Why we need law reform


back to issue

ECONOMISTS often talk about institutional constraints to growth. A good example is the legal system in India. Without reforms in the legal infrastructure, desired GDP (gross domestic product) growth rates of seven or eight per cent plus seem unlikely. Stated differently, post-1991 reforms involve a reliance on market mechanisms. Is the present legal system conducive to market-based behaviour?

Many laws continue to emphasize unnecessary state intervention. However, this is not the only angle to law reform. Even if the 1991 reforms had not happened, the legal system in India would come across as unsatisfactory and inefficient. Legal reform involves essentially two components, law reform and judicial reform. Law reform means changes in laws, while judicial reform refers more to the procedural aspect, that is, issues of reducing delays and speeding up dispute resolution.

Law as an expression needs to be defined. Not all law is statutory. Traditionally, India belongs to the common law tradition, i.e. some law has traditionally evolved, often through case law, without necessarily being codified. The counterpoint is civil law jurisdiction where most law is codified through statutes. Sometimes, this distinction is no longer very useful, viz. for the bulk of commercial law the law in India has been codified. In addition to statutory law, there is administrative law. Administrative law consists of government orders, regulations and rules. These, though not part of statutory law, are sanctified and allowed under some statutory law or the other.

In 1998, the Government of India set up a Commission on Review of Administrative Laws. Here is a quote from that report, submitted in September 1998.1 ‘The commission was seriously constrained by the fact that it did not have access to a complete set of subordinate legislation in the form of rules, regulations and administrative instructions, issued under different central Acts, by individual ministries and departments. It appears that the legislative department itself did not have such a complete compilation of rules, regulations and procedures issued by the ministries... Another handicap was that the central ministries did not have full information about the rules and regulations issued by state governments.’ This happened to a government-appointed commission on administrative law reform. It did not have access to all the administrative laws. Apparently, they were last collated and put together in one place in 1963.



However, there is a problem with statutory law as well. To quote from the Administrative Law Commission’s report:2 ‘There are now nearly 2500 central laws in force. While our focus in this study has been on central laws, it is worthwhile keeping in view the fact that there is not even a rough estimate available about the number of laws operating as state laws. In one state alone, the number is stated to be of the order of 1100. There might, thus, be 25000 to 30000 laws of states.’

The problem lies in Article 246 of the Indian Constitution. ‘(i) Notwithstanding anything in Clauses (2) and (3), Parliament has exclusive power to make laws with respect to any of the matters enumerated in List I in the Seventh Schedule (in this Constitution referred to as the "Union List". (ii) Notwithstanding anything in Clause (3), Parliament, and, subject to Clause (1), the Legislature of any state also, has power to make laws with respect to any of the matters enumerated in List III in the Seventh Schedule (in this Constitution referred to as the "Concurrent List". (iii) Subject to Clauses (1) and (2), the Legislature of any state has exclusive power to make laws for such state or any part thereof with respect to any of the matters enumerated in List II in the Seventh Schedule (in this Constitution referred to as the "State List".’



The centre and the states can both legislate and no one has the foggiest notion of how many state-level statutes there are. Apparently, in the state of Orissa, someone actually sat down, counted and discovered that there were 1015 in force. Extrapolated, one gets a figure of 25,000 or 30,000 state-level statutes.

The constraint is that these statutes are neither available and rarely published. In part, the problem concerns the Copyright Act of 1957. Under Section 2(k) of the act, ‘A work which is made or published under the direction or control of the government or any department of the government, any legislature in India, or any court, tribunal or other judicial authority in India’ is defined as ‘government work’.

Section 17(d) reads, ‘In the case of a government work, government shall, in the absence of any agreement to the contrary, be the first owner of the copyright therein.’ So the monopoly of publishing statutes is vested with the government, although Section 52(r) of the Copyright Act does offer some leeway. The point is that, given its monopoly, the government makes a mess of the business of publishing statutes, except the major ones. Hence the state-level statutes are neither published nor available.



There is a legal principle known as desuetude. This simply means that if a statute is not enforced for a long enough period, courts will regard that the statute has no legal effect. This holds even if the statute is not specifically repealed. But unlike Roman or Scot law, this principle is not accepted in English law, a tradition that India also follows. There is no desuetude. Consequently, unless a statute is specifically repealed, it continues to be on the statute books. The system is completely open-ended.

