> 00:13:29.985
the total return there is is not very strong

286
00:13:30.165 --> 00:13:33.105
and negative in some, in some time periods.

287
00:13:33.565 --> 00:13:35.465
We think this is kind of interesting to show

288
00:13:36.005 --> 00:13:38.465
the variation in housing as an investment

289
00:13:39.365 --> 00:13:41.785
across these different cities that it is quite different.

290
00:13:42.365 --> 00:13:47.025
Um, and, housing in terms, if we compare it

291
00:13:47.025 --> 00:13:50.505
to the return to stocks, investing in stocks,

292
00:13:50.615 --> 00:13:52.705
usually stocks looks like a better investment.

293
00:13:52.925 --> 00:13:54.585
That's actually an area where we agree

294
00:13:54.585 --> 00:13:55.825
with the previous literature.

295
00:13:56.165 --> 00:13:58.345
Uh, we, we would be in broad agreement with that.

296
00:13:58.345 --> 00:14:00.025
Even, showing

297
00:14:00.025 --> 00:14:03.145
that we have more has price inflation in our data.

298
00:14:03.765 --> 00:14:05.545
Uh, the housing return tends

299
00:14:05.545 --> 00:14:07.545
to be more stable low compared to stocks.

300
00:14:07.545 --> 00:14:10.785
So that of course may be a reason why people,

301
00:14:10.785 --> 00:14:13.865
definitely like to invest in, in the housing market.

302
00:14:14.075 --> 00:14:16.465
We're very excited to put this data set out there

303
00:14:16.845 --> 00:14:20.185
to have it freely available for other researchers to use it,

304
00:14:20.765 --> 00:14:23.585
in the paper there's even some other areas of,

305
00:14:23.585 --> 00:14:26.065
that we indicate people might work in.

306
00:14:26.445 --> 00:14:29.425
Uh, in particular, we think macro economists might be very

307
00:14:29.425 --> 00:14:30.825
interested in this data set

308
00:14:31.285 --> 00:14:35.265
and in thinking about how, the housing cycle is related

309
00:14:35.285 --> 00:14:36.865
to business cycle more generally.

310
00:14:37.365 --> 00:14:40.185
Um, that's also an area where,

311
00:14:40.185 --> 00:14:43.145
usually macro economists like high frequency data,

312
00:14:43.205 --> 00:14:46.465
and so they may be among the people who would like to go in

313
00:14:46.465 --> 00:14:47.665
and follow our methodology

314
00:14:48.205 --> 00:14:49.985
and do, you know, maybe quarterly data

315
00:14:49.985 --> 00:14:51.145
instead of annual data.

316
00:14:51.625 --> 00:14:53.725
Uh, and we encourage them to do, to do that

317
00:14:53.905 --> 00:14:56.285
and to, to use that for their very testing, their,

318
00:14:56.295 --> 00:14:59.125
their macro and housing finance models.

319
00:14:59.945 --> 00:15:02.885
Um, for our group, Ronan and Allison

320
00:15:03.225 --> 00:15:05.525
and I will be continuing to work in this area,

321
00:15:05.785 --> 00:15:09.445
and a, a follow up paper that we're very excited about is

322
00:15:09.645 --> 00:15:12.445
to think about new measures of housing affordability.

323
00:15:12.785 --> 00:15:14.605
Not just the prices and the rents,

324
00:15:14.985 --> 00:15:17.325
but thinking about affordability as well.

325
00:15:17.865 --> 00:15:20.965
And, measures of affordability are usually very simple.

326
00:15:21.265 --> 00:15:23.005
It, you know, we already have the price side

327
00:15:23.065 --> 00:15:26.205
and the rent side, and now we would be computing the wage

328
00:15:26.275 --> 00:15:28.845
side and thinking about people's incomes.

329
00:15:28.995 --> 00:15:30.405
What do households look like?

330
00:15:30.585 --> 00:15:32.205
How much could the typical household

331
00:15:32.545 --> 00:15:34.445
or the typical wage earner afford

332
00:15:34.985 --> 00:15:38.125
in these different time periods and across, across cities.

333
00:15:38.305 --> 00:15:41.485
So we have some more data work to do to compute those,

334
00:15:41.485 --> 00:15:43.325
measures over this long time period.

335
00:15:44.065 --> 00:15:46.445
Um, and, but we're excited to do that.

336
00:15:46.445 --> 00:15:48.365
We think that will be an important follow up

337
00:15:48.625 --> 00:15:51.005
to understanding housing markets more deeply

338
00:15:51.105 --> 00:15:53.125
and how it really affected the average person

339
00:15:53.675 --> 00:15:55.925
over this long, time period.

