├── examples
    ├── poverty_famine_sen.pdf
    └── poverty_famine_sen_cleaned_prompt_12_3.txt
├── images
    ├── example_pdf_screenshot.jpg
    └── example_pdf_paper_screenshot.jpg
├── LICENSE
├── Readme.md
├── clean_pdf.py
└── run_config.py


/examples/poverty_famine_sen.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/kanodiaayush/make-doc-listenable/HEAD/examples/poverty_famine_sen.pdf


--------------------------------------------------------------------------------
/images/example_pdf_screenshot.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/kanodiaayush/make-doc-listenable/HEAD/images/example_pdf_screenshot.jpg


--------------------------------------------------------------------------------
/images/example_pdf_paper_screenshot.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/kanodiaayush/make-doc-listenable/HEAD/images/example_pdf_paper_screenshot.jpg


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2023 Ayush Kanodia
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/Readme.md:
--------------------------------------------------------------------------------
  1 | # make-doc-listenable
  2 | 
  3 | The goal of this project is to convert pdfs to a format that is listenable by a screen reader or text to speech engine. It cleans out references, headers and footers, and outputs pdfs to a text file, which can then be listened to with say Speechify or say Apple's TTS Engine.
  4 | 
  5 | # Requirements
  6 | 
  7 | ## Mac
  8 | 
  9 | 1. Install poppler with homebrew - `brew install poppler`.
 10 | 2. Package requirements - `pip install -r requirements.txt`.
 11 | 3. Add in your OpenAI API key to your environment and make sure you have credits.
 12 | I simply add the following line to my `.bashrc`: `export OPENAI_API_KEY=<your api key>`.
 13 | See [this link](https://help.openai.com/en/articles/4936850-where-do-i-find-my-secret-api-key) for more information.
 14 | 4. If you want output pdf as well, install pandoc - `brew install pandoc`.
 15 |    
 16 | The install steps should be similar for other operating systems, but I have not tested them.
 17 | 
 18 | # Running
 19 | 1. Setup your configs in `run_config.py`. The variable names are self explanatory.
 20 | 2. Run `python clean_pdf.py` to clean the pdfs and output the text files.
 21 | 
 22 | # Example
 23 | 
 24 | ## Book
 25 | 
 26 | Here is a screenshot of the page of a book, which is in pdf format.
 27 | 
 28 | ![pdf screenshot](images/example_pdf_screenshot.jpg)
 29 | 
 30 | Here is the text output towards of page 10 of the pdf (not of the book).
 31 | ```
 32 | Chapter 1¬                                                                                                                                                                                                                                                   
 33 | Poverty and Entitlements¬                                                                                                                                                                                                                                    
 34 | ¬                                                                                                                                                                                                                                                            
 35 | I.I ENTITLEMENTS AND OWNERSHIP¬                                                                                                                                                                                                                              
 36 | Starvation is the characteristic of some people not having enough food to eat. It is not the characteristic of there being not enough food to eat. While the latter can be a cause of the former, it is but one of many possible causes. Whether and how starvation relates to food supply is a matter for factual investigation.¬                                                                                                                                                                                        
 37 | Food supply statements say things about a commodity (or a group of commodities) considered on its own. Starvation statements are about the relationship of persons to the commodity (or that commodity group). Leaving out cases in which a person may deliberately starve, starvation statements translate readily into statements of ownership of food by persons. In order to understand starvation, it is, therefore, necessary to go into the structure of ownership.¬                                               
 38 | Ownership relations are one kind of entitlement relations. It is necessary to understand the entitlement systems within which the problem of starvation is to be analysed. This applies more generally to poverty as such, and more specifically to famines as well.¬                                                                                                                                                                                                                                                     
 39 | of legitimacy. It is a recursive relation and the process of connecting can be repeated. Consider a private ownership market economy. I own this loaf of bread. Why is this ownership accepted? Because I got it by exchange through paying some money I owned. Why is my ownership of that money accepted? Because I got it by selling a bamboo umbrella owned by me. Why is my ownership of the bamboo umbrella accepted? Because I made it with my own¬ 
 40 | 
 41 | 
 42 | 
 43 | 
 44 | 
 45 | 
 46 |  PAGE 10 ENDS
 47 | 
 48 | 
 49 | 
 50 | 
 51 | 
 52 | 
 53 | Poverty and Famines
 54 | 
 55 | labour using some bamboo from my land. Why is my ownership
 56 | of the land accepted? Because I inherited it from my father. Why
 57 | ```
 58 | 
 59 | ## Paper
 60 | 
 61 | Here is a screenshot of the page of a paper, which is in pdf format.
 62 | 
 63 | ![pdf screenshot](images/example_pdf_paper_screenshot.jpg)
 64 | 
 65 | Here is the text output towards the end of page 1 of the pdf (not of the paper).
 66 | 
 67 | ```
 68 | The reporters ask questions and soon warm to the manifesto’s importance. The
 69 | next day, the manifesto is carried as front-page news of most the world’s major
 70 | newspapers. For the next several days, at least, it is the talk of the world.
 71 | The Russell–Einstein manifesto galvanized the peace and disarmament
 72 | movements. It led to the Pugwash conferences, for which Joseph Rotblat and
 73 | the conference-series itself would eventually win the Nobel Peace Prize (1995).
 74 | Rotblat credits the manifesto for helping to create the conditions that gave rise
 75 | to the Nuclear Non-Proliferation Treaty (NPT, 1970). In his Nobel Peace Prize
 76 | acceptance speech, Rotblat explains:
 77 | 
 78 | 
 79 | 
 80 | 
 81 | 
 82 | 
 83 |  PAGE 1 ENDS
 84 | 
 85 | 
 86 | 
 87 | 
 88 | 
 89 | 
 90 | From my earliest days I had a passion for science. But science, the exercise of the supreme power of the human intellect, was always linked in my mind with benefit to people. I did not imagine that the second half of my life would be spent on efforts to avert a mortal danger to humanity created by science.
 91 | ```
 92 | 
 93 | See example full text output in the `examples` folder. The source pdf for this example is [this Amartya Sen book](https://www.google.com/books/edition/Poverty_and_Famines/FVC9eqGkMr8C).
 94 | 
 95 | This software was made initially for treehacks 2023 hosted at Stanford University. See [video demo](https://www.youtube.com/watch?v=c56qSiLCR6k&ab_channel=AyushKanodia).
 96 | 
 97 | ## Future Work
 98 | 
 99 | - [ ] Multiple sequential cleanings to remove headers too.
100 | 


--------------------------------------------------------------------------------
/clean_pdf.py:
--------------------------------------------------------------------------------
  1 | #!/usr/bin/env python3
  2 | 
  3 | import os
  4 | import textract
  5 | from PyPDF2 import PdfReader, PdfWriter
  6 | 
  7 | import openai,os,sys
  8 | 
  9 | 
 10 | # read in the pdf file
 11 | def clean_pdf(configs):
 12 |     pdf_file = configs['input_file']
 13 |     tmpdir = configs['tmpdir']
 14 |     start_page = configs['start_page']
 15 |     if start_page == -1:
 16 |         start_page = 0
 17 |     end_page = configs['end_page']
 18 |     if tmpdir == '':
 19 |         tmpdir = pdf_file.split('.')[0] + '_tmp'
 20 | 
 21 |     # split into pages, save
 22 |     if not os.path.exists(tmpdir):
 23 |         os.mkdir(tmpdir)
 24 | 
 25 |         # get the number of pages
 26 |     pdf = PdfReader(open(pdf_file, 'rb'))
 27 |     num_pages = len(pdf.pages)
 28 |     assert end_page <= num_pages, 'end_page is greater than the number of pages in the pdf'
 29 |     if end_page == -1:
 30 |         end_page = num_pages
 31 | 
 32 |     def pdf_to_text():
 33 |     # write each page to tmpdir
 34 |         for i in range(start_page, end_page):
 35 |             pdf_writer = PdfWriter()
 36 |             pdf_writer.add_page(pdf.pages[i])
 37 |             output_filename = '{}/page_{}.pdf'.format(tmpdir, i)
 38 |             # page_text = textract.process('{}/page_{}.pdf'.format(tmpdir, ii))
 39 |             with open(output_filename, 'wb') as out:
 40 |                 pdf_writer.write(out)
 41 |             page_text = textract.process(output_filename)
 42 |             output_filename_txt = '{}/page_{}.txt'.format(tmpdir, i)
 43 |             with open(output_filename_txt, 'wb') as out:
 44 |                 out.write(page_text)
 45 |             # remove pdf page
 46 |             os.remove(output_filename)
 47 | 
 48 |     def text_to_cleaned():
 49 |         # now extract only the text from the first page using textract
 50 |         pre_prompt = configs['pre_prompt']
 51 |         for ii in range(start_page, end_page):
 52 |             print("Cleaning page {}".format(ii))
 53 |             output_filename_txt = '{}/page_{}.txt'.format(tmpdir, ii)
 54 |             text = textract.process(output_filename_txt).decode('utf-8')
 55 |             prompt = f'{pre_prompt}\nSTART\n{text}'
 56 | 
 57 |             completions = openai.Completion.create(
 58 |                 model="text-davinci-003",
 59 |                 prompt=prompt,
 60 |                 temperature=0.05,
 61 |                 max_tokens=2000,
 62 |             )
 63 |             message = completions['choices'][0]['text']
 64 | 
 65 |             # if 'pre_prompt_2' in configs:
 66 |             #     pre_prompt_2 = configs['pre_prompt_2']
 67 |             #     prompt = f'{pre_prompt_2}\nSTART\n{message}'
 68 | 
 69 |             #     completions = openai.Completion.create(
 70 |             #         model="text-davinci-003",
 71 |             #         prompt=prompt,
 72 |             #         temperature=0.05,
 73 |             #         max_tokens=2000,
 74 |             #     )
 75 |             #     message = completions['choices'][0]['text']
 76 | 
 77 |             output_filename = '{}/page_{}_cleaned{}.txt'.format(tmpdir, ii, configs['suffix'])
 78 |             #print(prompt)
 79 |             #print(message)
 80 |             #print("\n\n\n\n\n\n")
 81 |             with open(output_filename, 'w') as out:
 82 |                 out.write(message)
 83 | 
 84 |     def cleaned_to_merged():
 85 |         # now merge all the pages into one file
 86 |         with open(f"{configs['output_file']}", 'w') as out:
 87 |             for ii in range(start_page, end_page):
 88 |                 output_filename = '{}/page_{}_cleaned{}.txt'.format(tmpdir, ii, configs['suffix'])
 89 |                 with open(output_filename, 'r') as f:
 90 |                     if configs['remove_newlines']:
 91 |                         cleaned_text = f.read().replace('\n', ' ')
 92 |                     else:
 93 |                         cleaned_text = f.read()
 94 |                     out.write(cleaned_text)
 95 |                 if configs['print_page_breaks']:
 96 |                     out.write(f'\n\n\n\n\n\n\n PAGE {ii} ENDS\n\n\n\n\n\n\n')
 97 |         if configs['output_to_pdf']:
 98 |             # write the contents of configs['output_file'] to configs['output_file'].pdf in pdf format
 99 |             # use pandoc
100 |             pdf_output_file = configs['output_file'].split('.')[0] + '.pdf'
101 |             os.system(f'pandoc {configs["output_file"]} -o {pdf_output_file} --pdf-engine=xelatex')
102 | 
103 |     def raw_to_merged():
104 |         # now merge all the pages into one file
105 |         with open(configs['output_file_raw'], 'w') as out:
106 |             for ii in range(start_page, end_page):
107 |                 output_filename = '{}/page_{}.txt'.format(tmpdir, ii)
108 |                 with open(output_filename, 'r') as f:
109 |                     out.write(f.read())
110 |                 out.write(f'\n\n\n\n\n\n\n PAGE {ii} ENDS\n\n\n\n\n\n\n')
111 | 
112 |     pdf_to_text()
113 |     raw_to_merged()
114 |     text_to_cleaned()
115 |     cleaned_to_merged()
116 | 
117 | if __name__ == '__main__':
118 |     from run_config import configs
119 |     clean_pdf(configs)
120 | '''
121 |     # Validate OPENAI_API_KEY
122 |     if 'OPENAI_API_KEY' not in os.environ:
123 |         print('OPENAI_API_KEY not found in environment variables')
124 |         print('Please set the OPENAI_API_KEY environment variable')
125 |         print('Aborting...')
126 |         sys.exit(1)
127 |     else:
128 |         try:
129 |             openai.Engine.list()
130 |             print('OPENAI_API_KEY is valid')
131 |         except:
132 |             print('OPENAI_API_KEY is invalid')
133 |             print('Please check the OPENAI_API_KEY environment variable')
134 |             print('Aborting...')