Given this background, there are several elements to law reform. First, there are old and dysfunctional statutes which need to be junked. Sometimes, an entire statute cannot be junked, but dysfunctional and old sections can be scrapped.3 Second, despite the large number of statutes, there are areas where the necessary law does not exist in India today.4 Third, there is over-legislation and unnecessary state intervention, both in statutory law and administrative law.5 These increase transaction costs. Fourth, dispute resolution needs to be speeded up. This has several elements – improving the efficiency of alternative channels of dispute resolution and cutting down on the demand for adjudication, curbing appeals and government litigation, computerization and improving the efficiency of existing benches, increasing the number of judges and benches and reforming procedural law.

All these issues cannot be addressed in the space of a brief paper. For our purposes, we will only illustrate some of these points using labour laws as an example.



Labour market rigidities constrain growth in employment, is an argument often advanced by economists.6 In the absence of flexible labour markets in the organized sector,7 growth in output does not necessarily lead to an increase in employment because labour effectively becomes a fixed input. Hence, production becomes artificially capital intensive. In an attempt to protect existing jobs, future and potential jobs are lost. Thus, the system is not in the broader interests of labour either. In addition, labour legislation creates relatively high wage islands in the organized sector such that India’s comparative advantage in an abundant supply of labour cannot be tapped. It is not surprising that most of India’s export basket originates in the unorganized sector. Conversely, the unorganized sector has virtually no protection.

Under Article 246 of the Indian Constitution, labour is placed in the concurrent list,8 barring exceptions like labour and safety in mines and oilfields and industrial disputes concerning Union employees that are in the central list. This issue is important given inter-state variations in labour laws. Consequently, labour market rigidity or flexibility varies from state to state. Subject to the comment made about labour being on the concurrent list, the list of central labour laws that have something directly to do with labour runs to 45.



Do we really need 45 and more statutes? Apart from the constitutional angle of the Seventh Schedule, are special statutes needed for cine workers, dock workers, motor transport workers, sales promotion employees, plantation labour, working journalists and workers in mines? Consider also the time span of the legislation, from the Fatal Accidents Act of 1855 to the Public Liability Insurance Act of 1991. Over a period of time, concepts and definitions have changed. So has the case law, contributing to further confusion. The law does not agree on definitions of adolescent, child, contract labour, wages, employee, workman, factory and industry. The case law only makes it worse. The case law under the Industrial Disputes Act has held almost everything to be an industry9 – panchayat samitis, state hospitals, real estate companies, running of tubewells, primary health centres, Federation of Indian Chambers of Commerce and Industry, religious institutions, universities and research institutions.

All these concepts, definitions and provisions need to be unified. For instance, all social security type provisions can be unified into a single statute. Similarly, all wage type legislation can also be unified into a single statute. The National Labour Association attempted this through a Uniform Indian Labour Code.10 But this has not been implemented.



From unification and harmonization, we now move on to reductions in state intervention in areas other than industrial relations. We will come to industrial relations later. The Factories Act is a good example of unnecessary government stipulations. Do we need the government to lay down stipulations like the following?11

Section 20: ‘(i) In every factory there shall be provided a sufficient number of spittoons in convenient places and they shall be maintained in a clean and hygienic condition. (ii) The state government may make rules prescribing the type and the number of spittoons to be provided and their location in any factory and provide for such further matters relating to their maintenance in a clean and hygienic condition. (iii) No person shall spit within the premises of a factory except in the spittoons provided for the purpose and a notice containing this provision and the penalty for its violation shall be prominently displayed at suitable places in the premises.’

Section 43: ‘The state government may, in respect of any factory or class or description of factories, make rules requiring the provision therein of suitable places for keeping clothing not worn during working hours and for the drying of wet clothing.’