340
00:15:57.165 --> 00:16:00.225
Um, but yeah, there is also scope for, you know,

341
00:16:00.595 --> 00:16:03.865
other academics, policy makers as well

342
00:16:03.925 --> 00:16:07.505
to be looking at these data to get more context,

343
00:16:07.775 --> 00:16:09.345
more long run understanding

344
00:16:09.485 --> 00:16:11.745
of what's been happening in housing markets.

345
00:16:12.565 --> 00:16:15.185
And, you know, as an economist, of course,

346
00:16:15.365 --> 00:16:17.585
who who's think started thinking about housing in recent

347
00:16:17.635 --> 00:16:19.785
years, we are always wanting

348
00:16:19.785 --> 00:16:22.225
to see policymakers coming up with new measures

349
00:16:22.225 --> 00:16:23.705
to increase housing supply.

350
00:16:24.285 --> 00:16:25.305
And so, you know,

351
00:16:25.765 --> 00:16:27.785
it doesn't matter whether you live in Sacramento

352
00:16:27.785 --> 00:16:30.025
where I live, which actually has been at the forefront

353
00:16:30.025 --> 00:16:31.705
of trying to increase housing supply

354
00:16:31.725 --> 00:16:33.305
and has actually seen some results

355
00:16:33.305 --> 00:16:36.025
with rents falling over the last year or two.

356
00:16:36.165 --> 00:16:37.585
I'm sure London has a big housing

357
00:16:37.985 --> 00:16:39.105
shortage and a big housing crunch.

358
00:16:39.645 --> 00:16:42.465
Uh, you know, I'm from Dublin Ronan as well,

359
00:16:42.815 --> 00:16:45.945
looking at these housing markets,

360
00:16:46.205 --> 00:16:48.945
we also have our own shortage over there as well.

361
00:16:49.085 --> 00:16:51.785
So we really hope that policymakers will pay attention

362
00:16:52.045 --> 00:16:54.385
to seeing this long run,

363
00:16:54.385 --> 00:16:56.505
inflation in house prices in particular.

364
00:16:57.005 --> 00:16:59.105
And, think about, okay,

365
00:16:59.535 --> 00:17:01.625
this is a really important issue in people's lives.

366
00:17:01.805 --> 00:17:04.385
It makes up a large portion of Peoples's budget.

367
00:17:04.775 --> 00:17:07.145
It's also a large fraction of household wealth is

368
00:17:07.985 --> 00:17:09.225
captured in their, in their housing.

369
00:17:09.765 --> 00:17:11.545
And so it's a really important area

370
00:17:11.545 --> 00:17:14.505
where we need policymakers to be paying attention,

371
00:17:14.525 --> 00:17:15.865
to the latest information.

Bridging Gaps in Housing Price Research

00:00:00 - We wrote this study because we identified a, a gap in the literature. So, there are existing studies of house prices and rental prices over the long run in the United States. but they, you know, people had, researchers had done their best. They had cobbled together various different sources, whatever was available at the time, using the best methodologies available at the time. but we identified that, there's, other sources that could be used and that would allow us

00:00:33 - and would allow for a larger sample size that we could be more confident with the results and with the long run trends that we would, draw out of those data sets. We, particularly noticed this for rental prices. So since 1914, the Bureau of Labor Statistics has, done surveys, conducted surveys, where they, asked people about the rents that they pay on the units that they're in. but they do this for a limited range of cities,

00:01:04 - but makes things difficult to compare across a very long, over a century time period. There were also some, some issues in terms of tenants who moved or didn't respond to the survey in a particular year, which made the data less reliable. we also, wanted to construct a series, both for sales prices and rental prices that was very relevant for market rates. So we wanted to sort of ask the question, if you wanted to move to the average US city, you know,

Challenges and Opportunities in Housing Market Data Analysis

00:01:36 - so at these various different time periods? Maybe you're looking to move to opt opportunity to access high wage, higher wages and good jobs, and could you actually do that? Could you afford, to live in these various places? And some of these previous data sets were not quite capturing current market conditions. They might capture them with a lag, but they, they weren't capturing exactly what we wanted to capture with our methodology, the existing data sources. on the, on the sales side,

00:02:07 - most notably is the case Schiller Vice Index that people have used quite extensively, and which again, goes also back to 1890. it did a great job with the data that was available at the time in terms of, computing, house prices and coming up with a national index over time. but that's the one that we are trying to improve on, of having annual data at the city level. that goes back also to 1890.

00:02:39 - which began asking about house prices and house rents, back in 1930. but that's of course only every decade. So it has limitations in how we can use that dataset. if we want to know things at a higher frequency, we're, we're going to need, need to have a different Approach. And we don't have any city level, annual data set for sales prices until about 1975 in the us.

