BSP4811�Innovation & Productivity

Policy and managerial challenge

• “Our key goal is to grow our productivity by 2% to 3% per year over the next decade, more than double the 1% we achieved over the last decade. ... If we achieve this goal, we can raise real incomes by one-third in 10 years.” Tharman Shanmugaratnam, Minister for Finance, Budget Speech, B.2, 22 February 2010.
• “Between 2008 and 2018, Singapore’s Total Factor Productivity (TFP) increased by 0.5 per cent per annum.” Minister for Trade & Industry Mr Chan Chun Sing, Oral Answer to Parliamentary Question, 10 January 2020.
• How did government perform?

Productivity: Fundamental equation

• “What gets measured, gets managed.”
• Y_it = A_it F(K_it, L_it, M_it)

where

• Y_it = output
• K_it = capital stock (flow of services)
• L_it = labor (flow)
• M_it = intermediate inputs
• Examples
• Whole economy
• Business: Airline; electricity generation
• Labour productivity: Y_it / L_it
• YouTube: Total Factor Productivity: Estimation https://youtu.be/h4FupvFDyZQ

Total factor productivity

Compare

• Labour productivity
• Total factor productivity

Productivity

• Manufacturing: U.S. (Syverson 2004)
• Ln TFP(p90) - Ln TFP(p10) = 0.651
• TFP(p90) = exp(0.651) TFP(p10) ≃ 1.9 TFP(p10)
• Manufacturing: Chinese and Indian TFP compared to U.S. (Hsieh and Klenow 2009)
• China: 25%-33% lower
• India: 31%-38% lower
• Services: UK, 1984-2001: Dispersion of productivity higher in services than manufacturing (Faccio et al., ICC 2010)
• Huge scope to raise productivity – adopting state of the art

Productivity

• Work smarter – get more output from same inputs => raise productivity
• Applying state of the art – re-engineering
• Shifting state of the art: Innovation
• Alternative approach to increasing output: Work harder – increase inputs
• Contrast increasing
• Labour productivity
• Total factor productivity

Policy and managerial challenge

• How to realize higher productivity?
• Within organization (“within”): Business strategy
• Re-allocate resources across organizations (“between”) – from laggards to leaders: Economic policy

Outline

• Environment
• Strategy
• Policy
• Policy evaluation
• Module administration

ECSO

Environment: Customers, Collaborators, Competitors, Government

Competencies

Strategy

Organization

Adapted from Heracleous and Werres, 2016

• Airlines: Singapore Airlines / Scoot
• Banking: HSBC / Hang Seng

Knowledge: Domains

• Knowledge: Basis of productivity and innovation
• Technical
• Information and communications technology
• Physical sciences
• Life sciences
• Operations
• Infrastructure: Layout
• Workflow
• Inputs
• Marketing
• Finance and accounting
• Human resource management

Knowledge: Domains

• Cross functional: Organization and strategy
• Knowledge
• Codified: SOPs – can be easily conveyed to new staff
• Tacit – need familiarity, proximity (geography)

​

Knowledge: Environment

• Access knowledge directly or indirectly from
• Customers
• Collaborators
• Competitors: Benchmarking
• Government: Industry Transformation Maps

Environment -- Change

“We live in an era of disruption in which powerful global forces are changing how we live and work. The rise of China, India, and other emerging economies, the rapid spread of digital technologies, growing challenges to globalization, and, in some countries, the splintering of

long-held social contracts are all roiling business, the economy, and society. These trends offer considerable new opportunities to companies, sectors, countries, and individuals that embrace them successfully. They are bringing forth dynamic and innovative new players on the world stage and could give a much-needed boost to productivity and prosperity in many countries. Indeed, our research shows that the benefits for those in the forefront are larger than ever. At the same time, the downside for those who cannot keep up has also

grown disproportionately. For business leaders, policy makers, and individuals, figuring out how to navigate these skewed times may require some radical rethinking.”