135 |             sys.exit(1)
136 |     
137 |     print('Enter input file path:')
138 |     configs['input_file'] = input()
139 | 
140 |     configs['output_file'] = f"{configs['input_file'].split('.')[0]}_cleaned{configs['suffix']}.txt"
141 | 
142 |     print(f'Enter output file path (Press enter to default to {configs["output_file"]}):')
143 |     configs['output_file'] = input()
144 |     if configs['output_file'] == '':
145 |         configs['output_file'] = f"{configs['input_file'].split('.')[0]}_cleaned{configs['suffix']}.txt"
146 | 
147 |     configs['output_file_raw'] = f"{configs['input_file'].split('.')[0]}_raw.txt"
148 |     if configs['output_file'] == '':
149 |         configs['output_file_raw'] = f"{configs['input_file'].split('.')[0]}_raw.txt"
150 | 
151 |     configs['tmpdir'] = f"{configs['input_file'].split('.')[0]}_tmp"
152 | 
153 |     print('Enter start page (Press enter to default to 0):')
154 |     configs['start_page'] = input()
155 |     if configs['start_page'] == '':
156 |         configs['start_page'] = 0
157 |     else:
158 |         configs['start_page'] = int(configs['start_page'])
159 |     print('Enter end page (Press enter to default to -1):')
160 |     configs['end_page'] = input()
161 |     if configs['end_page'] == '':
162 |         configs['end_page'] = -1
163 |     else:
164 |         configs['end_page'] = int(configs['end_page'])
165 | 
166 |     print('Enter tmpdir (Press enter to default to {}):'.format(configs['tmpdir']))
167 |     configs['tmpdir'] = input()
168 |     if configs['tmpdir'] == '':
169 |         configs['tmpdir'] = f"{configs['input_file'].split('.')[0]}_tmp"
170 | 
171 |     print('Enter if you want output to pdf (Press enter to default to False): (True/False)')
172 |     print('Make sure pandoc is installed if you want to output to pdf')
173 |     configs['output_to_pdf'] = input()
174 |     if configs['output_to_pdf'] == '':
175 |         configs['output_to_pdf'] = False
176 |     else:
177 |         if configs['output_to_pdf'] == 'True':
178 |             configs['output_to_pdf'] = True
179 |         else:
180 |             configs['output_to_pdf'] = False
181 | 
182 |     print('Running:')
183 |     clean_pdf(configs)
184 | '''
185 | 


--------------------------------------------------------------------------------
/run_config.py:
--------------------------------------------------------------------------------
 1 | #!/usr/bin/env python3
 2 | 
 3 | configs = {
 4 |     'suffix' : '_prompt_12_3',
 5 |     'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START. Remove references. Remove isolated digits and numbers. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct. Do not complete the last sentence.",
 6 |     # 'pre_prompt_2' : "Produce an exact replica of the text provided. The text starts at START. Remove the header.",
 7 | }
 8 | 
 9 | configs.update({
10 |     'input_file': 'data/poverty_famine_sen.pdf',
11 |     'output_file': f'data/poverty_famine_sen_cleaned{configs["suffix"]}.txt',
12 |     'output_file_raw': 'data/poverty_famine_sen_raw.txt',
13 |     'tmpdir': 'data/poverty_famine_sen_tmp',
14 |     'start_page': 30,
15 |     'end_page' : 90,
16 |     'output_to_pdf' : True,
17 |     'print_page_breaks' : True,
18 |     'remove_newlines' : True,
19 | })
20 | 
21 | # configs.update({
22 | #     'input_file': 'data/moral-fn.pdf',
23 | #     'output_file': f'data/moral-fn_cleaned{configs["suffix"]}.txt',
24 | #     'output_file_raw': 'data/moral-fn_raw.txt',
25 | #     'tmpdir': 'data/poverty_famine_sen_tmp',
26 | #     'start_page': 1,
27 | #     'end_page' : 3,
28 | #     'output_to_pdf' : True,
29 | #     'print_page_breaks' : False,
30 | # })
31 | 
32 | # other prompts tried
33 |     # 'suffix' : '_prompt_1',
34 |     # 'pre_prompt' : "Please cleanup this text for me by removing any page numbers, references, footers, or headers, if any. Don't change the text if it is not a page number, reference, footer, or header. Please also remove any blank lines, and fix spelling. Please add punctuation wherever necessary. Please add anything that will help an automated text to speech read it out. Thank you!",
35 |     # 'suffix' : '_prompt_2',
36 |     # 'pre_prompt' : "Please copy the text, and cleanup this text by removing any page numbers, references, footers, or headers. Don't change the text if it is not a page number, reference, footer, or header. Please also remove any blank lines, and fix spelling. Please add punctuation wherever necessary. Please add anything that will help an automated text to speech read it out.",
37 |     # 'suffix' : '_prompt_3',
38 |     # 'pre_prompt' : "After START is text extracted from a pdf. Please copy the text, and cleanup this text by removing any page numbers, references, footers, or headers. At END, the text ends. Do not summarize the text or complete its end even if its incomplete.",
39 |     # 'suffix' : '_prompt_4',
40 |     # 'pre_prompt' : "After START is text extracted from a page of a paper or a book in pdf format. Please copy the text, and cleanup this text by removing any page numbers, footnotes and references. Note that the start and end of the page (outside of footnotes and references in footnotes) will be incomplete snippets, please do not alter them. At END, the text ends.",
41 |     # 'suffix' : '_prompt_5',
42 |     # 'pre_prompt' : "After START is text extracted from a page of a paper or a book in pdf format. Please copy the text, and cleanup this text by removing any page numbers, footnotes and references. Note that the start of the page and the part right before footnotes (outside of footnotes and references in footnotes) will be incomplete snippets, please do not alter them. At END, the text ends. Please capitalize headings, and add new lines before and after headings. Please add newlines between paragraphs.",
43 |     # 'suffix' : '_prompt_6',
44 |     # 'pre_prompt' : "After START is text extracted from a page of a paper or a book in pdf format. Please copy the text, and cleanup this text by removing any page numbers, footnotes and references. Note that the start of the page and the part right before footnotes (outside of footnotes and references in footnotes) will be incomplete snippets, please do not alter them. At END, the text ends. Please capitalize headings, and add new lines before and after headings. Please add newlines between paragraphs. Please do not autocomplete incomplete sentences.",
45 |     # 'suffix' : '_prompt_7',
46 |     # 'pre_prompt' : "After START is text extracted from a page of a paper or a book in pdf format. Please copy the text, and cleanup this text by removing any page numbers, footnotes and references. Note that the start of the page and the part right before footnotes (outside of footnotes and references in footnotes) will be incomplete snippets, please do not alter them. Right before the footnotes, if there is an incomplete sentence, do not autocomplete it. At END, the text ends. Please capitalize headings, and add new lines before and after headings. Please add newlines between paragraphs. Please do not autocomplete incomplete sentences.",
47 |     # 'suffix' : '_prompt_8',
48 |     # 'pre_prompt' : "Produce an exact replica of the text provided, minus the footnotes and references. Do not autocomplete any sentences.",
49 |     # 'suffix' : '_prompt_9',
50 |     # 'pre_prompt' : "Produce an exact replica of the text provided, minus the footnotes and references. Do not autocomplete any sentences at the end of the text produced.",
51 |     # 'suffix' : '_prompt_10',
52 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END.",
53 |     # 'suffix' : '_prompt_11',
54 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references. Do not auto complete any sentences. Do not finish incomplete sentences. Leave the end of the text as is, do not finish or autocomplete it. Only do edits, no additions.",
55 |     # 'suffix' : '_prompt_12',
56 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references. Only do edits, no additions.",
57 |     # 'suffix' : '_prompt_12_1',
58 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct.",
59 |     # 'suffix' : '_prompt_12_2',
60 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove START and END. Remove page numbers and isolated numbers. Remove references. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct.",
61 |     # 'suffix' : '_prompt_12_3',
62 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START. Remove references. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct. Do not complete the last sentence.",
63 |     # 'suffix' : '_prompt_12_4',
64 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START. Remove references. Remove isolated digits and numbers, but keep headings. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct. Do not complete the last sentence.",
65 |     # 'suffix' : '_prompt_12_5',
66 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START. Remove references. Remove isolated digits and numbers, but keep headings. Delete footers by detecting the start of it and deleting everything after that. Only do deletions, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct. Do not complete the last sentence.",
67 |     # 'suffix' : '_prompt_13',
68 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references and footnotes. Only do edits, no additions.",
69 |     # 'suffix' : '_prompt_14',
70 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references. Please capitalize headings, and add new lines before and after headings. Only do edits, no additions.",
71 |     # 'suffix' : '_prompt_14',
72 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START and ends at END. Remove references. Remove references both twoards the end of the page and inline. References are often in parantheses and contain the word see. Please capitalize headings, and add new lines before and after headings. Only do edits, no additions.",
73 |     # 'suffix' : '_prompt_12_3_1',
74 |     # 'pre_prompt' : "Produce an exact replica of the text provided. The text starts at START. Remove references. Remove isolated digits and numbers. Remove newlines and line breaks except after headings. Only do edits, no additions. Do not autocomplete any sentences. The end of output may seem abrupt, that is correct. Do not complete the last sentence.",
75 | 


--------------------------------------------------------------------------------
/examples/poverty_famine_sen_cleaned_prompt_12_3.txt:
--------------------------------------------------------------------------------
   1 | Chapter 1
   2 | Poverty and Entitlements
   3 | 
   4 | I.I ENTITLEMENTS AND OWNERSHIP
   5 | Starvation is the characteristic of some people not having enough food to eat. It is not the characteristic of there being not enough food to eat. While the latter can be a cause of the former, it is but one of many possible causes. Whether and how starvation relates to food supply is a matter for factual investigation.
   6 | Food supply statements say things about a commodity (or a group of commodities) considered on its own. Starvation statements are about the relationship of persons to the commodity (or that commodity group). Leaving out cases in which a person may deliberately starve, starvation statements translate readily into statements of ownership of food by persons. In order to understand starvation, it is, therefore, necessary to go into the structure of ownership.
   7 | Ownership relations are one kind of entitlement relations. It is necessary to understand the entitlement systems within which the problem of starvation is to be analysed. This applies more generally to poverty as such, and more specifically to famines as well.
   8 | An entitlement relation applied to ownership connects one set of ownerships to another through certain rules of legitimacy. It is a recursive relation and the process of connecting can be repeated. Consider a private ownership market economy. I own this loaf of bread. Why is this ownership accepted? Because I got it by exchange through paying some money I owned. Why is my ownership of that money accepted? Because I got it by selling a bamboo umbrella owned by me. Why is my ownership of the bamboo umbrella accepted? Because I made it with my own
   9 | 
  10 | 
  11 | 
  12 | 
  13 | 
  14 | 
  15 |  PAGE 10 ENDS
  16 | 
  17 | 
  18 | 
  19 | 
  20 | 
  21 | 
  22 | Poverty and Famines
  23 | 
  24 | labour using some bamboo from my land. Why is my ownership
  25 | of the land accepted? Because I inherited it from my father. Why
  26 | is his ownership of that land accepted? And so on. Each link in
  27 | this chain of entitlement relations 'legitimizes' one set of
  28 | ownership by reference to another, or to some basic entitlement
  29 | in the form of enjoying the fruits of one's own labour. Entitlement relations accepted in a private ownership market
  30 | economy typically include the following, among others:
  31 | (i) trade-based entitlement: one is entitled to own what one obtains
  32 | by trading something one owns with a willing party (or,
  33 | multilaterally, with a willing set of parties);
  34 | (2) production-based entitlement, one is entitled to own what one
  35 | gets by arranging production using one's owned resources, or
  36 | resources hired from willing parties meeting the agreed
  37 | conditions of trade;
  38 | (3) own-labour entitlement: one is entitled to one's own labour
  39 | power, and thus to the trade-based and production-based
  40 | entitlements related to one's labour power;
  41 | (4) inheritance and transfer entitlement: one is entitled to own what is
  42 | willingly given to one by another who legitimately owns it,
  43 | possibly to take affect after the latter's death (if so specified
  44 | by him).
