The Advantages of Probabilistic Survey Questions...May 19, 2016 . 2 A nationally representative, monthly internet-based survey of a rotating panel of ~ 1,300 household heads. Respondents

IT Forum and RCEA Bayesian Workshop Keynote Address

The Advantages of Probabilistic

Survey Questions

The views presented here are those of the author and do not necessarily reflect those of the Federal

Reserve Bank of New York, or the Federal Reserve System

Simon M. Potter

Federal Reserve Bank of New York May 19, 2016

2

A nationally representative, monthly internet-based survey of a rotating

panel of ~ 1,300 household heads.

Respondents on the panel for up to 12 months.

The survey has a core monthly module on expectations about macro-

economic and household level variables.

Inflation, earnings growth, change in home prices and individual spending

items, HH income growth, HH spending growth, taxes, government debt,

credit access, job search.

In addition, special monthly ad-hoc questions and annual surveys on

special topics.

Rotating ad-hocs: Credit access; Labor Market; Spending; Floating

(insurance; ACA; Gas prices).

Annual surveys: Housing; Labor Market; Saving and Assets.

Survey of Consumer Expectations

Slide

3

In your view, what would you say is the percent chance that, over the 12-

month period between May 2018 to May 2019, …

the rate of inflation will be 12% or higher ______

the rate of inflation will be between 8% and 12% ______




the rate of deflation will be between 0% and 2% ______




the rate of deflation will be 12% or more ______

% Total ______

Elicitation of Probabilistic Beliefs

Slide

4

Monthly Survey Public Release

http://www.newyorkfed.org/microeconomics/sce/ Slide

5

𝐼𝑄𝑅1

𝜇1

𝜇𝑛

𝐼𝑄𝑅𝑛

Median

across

respondents:

(𝜇𝐴, 𝐼𝑄𝑅𝐴)

Summarizing Density Forecasts

Slide

6

Expected Three-Year Ahead Inflation Rate

Slide

2.0

2.2

2.4

2.6

2.8

3.0

3.2

3.4

3.6

3.8

4.0

Jun-13 Sep-13 Dec-13 Mar-14 Jun-14 Sep-14 Dec-14 Mar-15 Jun-15 Sep-15 Dec-15 Mar-16

%

Expected Three-Year Ahead Inflation Rate

Source: Survey of Consumer Expectations

7

Expected Three-Year Ahead Inflation Rates, by Education

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5


%

Expected 3-Year Ahead Inflation Rates, by Education

High School or Less Some College BA or Higher


Slide

8

Three-Year Ahead Inflation Uncertainty, by Education

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0


%

3-Year Ahead Inflation Uncertainty, by Education

High School or Less Some College BA or Higher


Slide

9

Three-Year Inflation Expectations Cumulative Probability

Three-Year Inflation Expectations

Median

September

2015

Median

January

2016

Difference

Repeat Respondents

(Sept 2015 and Jan 2016)

2.98 2.45 -0.53 Slide

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

-2 -1 0 1 2 3 4 5 6 7 8

Cu

mu

lati

ve

Pro

ba

bil

ity

3-Year Ahead Expected Inflation Rate (%)

Three-Year Inflation Expecations

September 2015 January 2016

10

Point and Density Forecasts for the Policy Rate

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 1 2 3 4 5

prob. per

pctage

point

Year-End 2017 Target Federal Funds

Rate/Range (December '15 Survey)

PDF

PDF-implied Mean

Modal path

Source: Desk Surveys

0

0.1

0.2

0.3

0.4

0.5

0.6

0 1 2 3 4 5

prob. per

pctage

point


Rate/Range (January '16 Survey)

PDF

PDF-implied Mean

Modal Path


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 1 2 3 4 5

prob. per

pctage

point


Rate/Range (December '15 Survey)

PDF

PDF-implied

Mean

Modal path


0

0.1

0.2

0.3

0.4

0.5

0.6

0 1 2 3 4 5

prob. per

pctage

point


Rate/Range (January '16 Survey)