McKinsey Global Institute, January 2019

Environment -- Change

“All old-established national industries have been destroyed or are daily being destroyed. They are dislodged by new industries, whose introduction becomes a life and death question for all civilised nations, by industries that no longer work up indigenous raw material, but raw material drawn from the remotest zones; industries whose products are consumed, not only at home, but in every quarter of the globe. In place of the old wants, satisfied by the production of the country, we find new wants, requiring for their satisfaction the products of distant lands and climes. In place of the old local and national seclusion and self-sufficiency, we have intercourse in every direction, universal inter-dependence of nations. And as in material, so also in intellectual production. The intellectual creations of individual nations become common property.”

• When?
• Who?

Outline

• Environment
• Strategy
• Policy
• Policy evaluation
• Module administration

Why strategy?

• Productivity: Evolves over time
• Incremental gains – learning by doing
• Discrete steps – major innovations
• To achieve sustained, superior performance
• Appreciate environment
• Make superior plans
• Execute effectively

Why strategy?

• Competition need not be life-and-death struggle
• Less productive can continue for long time – just earn lower return on resources

Singapore food hawkers (Hou and Png)

Opportunity: Exploiting change

• Preferences
• Beverages: Bubble Tea – Boom and bust. How sustainable?
• Food: Beyond Meat
• Technology
• Computers: Vertically integrated to plug-and-play
• Electric cars => easier to manufacture but need complementary infrastructure (Adoption of innovations)
• Platforms: e-commerce, “sharing” services
• Deregulation: Budget carriers
• U.S. domestic deregulation: Southwest Airlines
• International deregulation: Air Asia

Opportunity: Managing change

• Challenge: When to switch to new and cannibalize the old? Innovation cycle.
• Product
• Kodak: Pioneered digital imaging, but unable or unwilling to switch
• Nokia: Dominant in feature phones
• Fashion: Giordano vs Uniqlo
• Vaccines: Inactivated virus => MRNA
• Service delivery
• Microsoft: Desktop computing => Internet => Cloud
• Comfort Delgro: Migrate to App?

More innovation need not be better

• Marginal benefit
• Marginal cost

Source: Insideretail.asia

Opportunity: Promoting change

• Circumventing cartel
• Private car hire services: Uber, Lyft in New York
• Who in taxi industry loses?
• Skirting regulation
• Globally: AirBNB
• Singapore: Uber, Grab – until June 2017
• China: P2P – until Nov 2020 (cancelled ANT IPO)

Outline

• Environment
• Strategy
• Policy
• Policy evaluation
• Module administration

Policy

• Enterprise Development Grant
• Supports projects to upgrade business, innovate or venture overseas
• Qualifying costs: third party consultancy fees, software and equipment, and internal manpower
• Conditions
• Registered and operating in Singapore
• Minimum of 30% local shareholding
• Financially viable
• Why should government subsidize efforts to increase productivity and innovation?

Policy

• Market imperfections
• Market power
• Externalities
• Asymmetric information

24

• Why central kitchen for eight restaurants?
• Why located in Tuas?
• Why cook on different days?
• Why SPRING and WDA subsidy?

Discussion question

• “In truth, innovation, not productivity, is almost everything” (Martin Wolf, Financial Times, 13 June 2018). Discuss.

Outline

• Environment
• Strategy
• Policy
• Policy evaluation
• Module administration

Econometrics: Refresher https://youtu.be/rahhtfvkmOY

Causal inference

• “peasants say that a cold wind blows in late spring because the oaks are budding, and really every spring cold winds do blow when the oak is budding. But though I do not know what causes the cold winds to blow when the oak buds unfold, I cannot agree with the peasants that the unfolding of the oak bud is the cause of the cold wind” (L.N. Tolstoy, War and Peace, Book Eleven)

​

Image: Vintage Classics

Investment and jobs, 2019

• Assoc Prof Walter Theseira asked the Minister for Trade and Industry regarding the $15.2 billion in fixed asset investments in 2019 (a) what is the dollar cost in grants and incentives to attract these investments; (b) how many of the 32,814 new jobs to be created are projected to be net job creation, rather than substitution of existing jobs in other firms or sectors; and (c) of the $15.2 billion, what amount is projected to be invested in the absence of grant and incentive support from EDB.