Enhancing Rental Data Through Historical Newspaper Listings

00:03:09 - between the previous literature and what we've tried to do with our current paper. On the rental side, as I said there, the Bureau of Labor Statistics has put together data since 1914, but with these limitations and limited to a smaller range of cities, we are able to do cover, 30 cities in, in our, in our study, which has improved upon data availability. Our main goal was to have annual data city level over this long 12 decade time period that was,

00:03:43 - and where we used the same consistent methodology across that whole, whole 12 decades. The hip hop dataset is now housed at the Philadelphia Federal Reserve Bank. We're very happy to partner with them and make this data freely available. The, the indexes are available for, for anybody to use or to just view on their site, and they've made it look very nice, I have to say. and, and it's now called the Historical Housing Project on that site, I should point out. and so yes, working with my co-authors, Ronan Lyons,

00:04:16 - who's now at the Phil Philadelphia Federal Reserve, and David Atos, who's, finishing his PhD at University of Pittsburgh. the four of us collaborated on this huge project, and we identified a new source of data that could help us do this big project of systematic data collection for house prices and rents, which was using new newspaper listings, which is what people would've used if they are thinking of buying a house or renting a property from 1890

Evolution of Housing Data: From Newspapers to Digital Era

00:04:47 - Once we get into the two thousands, you know, the internet comes along and competes for newspapers. So we haven't, yet, added maybe some, some internet listings. We may do that in the future, the near future, and we will certainly update the dataset when we, when we add more data. and so we were able to look at newspaper listings for 30 sort of mid-size and large US cities, 1890 to 2006. and this project wouldn't have been possible without the

00:05:19 - to get a National Science Foundation grant, which allowed us to hire a sort of army of maybe about 50, research assistants who worked on various aspects of the data collection. And it also wouldn't have been possible, maybe, you know, 20 years ago before newspapers had been, scanned in and available online on a wide scale. We encourage people in the future if they want to expand what we've done, you know, now that we have AI technology as well, allowing us to search through all

00:05:53 - you know, some people who have more computer skills will definitely be able to, to expand upon our initial approach. and so what we did was we went through these newspapers. We looked for listings that told us what the price was or the rent told us what the basic location within the city was and the size of the unit. And it also, most advertisements also give you a bunch of amenities and characteristics that,

Analyzing Trends in Housing Prices and Rents

00:06:24 - And so we record all of that. We then, run an analysis that controls for these different characteristics, the location and size being the most important ones, and allows us just to strip out the sort of annual variation that gives us the trend in house prices. And, housing rents the trend that's not driven just by the fact that, you know, house size changes over time, certain neighborhoods are more popular, and have more amenities and so cost more.

00:06:56 - to construct these trends across cities. And so, you know, folks could work with the city level trends over time, or they could, use our national level in indexes where we sort of aggregate these things up to the national level, using these cities. And because we have mid-size and large cities, you know, we're representing a, a large fraction of the US population with this data set. We do a lot in this project, and not least the data product that comes out of it,

00:07:26 - we were very happy to be able to, update, two stylized facts in the literature that people had already kind of questioned. and so one of those was that, we do find that there's a ramp up in housing, prices during the 1920s, the interwar period. So we see a 1920s boom that's discussed in the literature, and we also see a bust during the Great Depression when it,

Revising Historical Rental Trends and Price Inflation

00:07:57 - And so the previous data sets that had been put together due to various data limitations had actually not, they had very flat series during that time period, which was surprising to, economic historians of, of that period. So we were happy that our series seems to match, some other smaller studies, some case studies and the historical narrative for that time period. another fact that we were able to sort of update on was, thinking about real rans. So this is rents where we've adjusted

00:08:29 - in the existing, bureau of Labor Statistics data, it had looked like real rents were actually declining over the 20th century, which I think certainly anecdotally is not sort of what, what people imagine, although of, of course, we're very, you know, in this time period of high sales, prices and rents, it's, it's hard to look back very far and, and imagine that they were ever any different. but what we find is actually that real rents to have increased about 20% over this long century

00:09:00 - We think that conforms more to, again, the other narratives that we have in the literature and what economists had had previously, believed. and more generally, you know, again, people have a sense that hes, prices have risen quite a bit in the recent time periods. and, we do actually find more inflation in the, price series compared to the previous literature. We find that sort of the ramp up in sales prices,

Impact of Housing Price Inflation on CPI Measurements

00:09:33 - We're actually seeing that from the late 1960s onwards. and so that's an update to the literature where we're seeing this uptick already, a couple of decades earlier than previously thought. and in terms of thinking about, you know, what are the implications of finding this, somewhat higher, price inflation in sales prices and, and rental prices, you know, we, these are things that factor into the CPI, right?

