Exercise in Input-Output Analysis A Manual Compilation of

Total Economic Impacts

URP 102:290
Dave Swenson. Iowa State University. Spring 2010.
This exercise is designed to teach you the steps needed to take an industry X industry
matrix, either one of your own construction or one that has been made, and compile the
total economic impacts of particular sectors of the economy. This method is designed to
help you understand the sectoral or industrial location of all final demand: exports,
governments, or households, mainly, while still accounting for all endogenous
transactions (industry X industry X households).
Step 1. The first step is to get an industry by industry matrix (or SAM matrix) of your
study region. Ours comes directly from the IMPLAN model. In this example we are
using 2 digit NAICS summaries. That means that industries are broadly aggregated. The
matrix is then made into a table for use in the spreadsheet using the pivot table utility.
Transactions Table or
Social Accounts Matrix
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation
Utilities
Information
Fin/Insurance
Real Estate
Prof Svcs
Other Svcs
SL Gov
Fed Gov
Other
Emp Comp
Proprietors
Other Property
Ind Bus Tax
Households
Gov
Frgn Trade
Dom Trade
Total

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9
10
11
12
13.1
13.2
14.1
14.2
15
16
17
5000
6000
7000
8000
10000
11000
25000
28000
Grand Total

1
1466.085
367.933
1.651
43.211
428.474
9.479
221.006
147.578
19.970
201.981
519.772
88.222
100.268
116.703
0.431
0.000
480.955
1732.253
1469.457
317.176
0.000
27.728
142.095
3800.382
11702.811

4
0.041
132.744
0.006
19.872
98.130
36.874
117.032
41.150
96.204
94.634
106.577
114.066
243.300
24.225
0.108
0.000
2992.570
198.189
1250.592
1325.263
0.000
16.127
33.328
1023.948
7964.980

6
0.017
4.692
1.143
0.023
9.049
1.700
5.925
8.114
0.717
3.809
4.893
2.254
8.946
1.952
0.002
0.000
98.712
18.696
82.994
10.138
0.000
0.135
4.926
68.205
337.043

7
7.425
281.932
10.939
14.541
713.912
700.421
179.851
36.503
46.939
155.648
128.994
337.481
192.794
34.151
0.458
0.000
2872.483
781.575
261.561
47.056
0.000
2.749
263.753
2851.908
9923.074

8
4549.344
3071.869
5.766
145.566
4615.017
125.819
1546.692
696.404
412.835
694.227
415.780
635.015
1349.595
327.783
0.469
0.000
10925.490
387.792
5079.912
461.176
29.195
76.205
2486.517
20223.120
58261.587

9
0.023
25.444
0.009
32.356
79.791
61.609
115.206
89.574
143.548
110.007
212.588
103.123
281.800
47.410
0.212
0.000
4234.028
434.878
1553.302
1502.765
0.000
13.051
21.337
1200.690
10262.750

185810.172

This table is a statement of all productivity in a region to include the demand for all local
requirements. That means that the total requirements have been adjusted by the sectoral
Regional Purchasing Coefficients (RPCs). In and of itself, this is a very useful table for
analysis. The column sectors list the amount of purchases that are made from the row
sectors Agriculture, for example, purchased $1.466 billion in inputs from other

agriculture and $428.5 million in inputs from manufacturing. This table can be used to
identify inter-industrial dependencies and the extent to which specific Iowa industrial
sectors supply goods and services to other sectors in the Iowa economy or elsewhere.
Step 2. Next we copy the table directly below the original SAM and convert the values to
a table of local coefficients. First, we edit the matrix to only include by row and column
detail through households. That is all of the information that we need to work with for
the remainder of this exercise. Next, using the cell references in the first table, we divide
all cell values by the column totals in table 1. This gives us our local coefficients. (Note,
you should use an =if(columntotal >0, argument,0) check in case the sector total is zero
and to imbed zeros in blank cells). This table is analyzable, but we really dont do
anything else with it. Technically, we call this the A Matrix.
Direct (Local)
Coefficients or A Matrix
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation
Utilities
Information
Fin/Insurance
Real Estate
Prof Svcs
Other Svcs
SL Gov
Fed Gov
Other
Emp Comp
Proprietors
Other Property
Ind Bus Tax
Households

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9
10
11
12
13.1
13.2
14.1
14.2
15
16
17
5000
6000
7000
8000
10000

1
0.1252763
0.0314397
0.0001411
0.0036923
0.0366129
0.0008100
0.0188849
0.0126105
0.0017064
0.0172592
0.0444143
0.0075386
0.0085678
0.0099722
0.0000368
0.0000000
0.0410974
0.1480203
0.1255645
0.0271026
0.0000000