  45 | These are some entitlement relations of more or less straightforward kind, but there are others, frequently a good deal more
  46 | complex. For example, one may be entitled to enjoy the fruits of
  47 | some property without being able to trade it for anything else. Or
  48 | one may be able to inherit the property of a deceased relation
  49 | who did not bequeath it to anyone, through some rule of kinshipbased inheritance accepted in the country in question. Or one
  50 | may have some entitlements related to unclaimed objects on the
  51 | basis of discovery. Market entitlements may even be supplemented by rationing or coupon systems, even in private
  52 | ownership market economies, such as in Britain during the last
  53 | war.
  54 | This may or may not be combined with price 'control', and that in its turn may or
  55 | may not be combined with a flourishing 'black market'; see Dasgupta (1950) for an
  56 | illuminating analysis of black market prices.
  57 | 
  58 | 
  59 | 
  60 | 
  61 | 
  62 | 
  63 |  PAGE 11 ENDS
  64 | 
  65 | 
  66 | 
  67 | 
  68 | 
  69 | 
  70 | entitlement are the economic system, the ownership structure,
  71 | the production technology, the distribution of resources, and the
  72 | market structure. The economic system determines the
  73 | ownership structure, and the ownership structure determines
  74 | the exchange entitlement. The production technology and the
  75 | distribution of resources determine the production-based
  76 | entitlements, and the market structure determines the
  77 | exchange-based entitlements.
  78 | Poverty and entitlements are closely related. Poverty is
  79 | defined as a situation in which a person's ownership bundle
  80 | belongs to the starvation set. Entitlements are the rights and
  81 | obligations that arise from ownership relations. Poverty is
  82 | caused by a lack of entitlements, and entitlements are
  83 | determined by ownership relations. The scope of ownership
  84 | relations can vary greatly with economic systems. A socialist
  85 | economy may not permit private ownership of 'means of
  86 | production', thereby rendering 'production-based entitlements'
  87 | inoperative except when it involves just one's own labour and
  88 | some elementary tools and raw materials. A capitalist
  89 | economy will not only permit the private ownership of means
  90 | of production; that is indeed one of its main foundations. On
  91 | the other hand, a capitalist economy—like a socialist one—will
  92 | not permit ownership of one human being by another, as a
  93 | slave economy will. A socialist economy may restrict the
  94 | employment of one person by another for production
  95 | purposes, i.e. constrain the possibility of private trading of
  96 | labour power for productive use. A capitalist economy will
  97 | not, of course, do this, but may impose restrictions on binding
  98 | contracts involving labour-power obligations over long periods
  99 | of time. This, however, is the standard system under some
 100 | feudal practices involving bonded labour, and also in some
 101 | cases of colonial plantations.
 102 | In a market economy, a person can exchange what he owns for
 103 | another collection of commodities. He can do this exchange
 104 | either through trading, or through production, or through a
 105 | combination of the two. The set of all the alternative bundles of
 106 | commodities that he can acquire in exchange for what he owns
 107 | may be called the 'exchange entitlement' of what he owns.
 108 | The 'exchange entitlement mapping' is the relation that
 109 | specifies the set of exchange entitlements for each ownership
 110 | bundle. This relation—E-mapping for brevity—defines the
 111 | possibilities that would be open to him corresponding to each
 112 | ownership situation. A person will be exposed to starvation if, for
 113 | the ownership that he actually has, the exchange entitlement set
 114 | does not contain any feasible bundle including enough food.
 115 | Given the E-mapping, it is in this way possible to identify those
 116 | ownership bundles—call them collectively the starvation set—
 117 | that must, thus, lead to starvation in the absence of nonentitlement transfers. E-mappings, starvation sets, and related
 118 | concepts are discussed in Chapter 5 and are formally analysed
 119 | in Appendix A, and here we are concerned only with the
 120 | underlying ideas.
 121 | Among the influences that determine a person's exchange
 122 | entitlement are the economic system, the ownership structure,
 123 | the production technology, the distribution of resources, and the
 124 | market structure. The economic system determines the
 125 | ownership structure, and the ownership structure determines
 126 | the exchange entitlement. The production technology and the
 127 | distribution of resources determine the production-based
 128 | entitlements, and the market structure determines the
 129 | exchange-based entitlements. Poverty and entitlements are
 130 | closely related. Poverty is defined as a situation in which a
 131 | person's ownership bundle belongs to the starvation set.
 132 | Entitlements are the rights and obligations that arise from
 133 | ownership relations. Poverty is caused by a lack of
 134 | entitlements, and entitlements are determined by ownership
 135 | relations.
 136 | 
 137 | 
 138 | 
 139 | 
 140 | 
 141 | 
 142 |  PAGE 12 ENDS
 143 | 
 144 | 
 145 | 
 146 | 
 147 | 
 148 | 
 149 | 
 150 | Poverty and Famines
 151 | 
 152 | entitlement mapping can differ with the mode of production.
 153 | For example, in a subsistence economy, the exchange entitlements
 154 | of a person will depend on the amount of land he owns and
 155 | the amount of labour he can command. In a capitalist economy,
 156 | the exchange entitlements of a person will depend on his ownership
 157 | bundle (including labour power) and the wage rate he can
 158 | command. In a socialist economy, the exchange entitlements of
 159 | a person will depend on his ownership bundle (including labour
 160 | power) and the amount of resources he can command from
 161 | the state.
 162 | A person's ability to avoid starvation will depend both on his
 163 | ownership and on the exchange entitlement mapping that he
 164 | faces. A general decline in food supply may indeed cause him to
 165 | be exposed to hunger through a rise in food prices with an
 166 | unfavourable impact on his exchange entitlement. Even when
 167 | his starvation is caused by food shortage in this way, his immediate
 168 | reason for starvation will be the decline in his exchange
 169 | entitlement.
 170 | More importantly, his exchange entitlement may worsen for
 171 | reasons other than a general decline of food supply. For example,
 172 | given the same total food supply, other groups' becoming richer
 173 | and buying more food can lead to a rise in food prices, causing a
 174 | worsening of exchange entitlement. Or some economic change
 175 | may affect his employment possibilities, leading also to worse
 176 | exchange entitlement. Similarly, his wages can fall behind prices.
 177 | Or the price of necessary resources for the production he engages
 178 | in can go up relatively. These diverse influences on exchange
 179 | entitlements are as relevant as the overall volume of food supply
 180 | vis-à-vis population.
 181 | I.3 MODES OF PRODUCTION
 182 | The exchange entitlements faced by a person depend, naturally,
 183 | on his position in the economic class structure as well as the
 184 | modes of production in the economy. What he owns will vary
 185 | with his class, and even if exactly the same E-mapping were to
 186 | hold for all, the actual exchange entitlements would differ with
 187 | his ownership position.
 188 | But even with the same ownership position, the exchange 
 189 | entitlement mapping can differ with the mode of production.
 190 | For example, in a subsistence economy, the exchange entitlements
 191 | of a person will depend on the amount of land he owns and
 192 | the amount of labour he can command. In a capitalist economy,
 193 | the exchange entitlements of a person will depend on his ownership
 194 | bundle (including labour power) and the wage rate he can
 195 | command. In a socialist economy, the exchange entitlements of
 196 | a person will depend on his ownership bundle (including labour
 197 | power) and the amount of resources he can command from
 198 | the state.
 199 | A person's ability to avoid starvation will depend both on his
 200 | ownership and on the exchange entitlement mapping that he
 201 | faces. A general decline in food supply may indeed cause him to
 202 | be exposed to hunger through a rise in food prices with an
 203 | unfavourable impact on his exchange entitlement. Even when
 204 | his starvation is caused by food shortage in this way, his immediate
 205 | reason for starvation will be the decline in his exchange
 206 | entitlement.
 207 | More importantly, his exchange entitlement may worsen for
 208 | reasons other than a general decline of food supply. For example,
 209 | given the same total food supply, other groups' becoming richer
 210 | and buying more food can lead to a rise in food prices, causing a
 211 | worsening of exchange entitlement. Or some economic change
 212 | may affect his employment possibilities, leading also to worse
 213 | exchange entitlement. Similarly, his wages can fall behind prices.
 214 | Or the price of necessary resources for the production he engages
 215 | in can go up relatively. These diverse influences on exchange
 216 | entitlements are as relevant as the overall volume of food supply
 217 | vis-à-vis population.
 218 | I.3 MODES OF PRODUCTION
 219 | The exchange entitlements faced by a person depend, naturally,
 220 | on his position in the economic class structure as well as the
 221 | modes of production in the economy. What he owns will vary
 222 | with his class, and even if exactly the same E-mapping were to
 223 | hold for all, the actual exchange entitlements would differ with
 224 | his ownership position.
 225 | But even with the same ownership position, the exchange 
 226 | entitlement mapping can differ with the mode of production.
 227 | For example, in a subsistence economy, the exchange entitlements
 228 | of a person will depend on the amount of land he owns and
 229 | the amount of labour he can command. In a capitalist economy,
 230 | the exchange entitlements of a person will depend on his ownership
 231 | bundle (including labour power) and the wage rate he can
 232 | command. In a socialist economy, the exchange entitlements of
 233 | a person will depend on his ownership bundle (including labour
 234 | power) and the amount of resources he can command from
 235 | the state.
 236 | 
 237 | 
 238 | 
 239 | 
 240 | 
 241 | 
 242 |  PAGE 13 ENDS
 243 | 
 244 | 
 245 | 
 246 | 
 247 | 
 248 | 
 249 | Poverty and Entitlements
 250 | 
 251 | Entitlements will be different depending on what economic prospects are open to him, and that will depend on the modes of production and his position in terms of production relations. For example, while a peasant differs from a landless labourer in terms of ownership (since he owns land, which the labourer does not), the landless share-cropper differs from the landless labourer not in their respective ownerships, but in the way they can use the only resource they own, viz. labour power. The landless labourer will be employed in exchange for a wage, while the share-cropper will do the cultivation and own a part of the product.
 252 | This difference can lead not merely to contrasts of the levels of typical remuneration of the two, which may or may not be very divergent, but also to sharp differences in exchange entitlements in distress situations. For example, a cyclone reducing the labour requirement for cultivation by destroying a part of the crop in each farm may cause some casual agricultural labourers to be simply fired, leading to a collapse of their exchange entitlements, while others are retained. In contrast, in this case the sharecroppers may all operate with a lower labour input and lower entitlement, but no one may become fully jobless and thus incomeless.
 253 | Similarly, if the output is food, e.g. rice or wheat, the sharecropper gets his return in a form such that he can directly eat it without going through the vagaries of the market. In contrast, the agricultural labourer paid in money terms will have to depend on the exchange entitlement of his money wage. When famines are accompanied by sharp changes in relative prices—and in particular a sharp rise in food prices—there is much comparative merit in being a share-cropper rather than an agricultural labourer, especially when the capital market is highly imperfect. The greater production risk of the sharecropper compared with the security of a fixed wage on the part of the agricultural labourer has been well analysed; but a fixed money wage may offer no security at all in a situation of sharply varying food prices (even when employment is guaranteed). In contrast, a share of the food output does have some security advantage in terms of exchange entitlement.
 254 | 
 255 | 
 256 | 
 257 | 
 258 | 
 259 | 
 260 |  PAGE 14 ENDS
 261 | 
 262 | 
 263 | 
 264 | 
 265 | 
 266 | 
 267 | 
 268 | Poverty and Famines
 269 | Similarly, those who sell services (e.g. barbers or rickshawpullers) or handicraft products (e.g. weavers or shoemakers) are—like wage labourers—more exposed, in this respect, to famines involving unexpected rises of food prices than are peasants or share-croppers producing food crops. This is the case even when the typical standard of living of the latter is no higher than that of the former.
 270 | In understanding general poverty, or regular starvation, or outbursts of famines, it is necessary to look at both ownership patterns and exchange entitlements, and at the forces that lie behind them. This requires careful consideration of the nature of modes of production and the structure of economic classes as well as their interrelations. Later in the monograph, when actual famines are analysed, these issues will emerge more concretely.
 271 | I.4 SOCIAL SECURITY AND EMPLOYMENT ENTITLEMENTS
 272 | The exchange entitlements depend not merely on market exchanges but also on those exchanges, if any, that the state provides as a part of its social security programme. Given a social security system, an unemployed person may get 'relief, an old person a pension, and the poor some specified 'benefits'. These affect the commodity bundles over which a person can have command. They are parts of a person's exchange entitlements, and are conditional on the absence of other exchanges that a person might undertake. For example, a person is not entitled to unemployment benefit if he exchanges his labour power for a wage, i.e. becomes employed. Similarly, exchanges that make a person go above the specified poverty norm will make him ineligible for receiving the appropriate relief. These social security provisions are essentially supplementations of the processes of market exchange and production, and the two types of opportunities together determine a person's exchange entitlements in a private ownership market economy with social security provisions.