These are examples from only one act. It is no one’s case that welfare provisions should not exist. But are welfare provisions enacted in 1948 still relevant? Assuming that they are, is the present government-mandated system with a regime of inspectors the best way to achieve the objective? Each labour legislation has a separate inspector and visits of inspectors are not synchronized across all labour enactments. Barring the Payment of Wages Act, where a maximum period of three years is stipulated, no other labour statute prescribes a maximum period for which records and registers must be maintained. Compliance is thus impossible and visits of inspectors result in bribery and rent-seeking. This system is not distributionally neutral as it tends to hurt the small-scale sector much more than it hurts large-scale industry. That apart, returns under various labour laws are not standardized and inspectors insist on maintenance of manual records and registers.

* There can be a common format for computerization of required records.

* There should be a single inspector for a given area.

* Some inspections for site and building and site plans or testing equipment can be farmed out to recognized private agencies.

* With the opening up of insurance, some social security provisions can be farmed out. For example, the Employees’ State Insurance Act hasn’t worked at all well.



Let us now turn to industrial relations. The three statutes that impinge on industrial relations are the Contract Labour (Regulation and Abolition) Act, the Trade Unions Act and the Industrial Disputes Act.

The Contract Labour (Regulation and Abolition) Act was never meant to prohibit contract labour. Section 10 provided the government the discretion of prohibiting contract labour in selected areas. In fact, regulation comes before abolition in the title of the act. Contract labour allows flexibility and permits outsourcing. However, a few court judgements have affected this flexibility. In 1960 and again in 1972, the Supreme Court ruled that if work performed by contract labour was essential to the main activity of the industry, contract labour should be abolished.12 Work should be done by regular workmen and contract labour should be absorbed by the principal employer.



In 1976, the central government issued a notification that in establishments run by it, contract labour should not be used for sweeping, cleaning, dusting and watching of buildings. But the 1970 act was still not clear on whether contract labour should be absorbed after the abolition.13 However, in a recent judgement, the Supreme Court has ruled that contract labour must be absorbed.14

Such judgements reduce labour market flexibility. There is an argument doing the rounds that the Contract Labour (Regulation and Abolition) Act should be scrapped. This sounds like an attractive proposition, but is probably facile. If the 1970 statute is scrapped, decisions on abolition of contract labour will revert from the government to industrial tribunals. To take the Factories Act as an example, industrial tribunals are likely to conclude that since canteens are mandated under Section 46 of the Factories Act, no contract labour can be employed in canteens. It seems a better idea to retain the 1970 act and tighten up Section 10 so that ambiguity about continuance of contract labour and absorption following abolition is removed.

Next, one should mention the Trade Unions Act. As a minor point, child labour is not prohibited in India; it is only prohibited in hazardous processes. Yet, under Section 21 of the Trade Unions Act, those under 15 are not allowed to be members of trade unions and Section 21-A prevents those under 18 from becoming office bearers. But more important are provisions of the act that lead to multiplicity. Under Section 4 of the Trade Unions Act, any seven people can form and register a trade union and these seven people need not even be workers. There is no cap on office bearers being from outside either. Nor is there any test for representativeness of a trade union, through secret ballots or otherwise.



The multiplicity problem impinges on collective bargaining because an agreement with one union is not necessarily binding on others. This is partly due to Section 18(1) of the Industrial Disputes Act, which states, ‘A settlement arrived at by agreement between the employer and workman otherwise than in the course of conciliation proceeding shall be binding on the parties to the agreement.’ It is not mandatory on others. Maharashtra and Gujarat are the only states with laws requiring recognition of trade unions by employers for purposes of collective bargaining. Following recommendations of the Second Labour Commission, the Union cabinet has now approved amendments to the Trade Unions Act. The number of persons required for registration of a trade union will change from 7 to 10% of the labour force. Not more than one-third of office bearers (subject to a maximum of five) can be outsiders. And the holding of annual elections and auditing of accounts will be mandatory. If these changes are accepted by Parliament, many of the problems connected with the Trade Unions Act will disappear.



Next one moves on to the Industrial Disputes Act; the following is a list of sections where there are problems.

Section 9-A: This prevents technological upgradation since, ‘No employer, who proposes to effect any change in the conditions of service applicable to any workman in respect of any matter specified in the Fourth Schedule, shall effect such change (a) without giving notice to the workman likely to be affected by such change a notice in the prescribed manner of the nature of the change proposed to be effected, or (b) within twenty-one days of giving such notice.’

