Lessons from Israel’s Advanced Manufacturing Institute
Political and business leaders from across the political spectrum have begun to rethink manufacturing policy in the United States. Senator Marco Rubio recently called for an approach that “recognizes that government is often essential to identifying, organizing, and funding national projects beyond the scope of private initiative. It also recognizes that how best to carry out these projects is a prudential question, and that the answer will often involve market incentives as well as buy-in from the private sector.” My experience in the Israeli Ministry of Economy and Industry has taught me that this perspective offers the strongest foundation for good public policy.
I worked in the Israeli government for over eight years, most of them at the Ministry of Economy and Industry (MoEI). My job was to help strengthen the manufacturing sector in Israel by creating and implementing a national policy agenda as well as fostering a pro-business regulatory environment.
I would like to highlight one of these programs here—the Advanced Manufacturing Institute (“AMI”). I had the privilege of leading this program from concept to execution, and I believe it serves as a good example of the kind of manufacturing policy that Senator Rubio is advocating.
We created the AMI to encourage the Israeli manufacturing industry to adopt advanced manufacturing practices and technologies, and to assist individual manufacturers in beginning their own upgrades and transformations. The AMI works with manufacturers to create a roadmap based on their needs and opportunities, and then connects them to the best consultants and tech providers to implement their individual roadmap. While the program is partially subsidized by the government, manufactures who opt in face no government bureaucracy. The AMI is operated by a private company which can only make a profit if the services it provides bring measurable improvement to their client’s labor productivity.
The AMI has been operating since 2020 and has exceeded our expectations in terms of reach and impact. Demand far outpaced our initial projections, as more than 350 manufacturers signed on in the first three years. Our original goal was for two hundred enterprises within four years. Manufacturers who work with the AMI have experienced a 22 percent increase in labor productivity—twice the average rate of the Israeli manufacturing sector as a whole. These manufacturers have also averaged 16 percent growth in revenue—more than five times the typical growth for an Israeli manufacturer.
The Problem
Public policy is a messy process. Sometimes I think that the old saying about the sausage factory is unfair to sausage manufacturers. But good public policy, no matter how ugly or convoluted the process that produces it, can be situated within a logical and coherent story. A story starts with something that is wrong in the world (the “undesirable phenomenon”), explores the problems that cause it, and digs into the root causes of those problems. Then, the story explains how the policy addresses the problems and their root causes, and how it mitigates any undesired side effects of government intervention. The story should use clear quantitative indicators for evaluating improvements and rely on facts and data. Finally, the story ends with an evaluation of the policy, using the same quantitative indicators.
The AMI was created through a very messy process. But I believe it eventually fit the mold of a good story, one that offers several relevant lessons.
The manufacturing sector is an important driver of growth for the Israeli economy. It accounts for 18 percent of Israel’s private sector GDP and 50 percent of Israel’s exports. The manufacturing sector employs 10 percent of Israeli workers, and the average pay for a manufacturing job is 35 percent higher than the national average salary. Even when we examine subsets of workers—unskilled workers, workers with vocational training, workers with academic degrees—the average salary in manufacturing is higher for all of them than the national average for that group. The manufacturing sector is especially important as an employer outside of major urban areas—more than 15 percent of workers outside of Israel’s main metropolitan areas are employed in manufacturing, which is 50 percent higher than the national average. Manufacturing is also a key part of Israel’s national security and the stability of its economy. Israel’s ability to produce at least some of its own food, petroleum products, building materials, and weapons helps create an economy that is more resilient to external shocks.
For these reasons, the Israeli government believes it is important to not only maintain a strong manufacturing sector, but also to continue to improve it.
The Israeli manufacturing sector, however, has shown some indications of deterioration in the past two decades—the percentage of Israelis employed in manufacturing has slowly declined, along with the share of Israel’s total exports represented by manufactured goods. Most importantly, labor productivity in the manufacturing sector was 25 percent lower than the OECD average and 50 percent lower than the OECD top five. Additionally, Israeli manufacturers’ labor productivity has been growing at a slower rate than that of the OECD average for the past ten years.
This means that Israeli manufacturers are less efficient than their counterparts around the world, and that, in the long run, they will not be able to compete. Increased global competition is good for the private consumer, but it forces local manufacturers to become more efficient and innovate or be left behind. With this in mind, the problem that guided the majority of our policy efforts was the low labor productivity of the Israeli manufacturing sector.
There are many factors that can influence labor productivity—from national factors like infrastructure, education, and government regulation to specific factors such as how a single manufacturer runs its business, and everything in between.
We worked to address many of these factors. Here, I will focus on one of them—the adoption of advanced manufacturing practices and technologies. Advanced manufacturing, sometimes referred to as “Industry 4.0,” is the adoption of innovative practices and technologies that rely heavily on information, automation, computation, software, sensing, and networking. It is a strategic approach to production which utilizes these practices and technologies to create products as well as improve efficiency, flexibility, and competitiveness.