 273 | The social security arrangements are particularly important in the context of starvation. The reason why there are no famines in the rich developed countries is not because people are generally rich on the average. Rich they certainly are when they have jobs and earn a proper wage; but for large numbers of people this condition fails to hold for long periods of time, and the exchange
 274 | 
 275 | 
 276 | 
 277 | 
 278 | 
 279 | 
 280 |  PAGE 15 ENDS
 281 | 
 282 | 
 283 | 
 284 | 
 285 | 
 286 | 
 287 | Poverty and Entitlements
 288 | 
 289 | entitlements of their endowments in the absence of social security arrangements could provide very meagre commodity bundles indeed. With the proportion of unemployment as high as it is, say, in Britain or America today, but for the social security arrangements there would be widespread starvation and possibly a famine. What prevents that is not the high average income or wealth of the British or the general opulence of the Americans, but the guaranteed minimum values of exchange entitlements owing to the social security system.
 290 | Similarly, the elimination of starvation in socialist economies —for example in China—seems to have taken place even without a dramatic rise in food availability per head, and indeed, typically the former has preceded the latter. The end of starvation reflects a shift in the entitlement system, both in the form of social security and—more importantly—through systems of guaranteed employment at wages that provide exchange entitlement adequate to avoid starvation.
 291 | 1.5 FOOD SUPPLY AND STARVATION
 292 | There has been a good deal of discussion recently about the prospect of food supply falling significantly behind the world population. There is, however, little empirical support for such a diagnosis of recent trends. Indeed, for most areas in the world—with the exception of parts of Africa—the increase in food supply has been comparable to, or faster than, the expansion of population. But this does not indicate that starvation is being systematically eliminated, since starvation—as discussed—is a function of entitlements and not of food availability as such. Indeed, some of the worst famines have taken place with no significant decline in food availability per head.
 293 | To say that starvation depends 'not merely' on food supply but also on its 'distribution' would be correct enough, though not remarkably helpful. The important question then would be: what determines distribution of food between different sections of the community? The entitlement approach directs one to questions dealing with ownership patterns and—less obviously
 294 | 
 295 | 
 296 | 
 297 | 
 298 | 
 299 | 
 300 |  PAGE 16 ENDS
 301 | 
 302 | 
 303 | 
 304 | 
 305 | 
 306 | 
 307 | Poverty and Famines
 308 | 
 309 | but no less importantly—to the various influences that affect
 310 | exchange entitlement mappings (see Appendices A and B, and
 311 | Chapters 5-10). In so far as food supply itself has any influence on
 312 | the prevalence of starvation, that influence is seen as working
 313 | through the entitlement relations. If one person in eight starves
 314 | regularly in the world, this is seen as the result of his inability to
 315 | establish entitlement to enough food; the question of the physical
 316 | availability of the food is not directly involved.
 317 | The approach of entitlements used in this work is very general
 318 | and—I would argue—quite inescapable in analysing starvation
 319 | and poverty. If, nevertheless, it appears odd and unusual, this
 320 | can be because of the hold of the tradition of thinking in terms of
 321 | what exists rather than in terms of who can command what. The
 322 | mesmerizing simplicity of focusing on the ratio of food to
 323 | population has persistently played an obscuring role over
 324 | centuries, and continues to plague policy discussions today much
 325 | as it has deranged anti-famine policies in the past.
 326 | 
 327 | 
 328 | 
 329 | 
 330 | 
 331 | 
 332 |  PAGE 17 ENDS
 333 | 
 334 | 
 335 | 
 336 | 
 337 | 
 338 | 
 339 | Chapter 2
 340 | Concepts of Poverty
 341 | 
 342 | 2.1 REQUIREMENTS OF A CONCEPT OF POVERTY
 343 | On his deathbed in Calcutta, J. B. S. Haldane wrote a poem
 344 | called 'Cancer's a funny thing'. Poverty is no less funny.
 345 | Consider the following view of poverty:
 346 | People must not be allowed to become so poor that they offend or are
 347 | hurtful to society. It is not so much the misery and plight of the poor but
 348 | the discomfort and cost to the community which is crucial to this view of
 349 | poverty. We have a problem of poverty to the extent that low income
 350 | creates problems for those who are not poor.
 351 | To live in poverty may be sad, but to 'offend or [be] hurtful to
 352 | society', creating 'problems for those who are not poor' is, it
 353 | would appear, the real tragedy. It isn't easy to push much further
 354 | the reduction of human beings into 'means'.
 355 | The first requirement of the concept of poverty is of a criterion
 356 | as to who should be the focus of our concern. The specification of
 357 | certain 'consumption norms', or of a 'poverty line', may do part
 358 | of the job: 'the poor' are those people whose consumption
 359 | standards fall short of the norms, or whose incomes lie below that
 360 | line. But this leads to a further question: is the concept of poverty
 361 | to be related to the interests of: (i) only the poor, (iii) both the
 362 | poor and the non-poor?
 363 | It seems a bit grotesque to hold that the concept of poverty
 364 | should be concerned only with the non-poor, and I take the
 365 | liberty of dropping (ii)—and the 'view' quoted in the first
 366 | paragraph—without further ado. Alternative (iii) might,
 367 | however, appear to be appealing, since it is broad-based and
 368 | unrestrictive. There is little doubt that the penury of the poor
 369 | does, in fact, affect the well-being of the rich. The real question is
 370 | whether such effects should enter into the concept of poverty as
 371 | 
 372 | 
 373 | 
 374 | 
 375 | 
 376 | 
 377 |  PAGE 18 ENDS
 378 | 
 379 | 
 380 | 
 381 | 
 382 | 
 383 | 
 384 | Poverty and Famines: An Essay on Entitlement and Deprivation, Oxford: Clarendon
 385 | Press, 1981.
 386 | 
 387 | Poverty and Famines
 388 | 
 389 | Such, or whether they should figure under the possible effects of poverty. I believe a good case can be made for choosing the latter alternative, since in an obvious sense poverty must be a characteristic of the poor rather than of the non-poor. One can, for instance, argue that, if one considers a case of reduction of real income and increase in the suffering of all the poor, it must be described as an increase of poverty, no matter whether this change is accompanied by a reduction in the adverse effects on the rich (e.g. whether the rich are less 'offended' by the sight of penury).
 390 | This conception of poverty based on does not, of course, imply any denial of the fact that the suffering of the poor themselves may depend on the condition of the non-poor. It merely asserts that the focus of the concept of poverty has to be on the well-being of the poor as such, no matter what influences affect their well-being. Causation of poverty and effects of poverty will be important issues to study on their own rights, and the conceptualization of poverty in terms of the conditions only of the poor does not affect the worthwhileness of studying these questions. Indeed, there will be much to say on these questions later on in the book.
 391 | It is perhaps worth mentioning in this context that in some discussions one is concerned not with the prevalence of poverty in a country in the form of the suffering of the poor, but with the relative opulence of the nation as a whole. In those discussions it will, of course, be entirely legitimate to be concerned with the well-being of all the people in the nation, and the description of a nation as 'poor' must obviously relate to such a broader concept. These are different exercises, and so long as this fact is clearly recognized there need not be any confusion.
 392 | Even after we have identified the poor and specified that the concept of poverty is concerned with the conditions of the poor, much remains to be done. There is the problem of aggregation—often important—over the group of the poor, and this involves moving from the description of the poor to some over-all measure of'poverty' as such. In some traditions, this is done very simply by just counting the number of the poor, and then expressing poverty as the ratio of the number of the poor to the total number of people in the community in question.
 393 | 
 394 | 
 395 | 
 396 | 
 397 | 
 398 | 
 399 |  PAGE 19 ENDS
 400 | 
 401 | 
 402 | 
 403 | 
 404 | 
 405 | 
 406 | 
 407 | In this chapter, the focus is on concepts of poverty. A head-count measure, referred to as H, is the most widely used measure, but it has two serious drawbacks. It does not take into account the extent of the shortfall of incomes of the poor from the poverty line, nor is it sensitive to the distribution of income among the poor. This chapter is not concerned with problems of measurement, but rather with the general ideas on the conception of poverty. 
 408 | Two distinct exercises are required for a concept of poverty: identification of a group of people as poor, and aggregation of the characteristics of the set of poor people into an overall image of poverty. Alternative approaches to the concept of poverty can be found in the literature, and these will be evaluated in the following sections. 
 409 | The biological approach, as defined by Seebohm Rowntree in 1901, defines families as being in primary poverty if their total earnings are insufficient to obtain the minimum necessities for the maintenance of merely physical efficiency. 
 410 | The subsistence approach, which is based on the idea that poverty is a lack of basic needs, is also discussed. This approach is based on the notion that poverty is a lack of basic needs, and it is argued that this approach is inadequate in that it does not take into account the social and psychological aspects of poverty. 
 411 | The relative deprivation approach is based on the idea that poverty is a lack of resources relative to the resources of others. This approach is criticized for its lack of precision and its reliance on subjective judgments. 
 412 | Finally, the capability approach is discussed. This approach is based on the idea that poverty is a lack of capabilities, and it is argued that this approach is more comprehensive than the other approaches as it takes into account both the material and non-material aspects of poverty.
 413 | 
 414 | 
 415 | 
 416 | 
 417 | 
 418 | 
 419 |  PAGE 20 ENDS
 420 | 
 421 | 
 422 | 
 423 | 
 424 | 
 425 | 
 426 | Poverty and Famines
 427 | 
 428 | survival or work efficiency have been often used in defining the
 429 | poverty line. Starvation, clearly, is the most telling aspect of
 430 | poverty.
 431 | The biological approach has come under rather intense fire
 432 | recently. There are indeed several problems with its use. First,
 433 | there are significant variations related to physical features,
 434 | climatic conditions and work habits. In fact, even for a specific
 435 | group in a specific region, nutritional requirements are difficult
 436 | to define precisely. People have been known to survive with
 437 | incredibly little nutrition, and there seems to be a cumulative
 438 | improvement of life expectation as the dietary limits are raised.
 439 | In fact, physical opulence seems to go on increasing with
 440 | nutrition over a very wide range; Americans, Europeans and
 441 | Japanese have been growing measurably in stature as their diets
 442 | have continued to improve. There is difficulty in drawing a line
 443 | somewhere, and the so-called 'minimum nutritional requirements' have an inherent arbitrariness that goes well beyond
 444 | variations between groups and regions.
 445 | Second, the translation of minimum nutritional requirements
 446 | into minimum food requirements depends on the choice of
 447 | commodities. While it may be easy to solve the programming
 448 | exercise of a 'diet problem', choosing a minimum cost diet for
 449 | meeting specified nutritional requirements from food items sold
 450 | at specified costs, the relevance of such a minimum cost diet is not
 451 | clear. Typically, it turns out to be very low-cost indeed, but
 452 | monumentally boring, and people's food habits are not, in fact,
 453 | determined by such a cost minimization exercise. The actual
 454 | incomes at which specified nutritional requirements are met will
 455 | depend greatly on the consumption habits of the people in
 456 | question.
 457 | Third, for non-food items such minimum requirements are not
 458 | easy to specify, and the problem is usually solved by assuming
 459 | that a specified proportion of total income will be spent on food.
 460 | With this assumption, the minimum food costs can be used to
 461 | derive minimum income requirements. But the proportion spent
 462 | on food varies not merely with habits and culture, but also with
 463 | relative prices and availability of goods and services. It is not
 464 | 
 465 | 
 466 | 
 467 | 
 468 | 
 469 | 
 470 |  PAGE 21 ENDS
 471 | 
 472 | 
 473 | 
 474 | 
 475 | 
 476 | 
 477 | The concept of poverty, then, is a complex one, and it is
 478 | 
 479 | 
 480 | 
 481 | 
 482 | 
 483 | 
 484 |  PAGE 22 ENDS
 485 | 
 486 | 
 487 | 
 488 | 
 489 | 
 490 | 
 491 | Poverty and Famines
 492 | Third, even when we do go through the intermediary of income, the translation of a set of nutritional norms (or of alternative sets of such norms) into a 'poverty line' income (or poverty-line incomes) may be substantially simplified by the wide prevalence of particular patterns of consumption behaviour in the community in question. Proximity of actual habits and behaviour makes it possible to derive income levels at which the nutritional norms will be 'typically' met. (This question is discussed further in Chapter 3.)
 493 | Finally, while it can hardly be denied that malnutrition captures only one aspect of our idea of poverty, it is an important aspect, and one that is particularly important for many developing countries. It seems clear that malnutrition must have a central place in the conception of poverty. How exactly this place is to be specified remains to be explored, but the recent tendency to dismiss the whole approach seems to be a robust example of misplaced sophistication.