​

• Mr Chan Chun Sing: The Singapore Economic Development Board (EDB) uses a range of grant and tax incentives to attract investments into Singapore. These are compliant with the World Trade Organisation (WTO) rules, as well as with the Organisation for Economic Cooperation and Development (OECD) rules to address base erosion and profit shifting.

Investment and jobs, 2019

• The value of incentives awarded is a fraction of the projected total investment. Tax incentives are notional, and do not erode the tax base from existing activities. In addition, if a tax incentive recipient is not profitable, it will not enjoy any benefit. Other incentives such as training or research and development (R&D) grants are modest. Moreover, these grants are administered on a reimbursement basis. In other words, if the grant recipient does not incur the expenditure, it will not receive the grant.
• The number of jobs created are new jobs linked to new and expansion projects committed in that particular year.
• It is not meaningful to hypothesise the investments into Singapore absent these incentives. Competition for these investments is intensifying, and we must continue to remain globally competitive and an investment destination of choice. Doing so requires us to go beyond incentives, to strengthen our fundamentals such as our strong rule of law, pro-business environment, physical and data connectivity, and access to skilled talent. ...

​

Policy evaluation

• How to estimate impact of policy on productivity or innovation?
• Government policy: Does Enterprise Development Grant increase productivity?
• Business strategy: Does digital banking increase profit?
• Organization strategy: How to form teams?
• Incremental effect = Actual outcome – counterfactual outcome
• Always ask: “What is the counterfactual?”

DBS Bank: Digital strategy

• How should DBS adjust its strategy?
• If DBS converts all customers to digital, will ROE increase by 42 percent?

Source: Financial Times, 20 Nov 2017

Policy evaluation: Experimental design

• Prospective evaluation: Design and plan the evaluation before the policy is implemented
• Retrospective evaluation (observational study)
• Conduct the evaluation after the policy is implemented
• Study design constrained by
• Administrative policy: who can benefit from policy
• Available data: Policy administrator may not collect data useful for evaluation

Experimental design: Considerations

• Typically, design involves trade-off among
• External validity
• Ecological validity
• Internal validity

Prospective evaluation: �Randomized controlled trial

• Example: Evaluate new treatment for Covid-19
• Randomized controlled trial (A/B testing)
• Treatment/control: Why must have control?
• Why must randomize?
• Design
• Unit of randomization: Avoid spillovers (contamination of treatment)
• Intervention: Avoid Hawthorne Effect
• Diagnostics
• Test balance between treatment and control groups
• Manipulation check

Prospective evaluation: �Randomized controlled trial

• Other examples
• How to form a team (Chen and Gong, JEBO 2018)
• Self-formed
• Intelligently designed according to skills
• Random
• McDonalds [next slide]

Prospective evaluation: �Consumer behaviour

• McDonalds Singapore: Customers order more at self-service order stations
• Selection or
• Behavioral change?
• How to design experiment?

36

Prospective evaluation: �Enterprise Development Grant

• Randomized controlled trial
• Costly: Much more costly to experiment with businesses than individuals
• Fairness: Difficult for government to justify discriminating in policy between businesses (even randomly)
• Contamination of treatment – difficult to prevent control subjects from learning about treatment
• Randomized encouragement trial – but sufficient power?
• Regression discontinuity: Rank applicants by score and award to top N: (next slide)