It is almost self-evident that the adoption of these advanced practices and technologies contributes to productivity, but this has also been demonstrated by ample research. While advancements in manufacturing technology are rapid, the rate of adoption of these technologies has been much slower. For this reason, most Western economies are concerned with increasing the rate of adoption of advanced manufacturing and have therefore created government programs expressly for this purpose. We had data that demonstrated that Israeli manufacturers were adopting technology at an even slower rate than their international counterparts, which served as a good indication that there was room for improvement.
Moreover, the issue of advanced manufacturing became a national political priority, enabling cross-agency cooperation and buy-in from the manufacturers themselves. This meant that attempting to address the issue would meet less resistance and more good will from most actors and agencies we would have to deal with, and so increased the chances of success.
Prior to the formation of the AMI, there were already government programs aimed at encouraging the adoption of advanced manufacturing. The two major agencies that provide large-scale subsidy programs are the Capital Investments Authority and the Innovation Authority.
Israel’s Capital Investments Authority subsidizes 20–30 percent of large-scale capital investments for manufacturers under certain conditions. It offers a specific program which subsidizes investment in advanced manufacturing technologies. The program has a budget of around $150 million per year, and gives an average subsidy of just under $1 million per manufacturer.
Israel’s Innovation Authority subsidizes many types of innovation. At the time, it had a specific program which subsidized up to 50 percent of the cost of adopting innovative technologies for manufacturers. This program had a budget of roughly $150 million per year and would give an average of $150,000 per recipient.
These subsidy programs had been running for several years and, while they did help some manufacturers, they also had several issues. A significant portion of manufacturers didn’t opt in, either because they didn’t know that these programs existed, or because they couldn’t deal with the bureaucracy involved in applying.
The bureaucracy associated with applying for these programs created a cast of “intermediaries” who would promise to complete all of the relevant paperwork. At best, this meant that manufacturers had to pay a third party in order to access a government service, but frequently these intermediaries would do a poor job and end up preventing the manufacturer from accessing assistance. In some cases, manufacturers were even forced to repay funds they had already received. Those who did apply and receive subsidies were often the manufacturers who had more resources to begin with.
Manufacturers usually have a multiyear strategy for their capital investments, a strategy which rarely relies on government subsidies. Therefore, most of the investment subsidized by these programs would have happened even without the subsidies. Additionally, the government bureaucrats in charge of these programs were not experts in advanced manufacturing technologies or practices. They knew a lot about manufacturing on a macro scale, but they had never worked in a factory. This made it difficult for them to accurately assess subsidy requests.
These issues tend to afflict a wide variety of government programs. In this case, however, we believed that the bigger problem was that these programs were not addressing the root causes that led to the slow adoption of advanced manufacturing methods.
The Root Causes
Adopting new technologies often requires large capital investments, which can be a barrier for some manufacturers. That is not often the main barrier, however. Several surveys conducted in Israel showed that the two main barriers preventing manufacturers from adopting new technologies are a lack of knowledge and a lack of adequate internal capabilities, both in terms of workers and of management. These results are also supported by the findings of international academic research.
The reasons for this are easy to understand—it is hard, even for experts, to keep up with the constant technological advancements in any given field. It is impossible for anyone to remain on the cutting edge of all relevant technologies—robotics, 3-D printing, sensors, manufacturing information systems, AI, etc. On top of that, most manufacturers can’t afford to have workers and managers who can understand all of these emerging technologies, find the most relevant ones, and take on the project of integrating them into the factory. Most manufacturers have one IT person at best; how would they be able to take on a project of digitizing their entire factory alone?
In theory, the free market should be able to take care of these barriers. Companies that sell advanced manufacturing technologies advertise to manufacturers, consultants help the manufacturers understand their needs, workers and management are trained or recruited. Yet this rarely happens because no single actor has the ability or incentive to overcome these barriers.
To begin with the manufacturers, it is hard for them to even recognize that they are behind until it’s too late. Even if they do understand the need to improve, they lack the knowledge and expertise to clearly define their needs to a technology provider or consultant. Consultants, for their part, prefer to work with clients who know how to utilize them, meaning manufacturers who have the resources to clearly define their needs. On top of that, consultants have no incentive to ensure that their recommendations are adopted by their clients; their job ends with a presentation or a report. The actual implementation is up to the manufacturer who, again, often doesn’t have the people who can effectively lead it.