 494 | THE INEQUALITY APPROACH
 495 | The idea that the concept of poverty is essentially one of inequality has some immediate plausibility. After all, transfers from the rich to the poor can make a substantial dent on poverty in most societies. Even the poverty line to be used for identifying the poor has to be drawn with respect to contemporary standards in the community in question, so that poverty may look very like inequality between the poorest group and the rest of the community.
 496 | Arguments in favour of viewing poverty as inequality are presented powerfully by Miller and Roby, who conclude:
 497 | Casting the issues of poverty in terms of stratification leads to regarding poverty as an issue of inequality. In this approach, we move away from efforts to measure poverty lines with pseudo-scientific accuracy. Instead, we look at the nature and size of the differences between the bottom 20 or 10 per cent and the rest of the society. Our concern becomes one of narrowing the differences between those at the bottom and the better-off in each stratification dimension.
 498 | There is clearly quite a bit to be said in favour of this approach. But one can argue that inequality is fundamentally a different
 499 | 
 500 | 
 501 | 
 502 | 
 503 | 
 504 | 
 505 |  PAGE 23 ENDS
 506 | 
 507 | 
 508 | 
 509 | 
 510 | 
 511 | 
 512 | Concepts of Poverty
 513 | 
 514 | It is, of course, quite a different matter to recognize that inequality and poverty are associated with each other, and to note that a different distribution system may cure poverty even without an expansion of the country's productive capabilities. Recognizing the distinct nature of poverty as a concept permits one to treat it as a matter of interest and involvement in itself. The role of inequality in the prevalence of poverty can then figure in the analysis of poverty without making the two conceptually equivalent.
 515 | 2.4 RELATIVE DEPRIVATION
 516 | The concept of 'relative deprivation' has been fruitfully used in the analysis of poverty, especially in the sociological literature. Being poor has clearly much to do with being deprived, and it is natural that, for a social animal, the concept of deprivation will be a relative one. But within the uniformity of the term 'relative deprivation', there seem to exist some distinct and different notions.
 517 | One distinction concerns the contrast between feelings of
 518 | 
 519 | 
 520 | 
 521 | 
 522 | 
 523 | 
 524 |  PAGE 24 ENDS
 525 | 
 526 | 
 527 | 
 528 | 
 529 | 
 530 | 
 531 | Poverty and Famines
 532 | 
 533 | deprivation and 'conditions of deprivation'. Peter Townsend has
 534 | argued that 'the latter would be a better usage'. There is indeed
 535 | much to be said for a set of criteria that can be based on concrete
 536 | conditions, so that one could use 'relative deprivation' 'in an
 537 | objective sense to describe situations where people possess less of
 538 | some desired attribute, be it income, favourable employment
 539 | conditions or power, than do others'.
 540 | On the other hand, the choice of 'conditions of deprivation' can
 541 | not be independent of 'feelings of deprivation'. Material
 542 | objects cannot be evaluated in this context without reference to
 543 | how people view them, and even if 'feelings' are not brought in
 544 | explicitly, they must have an implicit role in the selection of
 545 | 'attributes'. Townsend has rightly emphasized the importance of
 546 | the 'endeavour to define the style of living which is generally
 547 | shared or approved in each society and find whether there
 548 | is ... a point in the scale of the distribution of resources below
 549 | which families find it increasingly difficult ... to share in the
 550 | customs, activities and diets comprising that style of living'.
 551 | One must, however, look also at the feelings of deprivation in
 552 | deciding on the style and level of living the failure to share which
 553 | is regarded as important. The dissociation of 'conditions' from
 554 | 'feelings' is, therefore, not easy, and an objective diagnosis of
 555 | 'conditions' requires an objective understanding of 'feelings'.
 556 | A second contrast concerns the choice of 'reference groups' for
 557 | comparison. Again, one has to look at the groups with which the
 558 | people in question actually compare themselves, and this can be
 559 | one of the most difficult aspects of the study of poverty based on
 560 | relative deprivation. The horizon of comparison is not, of course,
 561 | independent of political activity in the community in question,
 562 | since one's sense of deprivation is closely related to one's
 563 | expectations as well as one's view of what is fair and who has the
 564 | right to enjoy what.
 565 | 
 566 | 
 567 | 
 568 | 
 569 | 
 570 | 
 571 |  PAGE 25 ENDS
 572 | 
 573 | 
 574 | 
 575 | 
 576 | 
 577 | 
 578 | Poverty is a complex phenomenon, and it is not easy to define
 579 | it in a precise way. It is, however, possible to distinguish between
 580 | two approaches to the concept of poverty. The first is the
 581 | approach of relative deprivation, which is based on the idea that
 582 | poverty is a relative concept, and that it is defined in terms of the
 583 | distribution of resources within a given society. Thus, a person
 584 | may be said to be poor if he or she is deprived of resources
 585 | which are available to other members of the same society. This
 586 | approach has been widely used in the literature on poverty and
 587 | has been the basis of many attempts to measure the extent of
 588 | attitude to the subject. But it is important to distinguish between
 589 | the two. The exercise of poverty assessment is a descriptive one,
 590 | and not a prescriptive one. It is concerned with the facts of
 591 | deprivation, and not with the values of the society.
 592 | 
 593 | 
 594 | 
 595 | 
 596 | 
 597 | 
 598 |  PAGE 26 ENDS
 599 | 
 600 | 
 601 | 
 602 | 
 603 | 
 604 | 
 605 | Poverty and Famines
 606 | 
 607 | policy, the 'policy definition' may fail to catch the political issues
 608 | in policy-making. Second, even if 'policy' is taken to stand not for actual public policy, but for policy recommendations widely held in the society in question, there are problems. There is clearly a difference between the notion of 'deprivation' and the idea of what should be eliminated by 'policy'. For one thing, policy recommendations must depend on an assessment of feasibilities ('ought implies can'), but to concede that some deprivations cannot be immediately eliminated is not the same thing as conceding that they must not currently be seen as deprivations. (Contrast: 'Look here, old man, you aren't really poor even though you are starving, since it is impossible in the present economic circumstances to maintain the income of everyone above the level needed to eliminate starvation.') Adam Smith's notion of subsistence based on 'the commodities which are indispensably necessary for the support of life' and 'what ever the custom of the country renders it indecent' for someone 'to be without' is by no means identical with what is generally accepted as could and should be provided to all as a matter of policy. If in a country suddenly impoverished, say, by war it is agreed generally that the income maintenance programme must be cut down to a lower level of income, would it be right to say that the country does not have any greater poverty since a reduction of incomes has been matched by a reduction of the poverty line?
 609 | I would submit that the 'policy definiton' is based on a fundamental confusion. It is certainly true that with economic development there are changes in the notion of what counts as deprivation and poverty, and there are changes also in the ideas as to what should be done. But while these two types of changes are interdependent and also intertemporally correlated with each other, neither can be defined entirely in terms of the other. Oil-rich Kuwait may be 'more able to support their dependent citizens' with its new prosperity, but the notion of what is poverty may not go up immediately to the corresponding level. Similarly, the war-devastated Netherlands may keep up its standard of what counts as poverty and not scale it down to the level commensurate with its predicament.
 610 | 
 611 | 
 612 | 
 613 | 
 614 | 
 615 | 
 616 |  PAGE 27 ENDS
 617 | 
 618 | 
 619 | 
 620 | 
 621 | 
 622 | 
 623 | If this approach is accepted, then the measurement of poverty must be seen as an exercise of description assessing the predicament of people in terms of the prevailing standards of necessities. It is primarily a factual rather than an ethical exercise, and the facts relate to what is regarded as deprivation, and not directly to what policies are recommended. The deprivation in question has both absolute and relative aspects.
 624 | 
 625 | Two issues remain untouched. First, in comparing the poverty of two societies, how can a common standard of necessities be found, since such standards would vary from society to society? There are two distinct types of exercises in such inter-community comparisons. One is aimed at comparing the extent of deprivation in each community in relation to their respective standards of minimum necessities, and the other is concerned with comparing the predicament of the two communities in terms of some given minimum standard.
 626 | 
 627 | Second, while the exercise of 'identification' of the poor can be based on a standard of minimum needs, that of 'aggregation' requires some method of combining deprivations of different people into some overall indicator. In the latter exercise some relative scaling of deprivations is necessary.
 628 | 
 629 | 
 630 | 
 631 | 
 632 | 
 633 | 
 634 |  PAGE 28 ENDS
 635 | 
 636 | 
 637 | 
 638 | 
 639 | 
 640 | 
 641 | Poverty and Famines
 642 | Poverty is a matter of deprivation. The recent shift in focus from absolute to relative deprivation has provided a useful framework of analysis. But relative deprivation is incomplete as an approach to poverty and supplements, but cannot supplant, the earlier approach of absolute dispossession. The biological approach, which deserves reformulation but not rejection, relates to the irreducible core of absolute deprivation, keeping issues of starvation and hunger at the centre of the concept of poverty.
 643 | To view poverty as an issue in inequality does little justice to either concept. There is an element of arbitrariness in the description of poverty and making that element as clear as possible is important. Conventions on this are less firmly established and the constraints of acceptability would tend to leave one with a good deal of freedom. The problem is somewhat comparable with the criteria for making aggregative descriptive statements in such fields as comparisons of sporting achievements of different groups. While it is clear that certain circumstances would permit one to make an aggregative statement, and other circumstances would force one to deny this, there are intermediate cases in which either of the two aggregative descriptive statements would be disputable.
 644 | The ethical exercises involve similar ambiguities and end up answering a different question from the descriptive one that was originally asked.
 645 | 
 646 | 
 647 | 
 648 | 
 649 | 
 650 | 
 651 |  PAGE 29 ENDS
 652 | 
 653 | 
 654 | 
 655 | 
 656 | 
 657 | 
 658 | Poverty and inequality relate closely to each other, but they are distinct concepts and neither subsumes the other. There is a good case for viewing the measurement of poverty not as an ethical exercise, but primarily as a descriptive one. Furthermore, it can be argued that the frequently used 'policy definition' of poverty is fundamentally flawed. The exercise of describing the predicament of the poor in terms of the prevailing standards of 'necessities' does involve ambiguities, which are inherent in the concept of poverty. But ambiguous description isn't the same thing as prescription. Instead, the arbitrariness that is inescapable in choosing between permissible procedures and possible interpretations of prevailing standards requires recognition and appropriate treatment.
 659 | 
 660 | The underlying methodological issues have been discussed in Sen (1980a).
 661 | 
 662 | 
 663 | 
 664 | 
 665 | 
 666 | 
 667 |  PAGE 30 ENDS
 668 | 
 669 | 
 670 | 
 671 | 
 672 | 
 673 | 
 674 | 
 675 | The question of whether commodities or characteristics should
 676 | be used in specifying basic needs is not easy to answer. It is
 677 | possible to argue that the use of characteristics is more
 678 | appropriate since it is the characteristics of commodities that
 679 | consumers seek, and it is these characteristics that are
 680 | important in determining the level of welfare. On the other
 681 | hand, it is also possible to argue that the use of commodities
 682 | is more appropriate since it is the commodities that are
 683 | available in the market, and it is these commodities that
 684 | consumers have to buy in order to satisfy their needs.
 685 | 
 686 | The choice between commodities and characteristics is
 687 | further complicated by the fact that the characteristics of
 688 | commodities are not always known. For example, the
 689 | calorie content of a particular commodity may not be known
 690 | with any degree of accuracy. In such cases, it is not possible
 691 | to specify the characteristics needs of the consumers. In
 692 | addition, the characteristics of commodities may vary
 693 | considerably from one region to another, and from one
 694 | season to another. This makes it difficult to specify
 695 | characteristics needs in a uniform manner.
 696 | 
 697 | In view of these considerations, it appears that the use of
 698 | commodities is more appropriate in specifying basic needs.
 699 | This is not to deny the importance of characteristics in
 700 | determining the level of welfare, but merely to suggest that
 701 | commodities are more appropriate in specifying basic needs
 702 | since they are more easily available and more easily
 703 | measurable. It is also possible to argue that the use of
 704 | commodities is more appropriate since it is the commodities
 705 | that are actually consumed by the poor, and it is these
 706 | commodities that are important in determining their level of
 707 | welfare.
 708 | 
 709 | In conclusion, it appears that the use of commodities is
 710 | more appropriate in specifying basic needs than the use of
 711 | characteristics. This is not to deny the importance of
 712 | characteristics in determining the level of welfare, but merely
 713 | to suggest that commodities are more appropriate in
 714 | specifying basic needs since they are more easily available
 715 | and more easily measurable.