4
0.0000052
0.0166660
0.0000007
0.0024950
0.0123202
0.0046295
0.0146933
0.0051663
0.0120784
0.0118813
0.0133807
0.0143210
0.0305462
0.0030414
0.0000136
0.0000000
0.3757160
0.0248825
0.1570113
0.1663862
0.0000000

6
0.0000497
0.0139225
0.0033922
0.0000693
0.0268495
0.0050427
0.0175806
0.0240737
0.0021276
0.0113001
0.0145186
0.0066875
0.0265433
0.0057909
0.0000071
0.0000000
0.2928752
0.0554703
0.2462428
0.0300782
0.0000000

7
0.0007483
0.0284117
0.0011023
0.0014654
0.0719446
0.0705851
0.0181246
0.0036786
0.0047303
0.0156855
0.0129994
0.0340097
0.0194288
0.0034416
0.0000462
0.0000000
0.2894751
0.0787634
0.0263588
0.0047421
0.0000000

8
0.0780848
0.0527255
0.0000990
0.0024985
0.0792120
0.0021596
0.0265474
0.0119531
0.0070859
0.0119157
0.0071364
0.0108994
0.0231644
0.0056261
0.0000081
0.0000000
0.1875248
0.0066560
0.0871914
0.0079156
0.0005011

9
0.0000022
0.0024792
0.0000009
0.0031527
0.0077748
0.0060032
0.0112257
0.0087281
0.0139873
0.0107191
0.0207145
0.0100483
0.0274585
0.0046196
0.0000206
0.0000000
0.4125627
0.0423744
0.1513534
0.1464291
0.0000000

The next series of steps involve manipulating the data to arrive at a final or closed model.
Step 3. Compile an I Matrix (identity matrix) the same size as the second table (rows and
columns through households. Put zeros in all of the cells and 1s in the diagonals.
Identity (I) Matrix
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9

1
1.0
0.0
0.0
0.0
0.0
0.0

4
0.0
1.0
0.0
0.0
0.0
0.0

6
0.0
0.0
1.0
0.0
0.0
0.0

7
0.0
0.0
0.0
1.0
0.0
0.0

8
0.0
0.0
0.0
0.0
1.0
0.0

9
0.0
0.0
0.0
0.0
0.0
1.0

Step 4. Now we create the I-A table. You are simply subtracting the A-Matrix (Step 2)
from the I-Matrix (Step 3).
I - A Matrix
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9
10

1
0.87472
-0.03144
-0.00014
-0.00369
-0.03661
-0.00081
-0.01888

4
-0.00001
0.98333
0.00000
-0.00249
-0.01232
-0.00463
-0.01469

6
-0.00005
-0.01392
0.99661
-0.00007
-0.02685
-0.00504
-0.01758

7
-0.00075
-0.02841
-0.00110
0.99853
-0.07194
-0.07059
-0.01812

8
-0.07808
-0.05273
-0.00010
-0.00250
0.92079
-0.00216
-0.02655

9
0.00000
-0.00248
0.00000
-0.00315
-0.00777
0.99400
-0.01123

Step 5. Heres where it gets a little tricky. We need to invert the I-A matrix to create the
new total requirements estimate this is also called the SAM inverse, or more famously,
the Leontif inverse. This gives us, technically, the predictive multiplier model. To do the
inversion, we need to copy down a dummy table where the inverse is going to go. Then
we need to mark (block) the whole area to which we are going to write the inverse. Then
we use the =MINVERSE(,,,) command on the matrix above. Look in the first cell of the
example to see exactly how the equation is written to include the entire matrix range to be
inverted. Next, and this is important, to execute the MINVERSE command, you must
hit control-shift-enter. Once done, the inverted values are copied and you have a total
multiplier table.
-1

(I - A) Matrix or the (IA) Inverse

State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation
Utilities
Information
Fin/Insurance
Real Estate
Prof Svcs
Other Svcs
SL Gov
Fed Gov
Other
Emp Comp
Proprietors
Other Property
Ind Bus Tax
Households

Total Multipliers

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9
10
11
12
13.1
13.2
14.1
14.2
15
16
17
5000
6000
7000
8000
10000

1
1.15059
0.05605
0.00018
0.00746
0.07312
0.04252
0.03582
0.02397
0.01215
0.05583
0.06952
0.01956
0.10885
0.01885
0.00014
0.03696
0.21096
0.19136
0.25954
0.06531
0.44892

4
0.00562
1.04033
0.00002
0.00585
0.05025
0.06058
0.02920
0.01651
0.02521
0.05483
0.03592
0.02684
0.16171
0.01210
0.00013
0.05067
0.55151
0.04827
0.26604
0.20020
0.61546