PDF

PDF-implied

Mean

Modal Path


0

0.1

0.2

0.3

0.4

0.5

0.6

-1 0 1 2 3 4 5

prob. per

pctage

point

Year-End 2017 Target Fed Funds

Rate/Range (March '16 Survey)

PDF

PDF-implied

Mean

Modal Path


0

0.1

0.2

0.3

0.4

0.5

0.6

-1 0 1 2 3 4 5

prob. per

pctage

point

Year-End 2018 Target Fed Funds

Rate/Range (March '16 Survey)

PDF

PDF-implied

Mean

Modal Path


Slide

11

Probability Distribution for Change in Year-end 2014 SOMA Portfolio relative to June 2013 Level

0%

10%

20%

30%

40%

50%

60%

< -707 -707

to

207

-207

to

293

293

to

793

793

to

1293

1293

to

1793

> 1793

prob.

Change in SOMA Portfolio Size ($ billions)

by YE 2014 Conditional on < 7.3%

Unemployment at YE 2013

Conditional: Less than 7.3% Unconditional

Source: Survey of Primary Dealers

0%

10%

20%

30%

40%

50%

60%

< -707 -707

to

207

-207

to

293

293

to

793

793

to

1293

1293

to

1793

> 1793

prob.


at YE 2014 Conditional on 7.3-7.5%


Conditional: Between 7.3 and 7.5% Unconditional

Source: Survey of Primary Dealers

0%

10%

20%

30%

40%

50%

60%

< -707 -707

to

207

-207

to

293

293

to

793

793

to

1293

1293

to

1793

> 1793

prob.


at YE 2014 Conditional on > 7.5%


Conditional: Greater than 7.5% Unconditional

Source: Survey of Primary Dealers Slide

12

Market-Implied Rates, Modal Paths, Risk Premia and PDF-Implied Mean Rates

0

0.5

1

1.5

2

2.5

3

Dec-15 Jan-16 Mar-16

% Market-Implied Federal Funds Rates

YE 2016 YE 2017 YE 2018

Source: Bloomberg

0

0.5

1

1.5

2

2.5

3


% Survey PDF-Implied Mean Federal Funds

Rates

YE 2016 YE 2017 YE 2018


-1.2

-1

-0.8

-0.6

-0.4

-0.2

0


%

Model Estimates of Federal Funds Risk

Premia

YE 2016 YE 2017 YE 2018

Source: Bloomberg, Desk Calculations

0

0.5

1

1.5

2

2.5

3


%

Modal Path Estimates of Federal Funds

Rates

YE 2016 YE 2017 YE 2018

Source: Bloomberg, Desk Calculations

Slide

13

Federal Funds Rate Survey Probability Distributions versus Tilted Distributions

0%

10%

20%

30%

40%

50%

-1 0 1 2 3 4 5

prob. Year-End 2018 FF PDF

(December Survey)

Federal Funds Rates (%)

-1 0 1 2 3 4 5

Year-End 2018 FF PDF

(January Survey)

Federal Funds Rates (%)

Slide

14

KLIC Distributions Across Survey Respondents

0

5

10

15

20

25

30

35

0 0.5 1 1.5

count

Distance from Original Survey by Respondent

(Year-End 2018)

January Survey

December Survey


KLIC distance

Slide

0

5

10

15

20

25

30

35

0 0.5 1 1.5

count

Distance from Original Survey by Respondent

(Year-End 2018)

March Survey

December Survey


KLIC distance

15

Distribution of Probability of Returning to ZLB Across Survey Respondents

0

2

4

6

8

10

12

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%

count

Probability of Returning to ZLB Dispersion Among Respondents

January March


Slide

16

Individual Density Forecasts for Year-end 2017 Federal Funds Rate for March 2016 Survey Respondents

0

5

10

15

20

25

30

-0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

count

Modal Path Estimates for Year-end 2017

(March Survey)


Slide

Documents

The Advantages of Probabilistic Survey Questions...May 19, 2016 . 2 A nationally representative, monthly internet-based survey of a rotating panel of ~ 1,300 household heads. Respondents