 716 | 
 717 | 
 718 | 
 719 | 
 720 | 
 721 | 
 722 |  PAGE 31 ENDS
 723 | 
 724 | 
 725 | 
 726 | 
 727 | 
 728 | 
 729 | 
 730 | The characteristics needs are, in an obvious sense, prior to the needs for commodities, and translation of the former to the latter is possible only under special circumstances. Multiplicity of sources is, however, not uniform. Many commodities provide calories or proteins; rather few commodities provide shelter. Literacy comes almost entirely from elementary schooling, even though there are, in principle, other sources. In many cases, therefore, it is possible to move from characteristics requirements to commodity requirements—broadly defined—with rather little ambiguity. It is for this reason that 'basic' or 'minimum' needs are often specified in terms of a hybrid vector—e.g. amounts of calories, proteins, housing, schools, hospital beds—some of the components being pure characteristics while others are unabashed commodities. While there is some evidence that such mongrelism disconcerts the purist, it is quite economic, and typically does little harm.
 731 | 
 732 | An interesting intermediate case arises when a certain characteristic can be obtained from several different commodities, but the tastes of the community in question guarantee that the characteristic is obtained from one commodity only. A community may, for example, be wedded to rice, and may not treat the alternative sources of calories (or carbohydrates) as acceptable. A formal way of resolving the issue is to define the characteristic 'calories from rice' as the thing sought by the consumer in question, so that rice and rice alone can satisfy this. This is analytically adequate if a little underhand. But there are also other ways of handling the problem, e.g. the assumption that the group seeks calories as such but treats rice as its only feasible source. While these conceptual distinctions may not have much immediate practical importance, they tend to suggest rather different approaches to policy issues involving taste variations.
 733 | 
 734 | The role of knowledge accumulation in reforming ideas of feasible diets may in fact be an important part of nutritional planning. The knowledge in question includes both information about nutrition as such and experience of how things taste (once one breaks out of the barrier spotted by the old Guinness ad: 'I have never tasted it because I don't like it').
 735 | 
 736 | Dietary habits of a population are not, of course, immutable, but they have remarkable staying power. In making intercommunity comparisons of poverty, the contrast between for
 737 | 
 738 | 
 739 | 
 740 | 
 741 | 
 742 | 
 743 |  PAGE 32 ENDS
 744 | 
 745 | 
 746 | 
 747 | 
 748 | 
 749 | 
 750 | Poverty and Famines
 751 | In identifying the poor for a given set of basic needs, it is possible to use at least two alternative methods. One is simply to check the set of people whose actual consumption baskets happen to leave some basic need unsatisfied. This we may call the 'direct method', and it does not involve the use of any income notion, in particular not that of a poverty-line income. In contrast, in what may be called the 'income method', the first step is to calculate the minimum income at which all the specified minimum needs are satisfied. The next step is to identify those whose actual incomes fall below that poverty line.
 752 | In an obvious sense the direct method is superior to the income method, since the former is not based on particular assumptions of consumption behaviour which may or may not be accurate. Indeed, it could be argued that only in the absence of direct information regarding the satisfaction of the specified needs can there be a case for bringing in the intermediary of income, so that the income method is at most a second best.
 753 | There is much to be said for such a view, and the income method can indeed be seen as a way of approximating the results of the direct method. But the income method has the advantage of being more easily operationalized, and it is this which has led to its widespread use. Moreover, the income method has the advantage of being able to take account of the fact that formulating needs in terms of characteristics and formulating needs in terms of commodities may turn out to be significant. For example, the ranking of rural living standards in different states in India changes significantly when the basis of comparison is shifted from command over commodities to command over characteristics such as calories and protein. There is little doubt that ultimately characteristics provide the more relevant basis for specification of basic needs, but the relative inflexibility of taste factors makes the conversion of these basic needs into minimum cost diets a function not merely of prices but also of consumption habits. Explicit account would have to be taken of this issue in completing the identification exercise.
 754 | 
 755 | 
 756 | 
 757 | 
 758 | 
 759 | 
 760 |  PAGE 33 ENDS
 761 | 
 762 | 
 763 | 
 764 | 
 765 | 
 766 | 
 767 | The 'direct method' and the 'income method' are not, in fact, mutually exclusive. The direct method can be used to identify the poor, and the income method can be used to aggregate them. The direct method is based on the notion of deprivation, and it is used to identify those who are unable to satisfy their basic needs. The income method, on the other hand, is based on the notion of income, and it is used to measure the extent of poverty. The direct method is more accurate in identifying the poor, but it is difficult to implement due to the lack of detailed consumption data. The income method is easier to implement, but it does not take into account individual idiosyncrasies. 
 768 | There is a difficult line to draw here. If one were to look merely for the ability to meet minimum needs without being bothered by tastes, then one would, of course, set up a cost-minimizing programming problem and simply check whether someone's income falls short of that minimum cost solution. Such minimum cost diets are typically very inexpensive but exceedingly dull, and are very often regarded as unacceptable. In Indira Rajaraman's (1974) pioneering work on poverty in Punjab, in an initial round of optimization, unsuspecting Punjabis were subjected to a deluge of Bengal grams. Taste factors can be introduced through constraints, but it is difficult to decide how pervasive and severe these constraints should be. In the extreme case the constraints determine the consumption pattern entirely.
 769 | But there is, I believe, a difference in principle between taste constraints that apply broadly to the entire community and those that essentially reflect individual idiosyncrasies. If the poverty-level income can be derived from typical behaviour norms of society, a person with a higher income who is choosing to fast on a bed of nails can, with some legitimacy, be declared to be non-poor. The income method does, therefore, have some merit of its own, aside from its role as a way of approximating what would have been yielded by the direct method had all the detailed consumption data been available.
 770 | 
 771 | 
 772 | 
 773 | 
 774 | 
 775 | 
 776 |  PAGE 34 ENDS
 777 | 
 778 | 
 779 | 
 780 | 
 781 | 
 782 | 
 783 | Poverty and Famines
 784 | Two alternative ways of measuring the same thing, but represent two alternative conceptions of poverty. The direct method identifies those whose actual consumption fails to meet the accepted conventions of minimum needs, while the income method is after spotting those who do not have the ability to meet these needs within the behavioural constraints typical in that community. Both concepts are of some interest on their own in diagnosing poverty in a community, and while the latter is a bit more remote in being dependent on the existence of some typical behaviour pattern in the community, it is also a bit more refined in going beyond the observed choices into the notion of ability. A poor person, on this approach, is one whose income is not adequate to meet the specified minimum needs in conformity with the conventional behaviour pattern.
 785 | The income method has the advantage of providing a metric of numerical distances from the 'poverty line', in terms of income short-falls. This the 'direct method' does not provide, since it has to be content with pointing out the short-fall of each type of need. On the other hand, the income method is more restrictive in terms of preconditions necessary for the 'identification' exercise. First, if the pattern of consumption behaviour has no uniformity, there will be no specific level of income at which the 'typical' consumer meets his or her minimum needs. Second, if prices facing different groups of people differ, e.g. between social classes or income groups or localities, then the poverty line will be group-specific, even when uniform norms and uniform consumption habits are considered. These are real difficulties and cannot be wished away. That the assumption of a uniform poverty line for a given society distorts reality seems reasonably certain. What is much less clear, however, is the extent to which reality is thus distorted, and the seriousness of the distortion for the purposes for which the poverty measures may be used.
 786 | Another difficulty arises from the fact that the family rather than the individual is the natural unit as far as consumption behaviour
 787 | 
 788 | 
 789 | 
 790 | 
 791 | 
 792 | 
 793 |  PAGE 35 ENDS
 794 | 
 795 | 
 796 | 
 797 | 
 798 | 
 799 | 
 800 | Poverty: Identification and Aggregation
 801 | 
 802 | income they would need to maintain the same standard of living
 803 | if their family size were to change. This approach is based on
 804 | the assumption that the family's behaviour is such that the
 805 | additional income is used to maintain the same standard of living
 806 | for all family members. This approach is, however, difficult to
 807 | implement in practice, since it requires the collection of data on
 808 | the actual consumption patterns of families of different sizes.
 809 | 
 810 | Poverty estimation and social security operations require some
 811 | method of correspondence of family income with individual
 812 | income. The simplest method of doing this is to divide the family
 813 | income by the number of family members, but this overlooks the
 814 | economies of large scale that operate for many items of
 815 | consumption, and also the fact that the children's needs may be
 816 | quite different from those of adults. To cope with these issues,
 817 | the common practice is to convert each family into a certain
 818 | number of 'equivalent adults' by the use of some 'equivalence
 819 | scale', or, alternatively, to convert the families into 'equivalent
 820 | households'. There tends to be a lot of arbitrariness in any such
 821 | conversion. Much depends on the exact consumption pattern of
 822 | the people involved, which varies from family to family and with
 823 | age composition. The question of maldistribution within the
 824 | family is also an important issue requiring a good deal more
 825 | attention than it has received so far. There are also different
 826 | bases for deriving appropriate equivalence of needs. One
 827 | approach is to take the nutritional requirements for each age
 828 | group separately and then to take the ratios of their costs, given
 829 | established patterns of consumer behaviour. The acceptability of
 830 | this approach depends not merely on the validity of the
 831 | nutritional standards used, but also on the assumption that
 832 | family behaviour displays the same concern for fulfilling the
 833 | respective nutritional requirements of members of different age
 834 | groups in the family. It also ignores economies of scale in
 835 | consumption which seem to exist even for such items as food.
 836 | A second approach is to examine how the people involved
 837 | regard the equivalence question themselves, viz. how much extra
 838 | 
 839 | 
 840 | 
 841 | 
 842 | 
 843 | 
 844 |  PAGE 36 ENDS
 845 | 
 846 | 
 847 | 
 848 | 
 849 | 
 850 | 
 851 | Poverty and Famines
 852 | 
 853 | income they think is needed to make a larger family have the same standard of well-being as a smaller one. Empirical studies of these 'views' (e.g., Goedhart, Halberstadt, Kapteyn, and van Praag, 1977) have shown considerable regularities and consistency.
 854 | A third way is to examine the actual consumption behaviour of families of different size and to treat some aspect of this behaviour as an indicator of welfare. For example, the fraction spent on food has been treated as an indicator of poverty: two families of different size are regarded as having 'equivalent' incomes when they spend the same proportion of their incomes on food.
 855 | No matter how these equivalent scales are drawn up, there remains the further issue of the weighting of families of different size. Three alternative approaches may be considered: (1) put the same weight on each household, irrespective of size; (2) put the same weight on each person, irrespective of the size of the family to whom they belong; and (3) put a weight on each family equal to the number of equivalent adults in it.
 856 | The first method is clearly unsatisfactory since the poverty and suffering of a large family is, in an obvious sense, greater than that of a small family at a poverty level judged to be equivalent to that of the former. The third alternative might look like a nice compromise, but is, I believe, based on a confusion. The scale of 'equivalent adults' indicates conversion factors to be used to find out how well off members of that family are, but ultimately we are concerned with the sufferings of everyone in the family and not of a hypothetical equivalent number. If two can live as cheaply as one and a half and three as cheaply as two, these facts must be taken into account in comparing the relative well-beings of two-member and three-member families; but there is no reason why the suffering of two three-member families should receive any less weight than that of three two-member families at the same level of illfare. There is, thus, a good case for using procedure (2), after the level of well-being or poverty of each person has been ascertained by the use of equivalent scales taking note of the size and composition of the families to whom they belong.
 857 | 
 858 | 
 859 | 
 860 | 
 861 | 
 862 | 
 863 |  PAGE 37 ENDS
 864 | 
 865 | 
 866 | 
 867 | 
 868 | 
 869 | 
 870 | The income short-fall of a person whose income is less than the poverty-line income can be called his 'income gap'. In the aggregate assessment of poverty, these income gaps must be taken into account. But does it make a difference whether or not a person's short-fall is unusually large compared with those of others? It seems reasonable to argue that any person's poverty cannot really be independent of how poor the others are. Even with exactly the same absolute short-fall, a person may be thought to be 'poorer' if the other poor have short-falls smaller than his, in contrast with the case in which his short-fall is less than that of others. Quantification of poverty would, thus, seem to need the marrying of considerations of absolute and relative deprivation even after a set of minimum needs and a poverty line have been fixed.