00:10:04 - of the average consumer's budget, and, we need good accurate measures of, of that housing, right? and so having a different measure of that price is going to make some difference in, in the CPI measure. So we actually take that seriously in the paper. We do that calculation where we, put in our, rental price, inflation, which we find to be 3.5% per year.

00:10:36 - the existing, component of the CPI, they take the BLS measure, which increases about 2.6% per year. And so this actually makes a difference if we recompute the CPI. So we have prices increasing by about 3.3% annually. the existing measure would be about 3.1%. And that's important because, you know, we're basing our, our measures of cost of living and also our ideas of

Analyzing Housing Returns: Capital Gains vs. Rental Yields Over Time

00:11:07 - century on these, these price measures. And so, you know, it's, it's not maybe a dramatic change, but if we're thinking about a 0.2 difference per year, that is going to add up over a century. and so we would argue our, our measures of, of living standards today should be revised downwards somewhat due to this, this result. We look at the richness of our data across cities, and we actually think about how house prices have tra

00:11:38 - or within each city over time. and we've actually computed measures of our, total return to housing. And so this has two components. It has the capital gain component component, which is really about price appreciation, has, has your house appreciated in value over time, which of course, these days, everyone feels like they're an investor in their house, not just a consumer. And so we all want to achieve capital gains for our house. and we also look at the rental yield.

00:12:11 - house prices? So thinking about it as again, like if you're the landlord, you want some return each year in terms of your rental yield for your property. And so over this long time period, we find that, rental yields were very important in the early period, in terms of making up the total, the total return on housing. And then over time, it's really shifted toward the capital gain component being really important.

Regional Variations in Housing Market Returns

00:12:40 - because house prices have appreciated so much over, over that time period, especially since World War ii. so that's probably what we would expect. We do find that, you know, across cities there's huge variation. And so in the paper, one nice thing we do is to split it up into regions because, you know, it does look like the east coast, the west coast. Those housing markets tend to, on average, do fairly well and have had quite a lot of, appreciation.

00:13:11 - of not been a good investment from quite early in our time period. So even from sort of the late twenties, early thirties, when Detroit still seemed to be doing okay, you know, apart from the Great Depression, which was affecting everyone, we do see that, the total return there is is not very strong and negative in some, in some time periods. We think this is kind of interesting to show the variation in housing as an investment

00:13:42 - and, housing in terms, if we compare it to the return to stocks, investing in stocks, usually stocks looks like a better investment. That's actually an area where we agree with the previous literature. we, we would be in broad agreement with that. Even, showing that we have more has price inflation in our data. the housing return tends to be more stable low compared to stocks. So that of course may be a reason why people,

Advancing Housing Economics Research Through Accessible Data

00:14:14 - We're very excited to put this data set out there to have it freely available for other researchers to use it, in the paper there's even some other areas of, that we indicate people might work in. in particular, we think macro economists might be very interested in this data set and in thinking about how, the housing cycle is related to business cycle more generally. that's also an area where, usually macro economists like high frequency data,

00:14:46 - and follow our methodology and do, you know, maybe quarterly data instead of annual data. and we encourage them to do, to do that and to, to use that for their very testing, their, their macro and housing finance models. for our group, Ronan and Allison and I will be continuing to work in this area, and a, a follow up paper that we're very excited about is to think about new measures of housing affordability. Not just the prices and the rents,

00:15:17 - And, measures of affordability are usually very simple. It, you know, we already have the price side and the rent side, and now we would be computing the wage side and thinking about people's incomes. What do households look like? How much could the typical household or the typical wage earner afford in these different time periods and across, across cities. So we have some more data work to do to compute those, measures over this long time period. and, but we're excited to do that.

Understanding Long-Term Housing Markets and Policy Implications

00:15:48 - to understanding housing markets more deeply and how it really affected the average person over this long, time period. but yeah, there is also scope for, you know, other academics, policy makers as well to be looking at these data to get more context, more long run understanding of what's been happening in housing markets. And, you know, as an economist, of course, who who's think started thinking about housing in recent

00:16:19 - to see policymakers coming up with new measures to increase housing supply. And so, you know, it doesn't matter whether you live in Sacramento where I live, which actually has been at the forefront of trying to increase housing supply and has actually seen some results with rents falling over the last year or two. I'm sure London has a big housing shortage and a big housing crunch. you know, I'm from Dublin Ronan as well, looking at these housing markets, we also have our own shortage over there as well.

00:17:14 - to seeing this long run, inflation in house prices in particular. And, think about, okay, this is a really important issue in people's lives. It makes up a large portion of Peoples's budget. It's also a large fraction of household wealth is captured in their, in their housing. And so it's a really important area where we need policymakers to be paying attention, to the latest information.

Faculti AI can make mistakes

Research

Policy

Analysis

Data

Originals

Explainers

Podcasts