37

Prospective evaluation: �Regression discontinuity

• Hypothetical grant
• Assess each application for grant according to a standard rubric
• Award to applications that exceed Q points
• Regression discontinuity: Compare at the cutoff – subjects should be almost identical, except for the treatment
• Statistically, as good as randomized controlled trial
• Challenge: Estimate concentrated on discontinuity, so need large sample around discontinuity

​

38

Retrospective evaluation:�Administrative policy

• Did policy apply to all or only some?
• Challenge: If policy applies to all
• Before/after study: Effect of new digital banking system: Compare deposits and loans before/after new system
• Cross-section (intensity of treatment): “Impact of Enterprise Singapore’s Grants on Firms’ Revenue and Exports”, Economic Survey of Singapore, 2021Q3, 44-50 [CANVAS]
• What is the research question?
• Who are in the sample?

Retrospective evaluation:�Difference in differences

• Treatment administered according to some criteria (not cut-off rule) unrelated to outcome of interest
• Treatment group
• Control group – provides counterfactual
• Compare before/after outcome among treatment/control
• Example: Effect of incentives on worker productivity
• Safelite Glass changed from hourly pay to performance pay, 1994-95 (Lazear AER 2000)
• Compare productivity of
• Workers under hourly vis-à-vis performance pay (cross-section)
• Each worker under hourly vis-à-vis performance pay (before/after)

Retrospective evaluation:�Difference in differences

• Supermarket: Effect of scan-only/self-pay on cashier productivity
• Outlets converted during periodic renovation
• Compare productivity of
• Cashiers in conventional vis-à-vis scan-only stores (cross-section)
• Each cashier in conventional vis-à-vis scan-only checkout (before/after)

Retrospective evaluation:�Difference in differences

• DID basis: Treatment administered according to some criteria (not cut-off rule) unrelated to outcome of interest
• Lazear study
• Not ideal because management switched to performance pay to increase productivity
• Not so bad because implemented in stages not according to store productivity
• Ideal: natural or quasi-natural experiment
• Effect of hardship on risk preferences: Fukushima earthquake – study people by distance from epicenter
• Can we apply difference-in-differences to analyze DBS digital strategy? Use non-digital customers as control?

​

Retrospective evaluation: �Cut-off administration

• Administrative policy – award treatment according to cut-off rule
• Examples
• Medical treatment: If condition > k, treat with X, otherwise do not treat
• Admission to NUS
• Singapore Armed Forces selection to Officer Cadet School
• Can apply regression discontinuity
• Administrative over-ride? China Innofund (Wang et al. Research Policy 2017)

43

Outline

• Environment
• Strategy
• Policy
• Policy evaluation
• Module administration

Learning outcomes

• Better appreciation of research into
• Strategies to increase productivity and innovation
• Public policy towards productivity and innovation
• Research skills
• Reading current research
• Causal inference in practice
• Presentation
• Commentary

Research

• Objective: Organize and focus thinking of managers and policy-makers
• Economize on limited cognitive resources and attention
• Sources
• Peer-reviewed scholarship: Search through Google Scholar: Covid-19 vaccines – arXiv vis-à-vis journals
• Not newspapers or magazines – good for sound-bites, not for scientific study

​

Module administration

• Wed, 8.30-11.30am
• Saturday, Sep 10, 9am-12pm
• Field visit, Wed, Sep 14: Seoul Garden Restaurant
• Assessment (BSN4811)
• Class participation: 20%
• Research papers – presentation and slides (1-3): 30%
• Research papers – commentary and slides (1-3): 20%
• Final examination: 30% (short questions, open book)

Administration

• Schedule and topics subject to revision
• Syllabus: Canvas
• Files: Readings
• Course readings
• Research papers
• Presentation – assigned team [After 2nd session]
• Commentary – assigned team
• Teams
• Form teams – someone new
• Team assessment + individual assessment
• Examplify

Reminder

• YouTube lectures
• Total Factor Productivity: Estimation https://youtu.be/h4FupvFDyZQ
• Econometrics: Refresher https://youtu.be/rahhtfvkmOY