 871 | The question of relative deprivation can be viewed also in the context of a possible transfer of a unit of income from a poor person to another who is richer but still below the poverty line and remains so even after the transfer. Such a transfer will increase the absolute short-fall of the first person by exactly the same amount by which the absolute short-fall of person 2 will be reduced. Can one then argue that the over-all poverty is unaffected by the transfer? One can dispute this, of course, by bringing in some notion of diminishing marginal utility of income, so that the utility loss of the first may be argued to be greater than the utility gain of the second. But such cardinal utility comparisons for different persons involves the use of a rather demanding informational structure with well-known difficulties. In the absence of cardinal comparisons of marginal utility gains and losses, is it then impossible to hold that the overall poverty of the community has increased? I would argue that this is not the case.
 872 | Person 1 is relatively deprived compared with 2 (and there may be others in between the two who are more deprived than 2 but less so than 1). When a unit of income is transferred from 1 to 2, it increases the absolute short-fall of a more deprived person and reduces that of someone less deprived, so that in a straightforward
 873 | 
 874 | 
 875 | 
 876 | 
 877 | 
 878 | 
 879 |  PAGE 38 ENDS
 880 | 
 881 | 
 882 | 
 883 | 
 884 | 
 885 | 
 886 | Poverty and Famines
 887 | 
 888 | proportion of the total income of the community that is received
 889 | by the poor, i.e. the 'income gap' measure G given by
 890 | G = (1/T) Σqiyi/Σni
 891 | where qi is the number of people in the ith income class and yi
 892 | the mean income of that class.
 893 | 
 894 | The head-count measure H and the income gap measure G
 895 | have been the mainstay of poverty measurement. Both have
 896 | shortcomings. The head-count measure H is a crude measure of
 897 | poverty, since it does not take into account the degree of
 898 | deprivation of the poor. It is possible that the same head-count
 899 | measure H can be obtained with different distributions of income
 900 | among the poor. For example, if the income of the poorest
 901 | person is doubled, while the incomes of all other poor people
 902 | remain unchanged, the head-count measure H will remain
 903 | unchanged. But the degree of deprivation of the poorest person
 904 | will have been reduced. The income gap measure G is a better
 905 | measure of poverty than the head-count measure H, since it
 906 | takes into account the degree of deprivation of the poor. But it
 907 | too has shortcomings. It does not take into account the
 908 | distribution of income among the poor. For example, if the
 909 | income of the poorest person is doubled, while the incomes of all
 910 | other poor people remain unchanged, the income gap measure G
 911 | will remain unchanged. But the degree of deprivation of the
 912 | poorest person will have been reduced.
 913 | 
 914 | The head-count measure H and the income gap measure G
 915 | have been the mainstay of poverty measurement. Both have
 916 | shortcomings. The head-count measure H is a crude measure of
 917 | poverty, since it does not take into account the degree of
 918 | deprivation of the poor. It is possible that the same head-count
 919 | measure H can be obtained with different distributions of income
 920 | among the poor. For example, if the income of the poorest
 921 | person is doubled, while the incomes of all other poor people
 922 | remain unchanged, the head-count measure H will remain
 923 | unchanged. But the degree of deprivation of the poorest person
 924 | will have been reduced. The income gap measure G is a better
 925 | measure of poverty than the head-count measure H, since it
 926 | takes into account the degree of deprivation of the poor. But it
 927 | too has shortcomings. It does not take into account the
 928 | distribution of income among the poor. For example, if the
 929 | income of the poorest person is doubled, while the incomes of all
 930 | other poor people remain
 931 | 
 932 | 
 933 | 
 934 | 
 935 | 
 936 | 
 937 |  PAGE 39 ENDS
 938 | 
 939 | 
 940 | 
 941 | 
 942 | 
 943 | 
 944 | Poverty: Identification and Aggregation
 945 | 
 946 | of a country than the proportion of its population below the poverty
 947 | line'.18 This is a remarkable statement, since it implies that the
 948 | number of poor people is the only thing that matters, and that the
 949 | extent of their poverty is of no consequence.
 950 | 
 951 | 
 952 | 
 953 | 
 954 | 
 955 | 
 956 |  PAGE 40 ENDS
 957 | 
 958 | 
 959 | 
 960 | 
 961 | 
 962 | 
 963 | Poverty and Famines
 964 | 
 965 | of the nation than that which shows what proportion are in
 966 | poverty' (p. 214). The spirit of the remark is acceptable enough,
 967 | but surely not the gratuitous identification of poverty with the
 968 | head-count measure H.
 969 | What about a combination of these poverty measures? The
 970 | head-count measure H ignores the extent of income short-falls,
 971 | while the income-gap ratio / ignores the numbers involved: why
 972 | not a combination of the two? This is, alas, still inadequate. If a
 973 | unit of income is transferred from a person below the poverty line
 974 | to someone who is richer but who still is (and remains) below the
 975 | poverty line, then both the measures H and I will remain
 976 | completely unaffected. Hence any 'combined' measure based
 977 | only on these two must also show no response whatsoever to such
 978 | a change, despite the obvious increase in aggregate poverty as a
 979 | consequence of this transfer in terms of relative deprivation.
 980 | There is, however, a special case in which a combination of H
 981 | and I might just about be adequate. Note that, while individually
 982 | H is insensitive to the extent of income short-falls and I to the
 983 | numbers involved, we could criticize the combination of the two
 984 | only for their insensitivity to variations of distribution of income
 985 | among the poor. If we were, then, to confine ourselves to cases in
 986 | which all the poor have precisely the same income, it may be
 987 | reasonable to expect that H and / together may do the job.
 988 | Transfers of the kind that have been considered above to show
 989 | the insensitivity of the combination of//and / will not then be in
 990 | the domain of our discourse.
 991 | The interest of the special case in which all the poor have the
 992 | same income does not arise from its being a very likely
 993 | occurrence. Its value lies in clarifying the way absolute deprivation vis-à-vis the poverty line may be handled when there isn't
 994 | the additional feature of relative deprivation among the poor. It
 995 | helps us to formulate a condition that the required poverty
 996 | measure P should satisfy when the problem of distribution among
 997 | the poor is assumed away by postulating equality. It provides one
 998 | regularity condition to be satisfied among others.
 999 | 
1000 | 
1001 | 
1002 | 
1003 | 
1004 | 
1005 |  PAGE 41 ENDS
1006 | 
1007 | 
1008 | 
1009 | 
1010 | 
1011 | 
1012 | END
1013 | 
1014 | 
1015 | 
1016 | 
1017 | 
1018 | 
1019 |  PAGE 42 ENDS
1020 | 
1021 | 
1022 | 
1023 | 
1024 | 
1025 | 
1026 | Poverty and Famines
1027 | A rather distinguished and simple case of such a relationship is to make the weight on any person i's income gap equal the rank value r(i). This makes the weights equidistanced, and the procedure is in the same spirit as Borda's (1781) famous argument for the rank-order method of voting, choosing equal distances in the absence of a convincing case for any alternative assumption. While this too is arbitrary, it captures the notion of relative deprivation in a simple way, and leads to a transparent procedure, making it quite clear what precisely is being assumed. It is, in fact, possible to derive the characteristic of equidistance from other—more primitive—axioms (see Sen, 1973b, 1974).
1028 | It should be remembered that in fixing the poverty line considerations of relative deprivation have already played a part, so that absolute deprivation vis-à-vis the poverty line is non-relative only in the limited context of the 'aggregation' exercise. As was discussed earlier, the concepts of absolute and relative deprivation are both relevant to each of the two exercises in the measurement of poverty, viz. identification and aggregation. Axioms A and R are each concerned exclusively with the aggregation exercise.
1029 | 
1030 | 
1031 | 
1032 | 
1033 | 
1034 | 
1035 |  PAGE 43 ENDS
1036 | 
1037 | 
1038 | 
1039 | 
1040 | 
1041 | 
1042 | Poverty: Identification and Aggregation
1043 | 
1044 | poverty more meaningful. The index P is a useful tool for
1045 | aggregating the information on poverty, and for making
1046 | comparisons across different situations.
1047 | 
1048 | 
1049 | 
1050 | 
1051 | 
1052 | 
1053 |  PAGE 44 ENDS
1054 | 
1055 | 
1056 | 
1057 | 
1058 | 
1059 | 
1060 | Poverty and Famines
1061 | 
1062 | Poverty is sensitive to the different features that are implicit in our
1063 | ideas on poverty. In particular, the question of distribution
1064 | remains relevant even when incomes below the poverty line are
1065 | considered. It will be necessary to go into this question further in
1066 | the context of analysing starvation and famines, as is done in the
1067 | chapters that follow.
1068 | 
1069 | 
1070 | 
1071 | 
1072 | 
1073 | 
1074 |  PAGE 45 ENDS
1075 | 
1076 | 
1077 | 
1078 | 
1079 | 
1080 | 
1081 | 
1082 | distinguish famines from regular starvation by the intensity of the
1083 | phenomenon, the extent of the area affected, and the duration of
1084 | the crisis.
1085 | 
1086 | Famines are not only disasters in terms of human mortality, but
1087 | also in terms of the disruption of the social fabric. In the words
1088 | of one observer:
1089 | 
1090 | Famines are not only physical disasters, but also social and
1091 | political ones. They are not only a matter of food supply, but
1092 | also of social and political organization. They are not only
1093 | caused by natural disasters, but also by human failure.
1094 | 
1095 | Famines, then, are not only disasters in terms of human mortality,
1096 | but also in terms of the disruption of the social fabric. They are
1097 | not only a matter of food supply, but also of social and political
1098 | organization. They are not only caused by natural disasters, but
1099 | also by human failure.
1100 | 
1101 | Famines are not only physical disasters, but also social and
1102 | political ones. They are not only a matter of food supply, but
1103 | also of social and political organization. They are not only
1104 | caused by natural disasters, but also by human failure.
1105 | 
1106 | Famines are not only disasters in terms of human mortality, but
1107 | also in terms of the disruption of the social fabric. They can
1108 | have a devastating effect on the economic and social life of a
1109 | region, leading to a breakdown of the social order and a
1110 | deterioration of the quality of life.
1111 | 
1112 | Famines can be caused by a variety of factors, including natural
1113 | disasters such as drought, floods, and pestilence; political
1114 | instability; war; and economic mismanagement. In some cases,
1115 | famines can be caused by a combination of factors.
1116 | 
1117 | Famines can have a devastating effect on the economic and
1118 | social life of a region, leading to a breakdown of the social
1119 | order and a deterioration of the quality of life. They can also
1120 | have a long-term impact on the health and well-being of the
1121 | population, leading to increased morbidity and mortality.
1122 | 
1123 | Famines can be prevented or alleviated through a variety of
1124 | measures, including the provision of food aid, the promotion of
1125 | agricultural production, and the implementation of social and
1126 | economic policies that are designed to reduce poverty and
1127 | inequality.
1128 | 
1129 | Famines can have a devastating effect on the economic and
1130 | social life of a region, leading to a breakdown of the social
1131 | order and a deterioration of the quality of life. They can also
1132 | have a long-term impact on the health and well-being of the
1133 | population, leading to increased morbidity and mortality.
1134 | Famines imply starvation, but not vice versa. And starvation
1135 | implies poverty, but not vice versa. The time has come for us to
1136 | move from the general terrain of poverty to the disastrous
1137 | phenomenon of famines.
1138 | Poverty, as was discussed in Chapter 2, can reflect relative
1139 | deprivation as opposed to absolute dispossession. It is possible for
1140 | poverty to exist, and be regarded as acute, even when no serious
1141 | starvation occurs. Starvation, on the other hand, does imply
1142 | poverty, since the absolute dispossession that characterizes
1143 | starvation is more than sufficient to be diagnosed as poverty, no
1144 | matter what story emerges from the view of relative deprivation.
1145 | Starvation is a normal feature in many parts of the world, but
1146 | this phenomenon of 'regular' starvation has to be distinguished
1147 | from violent outbursts of famines. It isn't just regular starvation
1148 | that one sees in famines.
1149 | 
1150 | Famines are not only disasters in terms of human mortality, but
1151 | also in terms of the disruption of the social fabric. In the words
1152 | of one observer:
1153 | 
1154 | Famines are not only physical disasters, but also social and
1155 | political ones. They are not only a matter of food supply, but
1156 | also of social and political organization. They are not only
1157 | caused by natural disasters, but also by human failure.
1158 | 
1159 | Famines can be caused by a variety of factors, including natural
1160 | disasters such as drought, floods, and pestilence; political
1161 | instability; war; and economic mismanagement. In some cases,
1162 | famines can be caused by a combination of factors.
1163 | 
1164 | Famines can have a devastating effect on the economic and
1165 | social life of a region, leading to a breakdown of the social
1166 | order and a deterioration of the quality of life. They can also
1167 | have a long-term impact on the health and well-being of the
1168 | population, leading to increased morbidity and mortality.