6
0.00691
0.03767
1.00342
0.00349
0.06463
0.05904
0.03331
0.03530
0.01415
0.05312
0.03622
0.01872
0.15341
0.01467
0.00012
0.04899
0.46678
0.07757
0.35977
0.06469
0.59504

7
0.01238
0.05650
0.00113
1.00534
0.11633
0.12781
0.03641
0.01674
0.01946
0.06240
0.03849
0.04882
0.15656
0.01348
0.00018
0.05123
0.52418
0.11131
0.15955
0.05246
0.62222

8
0.10057
0.07842
0.00014
0.00590
1.11759
0.04244
0.04465
0.02271
0.01822
0.04806
0.02977
0.02304
0.12127
0.01403
0.00010
0.03562
0.37303
0.04223
0.20277
0.04573
0.43268

9
0.00549
0.02728
0.00002
0.00673
0.04780
1.06657
0.02604
0.02084
0.02784
0.05632
0.04456
0.02308
0.16866
0.01428
0.00015
0.05498
0.59007
0.06599
0.26833
0.18084
0.66782

1.71157

1.57578

1.58317

1.76326

1.70253

1.59063

Summing all of the industrial values in the I-A Inverse Matrix gives us the total
requirements multiplier for that industry. Summing the value added requirements gives
us the amount of value added per $1 of output.
Step 6. The final demands matrix. Now go back up to the first step. We need to
calculate total final demand and create a final demand vector that goes down through the
industries and households. We do that by summing the exogenous demand
(governments, investment and trade).

Gov Frgn Trade Dom Trade

11000
257.701
347.048
4.089
8900.338
5874.124
290.712
421.385
479.537
367.980
338.753

Final
Demand

Total

25000
28000 Grand Total
2220.339 3059.374
11702.812
734.956
0.000
7964.979
10.315
302.127
337.043
0.982
214.127
9923.073
8446.518 33080.433
58261.588
0.008 1270.184
10262.750
330.875 1933.598
7028.698
4.548
416.818
3441.841
104.418 2402.539
5518.285
261.679 6092.389
15242.033

5537.414
1082.004
316.531
9115.447
47401.075
1560.903
2685.858
900.904
2874.936
6692.821

We use these data to create a final demand matrix with zeros in all of the cells and final
demand values in the diagonals. This is boring, but simply use the copy/paste function to
do it.
Final Demands Matrix
(FD)
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9

1
5537.414
0.0
0.0
0.0
0.0
0.0

4
0.0
1082.004
0.0
0.0
0.0
0.0

6
0.0
0.0
316.531
0.0
0.0
0.0

7
0.0
0.0
0.0
9115.447
0.0
0.0

8
0.0
0.0
0.0
0.0
47401.075
0.0

9
0.0
0.0
0.0
0.0
0.0
1560.903

Step 7. Now we multiply our 22 X 22 SAM inverse matrix by our 22 X 1 vector of final
demands. (Note, my model is one vector smaller 21 X 21 and 21 X 1). We do that
using the =MMULT(,,) command where the first values are the SAM inverse values and
the second are the final demand vectors. To do this you highlight just the first vector
(column). Then write your =MMULT (SAM,Vector) command. Like before, you must
hit control-shift-enter to execute the command. Next, you need to copy the first vector
across to the remaining vectors. Just drag the highlighted column to the right until you
have finished the final demand matrix.

Final Demands Times

the Sam Inverse
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation

Industry
Payments:

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

Retail

1
4
6
7
8
9
10

1
6371.276
310.347
1.018
41.298
404.905
235.441
198.361

4
6.082
1125.644
0.020
6.325
54.367
65.551
31.598

6
2.189
11.922
317.614
1.105
20.458
18.687
10.544

7
112.824
515.034
10.313
9164.136
1060.430
1165.015
331.878

8
4767.151
3717.406
6.486
279.663
52974.746
2011.934
2116.247

9
8.572
42.578
0.031
10.500
74.610
1664.819
40.640

Step 8. From here on out we are summarizing the data. First we need to sum the
industrial amounts, the data from agriculture through other (sectors 17). We simply
compile these values using the =cell reference command (e.g, =c185) and copy that
relational reference down and across the industrial output data array.
We next sum the data by column to get an idea of total output by industry (I have also
compiled a percent of row total value for use later).
Summary of Industrial
Output
State of Iowa
Industry Receipts
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation
Utilities
Information
Fin/Insurance
Real Estate
Prof Svcs
Other Svcs
SL Gov
Fed Gov
Other

Industry
Payments:

1
4
6
7
8
9
10
11
12
13.1
13.2
14.1
14.2
15
16
17

Total Output
Percent of Total

Agriculture

Wholesale

Mining

Construction

All
Manufacturing

6371.276
310.347
1.018
41.298
404.905
235.441
198.361
132.748
67.266
309.156
384.958
108.325
602.720
104.405
0.782
204.667
9477.673
5.1%