Section 11: Not all disputes need to go through a process of adjudication. But Section 11 of the Industrial Disputes Act does not make it mandatory for conciliation officers to try for conciliation. This is despite Section 12(2), which merely states that the conciliation officer ‘may do all such things as he thinks fit for the purpose of inducing the parties to come to a fair and amicable settlement of the dispute.’ Nor do Sections 10 and 11 prescribe a maximum time limit for raising old disputes.

Section 11-A: This section needs to be quoted. ‘Where an industrial dispute relating to the discharge or dismissal of a workman has been referred to a labour court, tribunal or national tribunal for adjudication and, in the course of the adjudication proceedings, the labour court, tribunal or national tribunal, as the case may be, is satisfied that the order of discharge or dismissal was not satisfied, it may, by its award, set aside the order of discharge or dismissal and direct reinstatement of the workman on such terms and conditions, if any, as it may think fit, or give such other relief to the workman including the award of any lesser punishment in lieu of discharge or dismissal as the circumstances of the case may require.’ Should such blanket powers be granted to labour courts or tribunals? Directing appropriate retrenchment compensation is one thing, but is directing reinstatement necessary? Is such discretion to labour courts and tribunals necessary? A reversion to the pre-1971 statute, before Section 11-A was inserted, is probably better.

Section 17-B: If the employer appeals to a High Court or Supreme Court against the award of a labour court or tribunal, under Section 17-B, full wages have to be paid to the workman pending such proceedings, even if the appeal is admitted.

Sections 22/23: Section 22 prohibits strikes and lockouts without notice. But this section only applies to public utility services. Section 23 prohibits strikes and lockouts for all industrial establishments, but only during the pendency of conciliation or arbitration proceedings. These sections can be amended to require prior notice in the case of strikes and lock-outs for all industrial establishments. There can even be a requirement that a certain threshold percentage of workers must be in favour of the strike or lockout.



Chapter V-B/Sections 25-K, 25-L, 25-M, 25-N and 25-0: These provisions apply to industrial establishments that employ more than 100 workers and require prior permission of the appropriate government before layoffs, retrenchment and closure. Most problems connected with the Industrial Disputes Act arise from Chapter V-B, since the government becomes a third party to the dispute even if the employee is satisfied with the severance package. These sections need to be considered in conjunction with Section 2-A, which makes any dispute between an employer and an individual workman an industrial dispute ‘notwithstanding that no other workman nor any union of workmen is a party to the dispute.’ Note also the judgement of the Supreme Court in the Sundara Money case.15 Even if there is surplus labour force, that is no ground for retrenchment. A reversion to the pre-1976 statute, when Chapter V-B did not exist, is again desirable.



Consider now the case law.

* The discharge of an employee appointed on probation, during or at the end of the probationary period, is retrenchment.

* If there is a purely temporary appointment for nine days, terminated automatically at the end of nine days, that is retrenchment.

* A workman whose services are terminated because he failed to pass a test required for confirmation is retrenched.

* The termination of a workman’s service on account of unauthorized absence is retrenchment.

Perhaps a quote from a Supreme Court judgement is relevant.16 ‘Gradually, the net was cast too wide and the freedom of the employer tightened to such an extent by introduction of the impugned provisions that it has come to a breaking point from the point of view of the employers... It is not quite correct to say that because compensation is not a substitute for the remedy of prevention of unemployment, the latter remedy must be the only one. If it were so, then in no case closure can be or should be allowed... But, so long as the private ownership of an industry is recognised and governed on an overwhelmingly large proportion of our economic structure, is it possible to say that principles of socialism and social justice can be pushed to such an extreme so as to ignore completely, or to a very large extent, the interest of another section of the public, viz. the private owners of the undertakings?’



Next, a few comments about the adjudication process are in order, although this gets into general problems of dispute resolution. Under the Industrial Disputes Act, termination disputes are expected to be decided within three months. This rarely happens such that termination disputes have often remained pending for more than eight years. Beyond these eight years, there can be writ petitions before High Courts and special leave petitions as well. So a termination dispute can take more than 30 years.