1169 | 
1170 | Famines can be prevented or alleviated through a variety of
1171 | measures, including the provision of food aid, the promotion of
1172 | agricultural production, and the implementation of social and
1173 | economic policies that are designed to reduce poverty and
1174 | inequality.
1175 | 
1176 | 
1177 | 
1178 | 
1179 | 
1180 | 
1181 |  PAGE 46 ENDS
1182 | 
1183 | 
1184 | 
1185 | 
1186 | 
1187 | 
1188 | Poverty and famines are often seen as an inevitable part of the economic landscape. However, it is possible to diagnose it—like a flood or a fire—even without being armed with a precise definition. In distinguishing between starvation and famine, it is not my intention here to attribute a sense of deliberate harming to the first absent in the second.
1189 | 
1190 | The history of famines as well as of regular hunger is full of bloodboiling tales of callousness and malevolence—and I shall have something to say on this—but the distinction between starvation and famine used in this work does not relate to this. Starvation is used here in the wider sense of people going without adequate food, while famine is a particularly virulent manifestation of its causing widespread death.
1191 | 
1192 | In analysing starvation in general, it is important to make clear distinctions between three different issues. Firstly, lowness of the typical level of food consumption; secondly, declining trend of food consumption; resulting in widespread and persistent hunger, evidenced by loss of body weight and emaciation and increase in the death rate caused either by starvation or disease resulting from the weakened condition of the population. In statistical term, it can be defined as a severe shortage of food accompanied by a significant increase in the local or regional death rate. Famine is an economic and social phenomenon characterised by the widespread lack of food resources which, in the absence of outside aid, leads to death of those affected.
1193 | 
1194 | The definitional exercise is more interesting in providing a pithy description of what happens in situations clearly diagnosed as one of famine than in helping us to do the diagnosis—the traditional function of a definition. For example, Gale Johnson's pointer to disease in addition to starvation directs our attention to an exceptionally important aspect of famines. See also Morris.
1195 | 
1196 | 
1197 | 
1198 | 
1199 | 
1200 | 
1201 |  PAGE 47 ENDS
1202 | 
1203 | 
1204 | 
1205 | 
1206 | 
1207 | 
1208 | The underlying empirical generalisation about trends of food availability has been, however, the subject of some controversy. See also Ohkawa (1957) and Ohkawa and Rosovsky (1973). Nakamura (1966) argues that in the Tokugawa period food consumption was more regular but less on average than in the Meiji era. Hobsbawm (QS?) also suggests a similar contrast in the British standard of living during 1790-1850. The empirical issue as to whether the quoted views of the Indian or Japanese economic history are correct is, of course, a different question.
1209 | 
1210 | 
1211 | 
1212 | 
1213 | 
1214 | 
1215 |  PAGE 48 ENDS
1216 | 
1217 | 
1218 | 
1219 | 
1220 | 
1221 | 
1222 | Poverty and Famines
1223 | 
1224 | the fact that a rising trend need not eliminate big fluctuations.
1225 | Indeed, there are good reasons to think that the trend of food
1226 | availability per head in recent years has been a rising one in most
1227 | parts of the world, but nevertheless acute starvation has
1228 | occurred quite often, and there is some evidence of intensification
1229 | of famine threats. While this is partly a problem of distribution
1230 | of food between different groups in a nation—an issue to which I
1231 | shall turn presently—there is also the time contrast (in
1232 | particular, the problem of sharp falls against a generally rising
1233 | trend). Famines can strike even when regular starvation is on
1234 | firm decline.
1235 | The food crisis of 1972 is a global example of this time contrast.
1236 | Colin Tudge (1977) describes the development in dramatic
1237 | terms:
1238 | The 1960s brought good harvests, augmented by the Third World's
1239 | 'green revolution', based on American-developed dwarf strains of
1240 | wheat and rice. The world's food problem was not shortage,
1241 | apparently, but over-production, leading to low prices and agricultural
1242 | depression. The US took land out of production, and in the early 1970s
1243 | both the US and Canada ran down their grain stores. Then the bad
1244 | weather of 1972 brought dismal harvests to the USSR, China, India,
1245 | Australia and the Sahel countries south of Sahara. Russia bought
1246 | massively in the world grain markets before others, including the US,
1247 | realized what was happening. By mid-1974 there was only enough
1248 | grain left in store to feed the world's population for three-and-a-half
1249 | weeks; terrifying brinkmanship.
1250 | In all this the focus has been on the total availability of food—
1251 | for the nation as a whole, or even for the world as a whole. But
1252 | exactly similar contrasts hold for food availability to a particular
1253 | section of a given community. A sudden collapse of the command
1254 | of a group over food can go against a rising trend (or against a
1255 | typically high level of food consumption). Problems of
1256 | (i) existence of much regular starvation, (ii) worsening trend of
1257 | regular starvation, and (iii) sudden outbreak of acute starvation,
1258 | are quite distinct. While they can accompany each other, they
1259 | need not, and often do not, do so.
1260 | 
1261 | 
1262 | 
1263 | 
1264 | 
1265 | 
1266 |  PAGE 49 ENDS
1267 | 
1268 | 
1269 | 
1270 | 
1271 | 
1272 | 
1273 | Starvation and famines
1274 | While famines involve fairly widespread acute starvation, there is no reason to think that it will affect all groups in the famine-affected nation. Indeed, it is by no means clear that there has ever occurred a famine in which all groups in a country have suffered from starvation, since different groups typically do have very different commanding powers over food, and an over-all shortage brings out the contrasting powers in stark clarity.
1275 | 
1276 | There has been some speculation as to whether such a comprehensive famine was not observed in India in 1344-5 (see Walford 1878, and Alamgir 1980, p. 14). There is indeed some evidence for this famine being a very widespread one. In fact, the authoritative Encyclopaedia Britannica saw the famine as one in which even 'the Mogul emperor was unable to obtain the necessaries for his household' (Eleventh Edition, 1910-1, vol. X, p. 167). This is most unlikely since the Mogul empire was not established in India until 1526! But it is also doubtful that the Tughlak king then in power—Mohammad Bin Tughlak—was really unable to obtain his household necessities, since he had the resources to organize one of the most illustrous famine relief programmes, including remitting taxes, distributing cash, and opening relief centres for the distribution of cooked food (see Loveday, 1916). One has to be careful about anecdotal history, just as a companion volume of the same Encyclopaedia points out.
1277 | 
1278 | The importance of inter-group distributional issues rests not merely in the fact that an over-all shortage may be very unequally shared by different groups, but also in the recognition that some groups can suffer acute absolute deprivation. One contrast that has received much professional attention recently is that between urban and rural population (see particularly Lipton, 1977). This contrast is indeed relevant to conflicts implicit in some famines (see for example Chapter 6 below), but there are other, more specialized, group conflicts which deserve more attention (some of these contrasts are taken up in Chapters 6, 7, 8, and 9).
1279 | 
1280 | 
1281 | 
1282 | 
1283 | 
1284 | 
1285 |  PAGE 50 ENDS
1286 | 
1287 | 
1288 | 
1289 | 
1290 | 
1291 | 
1292 | Poverty and Famines
1293 | 
1294 | even when there is no over-all shortage. There is no reason
1295 | whatsoever to think that the food consumption of different groups
1296 | must vary in the same direction, and in later chapters cases will be encountered in
1297 | which different groups' fortunes moved sharply in opposite
1298 | directions.
1299 | 
1300 | 
1301 | 
1302 | 
1303 | 
1304 | 
1305 |  PAGE 51 ENDS
1306 | 
1307 | 
1308 | 
1309 | 
1310 | 
1311 | 
1312 | 
1313 | opportunities available to him. The endowment of a person is
1314 | the bundle of commodities he owns, and the exchange opportunities are the terms on which he can exchange his endowment for other commodities.
1315 | 
1316 | 5.2 THE ENTITLEMENT APPROACH
1317 | The entitlement approach to starvation and famines concentrates on the ability of people to command food through the legal means available in the society, including the use of production possibilities, trade opportunities, entitlements vis-àvis the state, and other methods of acquiring food. A person starves either because he does not have the ability to command enough food, or because he does not use this ability to avoid starvation. The entitlement approach concentrates on the former, ignoring the latter possibility. Furthermore, it concentrates on those means of commanding food that are legitimized by the legal system in operation in that society. While it is an approach of some generality, it makes no attempt to include all possible influences that can in principle cause starvation, for example illegal transfers (e.g. looting), and choice failures (e.g. owing to inflexible food habits).
1318 | Ownership of food is one of the most primitive property rights, and in each society there are rules governing this right. The entitlement approach concentrates on each person's entitlements to commodity bundles including food, and views starvation as resulting from a failure to be entitled to a bundle with enough food.
1319 | In a fully directed economy, each person may simply get a particular commodity bundle which is assigned to him. To a limited extent this happens in most economies, e.g. to residents of old people's homes or of mental hospitals. Typically, however, there is a menu—possibly wide—to choose from. E, is the entitlement set of person 1 in a given society, in a given situation, and it consists of a set of alternative commodity bundles, any one of which the person can decide to have. In an economy with private ownership and exchange in the form of trade (exchange with others) and production (exchange with nature), E, can be characterized as depending on two parameters, viz. the endowment of the person (the ownership bundle) and the exchange
1320 | 
1321 | 
1322 | 
1323 | 
1324 | 
1325 | 
1326 |  PAGE 52 ENDS
1327 | 
1328 | 
1329 | 
1330 | 
1331 | 
1332 | 
1333 | Poverty and Famines
1334 | 
1335 | Entitlement mapping (the function that specifies the set of alternative commodity bundles that the person can command respectively for each endowment bundle) will depend on the legal, political, economic and social characteristics of the society in question and the person's position in it. Perhaps the simplest case in terms of traditional economic theory is one in which the endowment bundle can be exchanged in the market at fixed relative prices for any bundle costing no more, and here the exchange entitlement will be a traditional 'budget set'.
1336 | Bringing in production will make the E-mapping depend on production opportunities as well as trade possibilities of resources and products. It will also involve legal rights to apportioning the product, e.g. the capitalist rule of the 'entrepreneur' owning the produce. Sometimes the social conventions governing these rights can be very complex indeed—for example those governing the rights of migrant members of peasant families to a share of the peasant output.
1337 | Social security provisions are also reflected in the E-mapping, such as the right to unemployment benefit if one fails to find a job, or the right to income supplementation if one's income would fall otherwise below a certain specified level. And so are employment guarantees when they exist—as they do in some socialist economies—giving one the option to sell one's labour power to the government at a minimum price. E-mappings will depend also on provisions of taxation.
1338 | 
1339 | 
1340 | 
1341 | 
1342 | 
1343 | 
1344 |  PAGE 53 ENDS
1345 | 
1346 | 
1347 | 
1348 | 
1349 | 
1350 | 
1351 | The Entitlement Approach
1352 | Let the set of commodity bundles, each of which satisfies person z's minimum food requirement, be F,. Person i will be forced to starve because of unfavourable entitlement relations if and only if he is not entitled to any member of Fi given his endowment and his exchange entitlement mapping. The 'starvation set' S, of endowments consists of those endowment bundles such that the exchange entitlement sets corresponding to them contain no bundles satisfying his minimum food requirements.
1353 | Person i can be plunged into starvation if his endowment collapses into the starvation set S, either through a fall in the endowment bundle, or through an unfavourable shift in the exchange entitlement mapping. The distinction is illustrated in Figure 5. i in terms of the simple case of pure trade involving only two commodities, food and non-food. The exchange entitlement mapping is taken to assume the simple form of constant price exchange. With a price ratio p and a minimum food requirement OA, the starvation set Sl is given by the region O AB. If the endowment vector is x, the person is in a position to avoid starvation. This ability can fail either (1) through a lower endowment vector, e.g. x*, or (2) through a less favourable exchange entitlement mapping, e.g. that given by p*, which would make the starvation set OAC.
1354 | It is easy to see that starvation can develop for a certain group of people as its endowment vector collapses, and there are indeed many accounts of such endowment declines on the part of sections of the poor rural population in developing countries through alienation of land, sale of livestock, etc. Asset loss affects not merely the ability to exchange the asset directly with food, but also the ability to borrow against one's future earning power. Given the nature of the capital markets, substantial borrowing is typically impossible without tangible securities. The limitations of the capital markets often constitute an important aspect of famine conditions.
1355 | 
1356 | 
1357 | 
1358 | 
1359 | 
1360 | 
1361 |  PAGE 54 ENDS
1362 | 
1363 | 
1364 | 
1365 | 
1366 | 
1367 | 
1368 | 


--------------------------------------------------------------------------------