6.082
1125.644
0.020
6.325
54.367
65.551
31.598
17.860
27.273
59.328
38.863
29.046
174.976
13.097
0.146
54.828
1705.003
0.9%

2.189
11.922
317.614
1.105
20.458
18.687
10.544
11.172
4.477
16.814
11.466
5.924
48.560
4.645
0.039
15.507
501.123
0.3%

112.824
515.034
10.313
9164.136
1060.430
1165.015
331.878
152.554
177.432
568.848
350.820
445.000
1427.095
122.910
1.595
466.979
16072.864
8.7%

4767.151
3717.406
6.486
279.663
52974.746
2011.934
2116.247
1076.442
863.489
2277.871
1411.207
1092.229
5748.389
664.954
4.961
1688.603
80701.778
43.4%

8.572
42.578
0.031
10.500
74.610
1664.819
40.640
32.528
43.452
87.906
69.554
36.018
263.259
22.289
0.234
85.824
2482.815
1.3%

Retail

Ive done the same thing for the value added array and the households array.
Summary of Value
Added
Emp Comp
Proprietors
Other Property
Ind Bus Tax

5000
6000
7000
8000

Total Value Added

Percent of Total
Summary of HH Income
Household Income
Percent of Total

10000

1168.173
1059.619
1437.175
361.625

596.741
52.224
287.861
216.616

147.751
24.555
113.879
20.477

4778.170
1014.606
1454.328
478.161

17682.137
2001.544
9611.725
2167.730

921.035
102.998
418.831
282.271

4026.592
4.3%

1153.443
1.2%

306.662
0.3%

7725.266
8.3%

31463.136
33.8%

1725.135
1.9%

2485.845
2.9%

665.933
0.8%

188.347
0.2%

5671.836
6.6%

20509.430
23.8%

1042.404
1.2%

Next we begin to construct a summary table of the output, value added, and the
household values. We copy the two summary rows for each array, the totals and the
percent of totals and we paste them to a new summary table using the paste special,
values, transpose options. We create three columns of data that are industry and
household specific: output, value added, and households. We can then format the table
and organize the information for additional analysis. This final table is a compilation of
the combined direct, indirect, and induced effects attributable to intermediate activity and
final demand. The totals should sum to the totals in the summaries table. (note:
occasionally the totals are slightly off. Dont worry about it unless you really, really
messed up. Its due to matrix accounting in step one that makes the SAM whole, but
might mess up one portion of the inverse calculations).
Final Demand Input Output Summary Table for Iowa

State of Iowa
Agriculture
Wholesale
Mining
Construction
All Manufacturing
Retail
Transportation
Utilities
Information
Fin/ Insurance
Real Estate
Prof Svcs
Other Svcs
SL Gov
Fed Gov
Other
Emp Comp
Proprietors
Other Property
Ind Bus Tax
Households

Total

Total Output

Percent of
Total

Total Value
Added

Percent of
Total

Total HH
Income

Percent of Total

9,477.67
1,705.00
501.12
16,072.86
80,701.78
2,482.81
4,734.55
1,285.31
4,688.46
11,317.37
1,943.38
2,159.58
12,771.08
13,706.95
1,626.71
(33.59)
20,669.13
185,810.17

5.1%
0.9%
0.3%
8.7%
43.4%
1.3%
2.5%
0.7%
2.5%
6.1%
1.0%
1.2%
6.9%
7.4%
0.9%
0.0%
11.1%
100.0%

4,026.59
1,153.44
306.66
7,725.27
31,463.14
1,725.13
2,570.20
823.18
2,428.42
6,276.41
1,299.38
1,446.46
7,001.70
10,929.31
1,345.31
(24.98)
12,470.68
92,966.31

4.3%
1.2%
0.3%
8.3%
33.8%
1.9%
2.8%
0.9%
2.6%
6.8%
1.4%
1.6%
7.5%
11.8%
1.4%
0.0%
13.4%
100.0%

2,485.84
665.93
188.35
5,671.84
20,509.43
1,042.40
1,802.44
399.28
1,514.12
3,821.27
595.89
1,085.16
5,004.21
8,207.36
1,069.94
(7.91)
32,063.23
86,118.77

2.9%
0.8%
0.2%
6.6%
23.8%
1.2%
2.1%
0.5%
1.8%
4.4%
0.7%
1.3%
5.8%
9.5%
1.2%
0.0%
37.2%
100.0%

We check our work by looking at the summary tables that came from the original model.
In particular, the Total Output and the Value Added amounts should be identical to the
values in the summary tables worksheet.
Note: in this table all of the value added row data have been erased as they are zero in
the penultimate table.