There are several not mutually exclusive ways to solve this problem. First, there can be a greater emphasis on conciliation and mediation. Second, the working efficiency of labour courts and tribunals can be improved. If an average number of 65 cases per day are posted, why should on average only one case per day be heard and the remaining 64 adjourned?17 Benchmarks for daily performance can be set. Third, there is need to stick to the maximum of three adjournments permitted and not deviate from this principle. Fourth, there may be a need to create a labour judiciary delinked from the civil judiciary, as labour cases often require special expertise. Fifth, there should be no delay in filling vacancies. Sixth, the principle that advocates cannot appear before labour courts and tribunals without permission of the opposite party, needs to be enforced. That is, there is need to revise Section 38(2)(f) of the Industrial Disputes Act.18

Seventh, the requirement that awards must be published can be scrapped. This only contributes to additional delays. Eighth, an autonomous Industrial Relations Commission can be set up in each state. This is needed because the awards of labour courts and tribunals are meant to be final, but writ petitions are routinely admitted before High Courts under Articles 226 and 227 of the Constitution. Once there is an Industrial Relations Commission, appeals before High Courts can be scrapped, retaining appeals to the Supreme Court under Article 136 of the Constitution.



If some of these changes are implemented, labour markets will become more flexible, the segmentation between organized and unorganized labour markets will break down and India will be able to tap the comparative advantage of an abundant supply of skilled and unskilled labour. Reportedly, other than what has been said about the Trade Unions Act, the government is planning an Industrial Relations Act to replace the Industrial Disputes Act and the Contract Labour (Regulation and Abolition) Act. However, the contours of the proposed Industrial Relations Act are yet unclear but for the indication that Chapter V-B of the Industrial Disputes Act is likely to be retained.

Labour markets are but one example. The entire canvas of legal reform is a much broader one and will demand a long haul. Unfortunately, not much has happened since 1991.




1. Report of the Commission on Review of Administrative Laws, Department of Administrative Reforms and Public Grievances, Ministry of Personnel, Public Grievances and Pensions, Government of India, September 1998.

2. Ibid.

3. B. Debroy, In the Dock: Absurdities of Indian Law, Konark Publishers, 2000, is a collation of such dysfunctional elements in central statutes.

4. Having made the point, we will ignore this issue. Examples of missing legislation are in areas like electronic commerce, credit cards, hire purchase and leasing and some areas of intellectual property.

5. Over-legislation exists simultaneously with under-governance.

6. India: Policies to Reduce Poverty and Accelerate Sustainable Development, World Bank, 31 January 2000 is a recent example.

7. Which accounts for less than 8% of the labour force.

8. Item 22 on trade unions, industrial and labour disputes; item 23 on social security and social insurance, employment and unemployment; item 24 on welfare of labour including conditions of work, provident funds, employer’s liability, workmen’s compensation, invalidity and old age pensions and maternity benefits.

9. Especially after the Bangalore Water Supply and Sewerage Board vs. A. Rajappa case in 1978 and an amendment to the Industrial Disputes Act in 1982.

10. Uniform Indian Labour Code – A Draft, National Labour Association and FES, 1994.

11. This is only a small list. There are many more along similar lines.

12. Standard Vacuum Refining Company of India Limited vs. Its Workmen (1960 III SCR 466) and Vegoils Private Limited vs. The Workmen (1972 1 SCR 673).

13. The Supreme Court’s judgements in Denanath vs. National Fertilizers Limited (1992 1 SCC 695) and Gujarat Electricity Board vs. Hind Mazdoor Sabha (1995 5 SCC 27). The decision was effectively left to the Industrial Tribunal.

14. Air India vs. United Labour Union (1996 9 SCC 70).

15. State Bank of India vs. Sundara Money, AIR 1976 SC 1111.

16. Excel Wear vs. Union of India AIR 1979 SC 25.

17. These are actual sample survey figures from Karnataka, obtained through a study done by the National Law School of India University, Bangalore and quoted in V. Nagraj, ‘Labour Laws’, in N.R. Madhava Menon and Bibek Debroy (ed.), Legal Dimensions of Economic Reforms, Allied Publishers, 1995.

18. This governs conditions under which parties can be represented by legal practitioners.