Technology providers have a specific product they want to sell, and so look for manufacturers with problems that can be solved by their product. Simply because a manufacturer has a problem that could potentially be solved by your product, however, doesn’t mean it is the most important problem, or even that your product would actually be able to solve it. A certain robotics technology could be useful to a specific manufacturer, but if the firm doesn’t have the surrounding information systems, it might end up being less productive than its current technology. Technology providers, just like consultants, don’t have a stake in their clients’ success—they are merely hired to bring in the new technology.
The bottom line is that most manufacturers don’t have the internal capabilities to lead an effective transition into advanced manufacturing, and outside actors can only provide part of what is needed. Without a capable actor with a view of the entire picture, who is dedicated to the success of the manufacturer, the adoption of advanced technologies remains slow.
Thus, if the government wants to accelerate the adoption of advanced manufacturing technologies and practices, the most effective way would be to help bridge this gap. This could be done by creating a capable organization that would help manufacturers understand that they need to be more efficient in order to survive, help individual manufacturers understand their needs, and find the right providers and consultants to work with. Such an actor would also have to be invested in the success of its clients.
The Formation of the
Advanced Manufacturing Institute
The Advanced Manufacturing Institute was created to help the Israeli manufacturing industry adopt advanced manufacturing practices and technologies. It does so in two main ways.
The first is via direct assistance to individual manufacturers (“the assistance program”). The assistance program is made up of three parts. During the initial roadmap, AMI staff meet with the manufacturer, talk to workers and management, observe the factory floor, and collect data in order to create a preliminary assessment—mapping the client’s needs and opportunities for improvement. This step takes a few weeks and results in a “roadmap” for a project with specific recommendations. This costs about $10,000. The manufacturer pays 10 percent of that, and the rest is subsidized by the AMI. Moving onto the implementation stage, AMI staff connect the manufacturer with relevant consultants and/or technology providers to begin implementation of the roadmap. AMI staff accompany and assist throughout the process. This stage is subsidized at 30–60 percent, but hardware and software are not subsidized, unlike consulting and the “soft” components of tech integration. The government pays the subsidy to the consultants through the AMI, who act as a pipeline. Subsidies are capped at around $30,000 per manufacturer. Finally, during the follow up, AMI staff remain in contact with the manufacturer for at least two years, in order to collect data on labor productivity and other key performance indicators (KPIs).
Additionally, the AMI collects and disseminates relevant information regarding advanced manufacturing, acting as an information hub for the benefit of the whole Israeli manufacturing industry. This is done through publications, conferences, and professional training, among other things.
The AMI is run by a private contractor for the Israeli government. This contractor is a single-project company (SPC), meaning it was created for the sole purpose of running this project. This company was created by a consortium formed out of a manufacturing consulting firm and an engineering college, who submitted the best bid to run the project.
The contractor receives three types of payments. A flat quarterly payment which covers some of the contractor’s operating costs. A payment is also given for each manufacturer who completes the initial roadmap and begins the implementation stage. In theory, these two payments were designed to cover all of the AMI’s costs, so that the contractor would break even if they met our recruitment goals. The rates, however, are fixed, meaning they do not reference the actual cost for the contractor of providing the services. A bonus is given for each manufacturer who has gone through the assistance program and has shown measurable improvements in labor productivity, or other pre-defined KPIs. This bonus was designed so that the contractor could make up to a 15 percent profit.
This is a rough outline of the AMI’s model of operation. Considering the relatively low budget for the program, around $5 million per year, and the relatively low subsidies, under $40,000 per manufacturer, we hoped the AMI would become one of the most cost-effective government programs in Israel. There are several key features of this model that we predicted would lead to its success.
Incentives. The AMI is operated by a private contractor, whose main incentive is to make money. The model created two avenues for the AMI to generate revenue: (1) Getting manufacturers to opt in to the assistance program, complete the initial roadmap, and begin implementation, as well as (2) making sure the implementation led to measurable improvement in labor productivity. These two avenues are obviously complementary; proof of success helps convince more manufacturers to opt in.
This incentive structure pushes the AMI to become the missing actor to address the root causes of the low adoption rate of advanced manufacturing. The AMI is incentivized to actively recruit manufacturers, to help them understand where they need to improve, and to determine how advanced manufacturing can help them. It then helps them to find the most relevant and useful consultant or tech provider and makes sure that this leads to measurable improvements for the manufacturer.
It is important to note that the AMI has no financial stake in the implementation stage—it does not earn more if the manufacturer implements specific projects, or if the projects are more or less costly. The only incentive for the AMI is to push for projects that are more likely to lead to measurable benefits in productivity for the manufacturer.
We also looked at the incentives for the manufacturers. We wanted to make sure we weren’t driving them to acquire services they didn’t need simply because they could receive a subsidy. On this point, the relatively low caps on subsidies, along with the time and effort that the manufacturer had to put into the process, meant that it was unlikely that any would opt in if they didn’t believe they could benefit from these services.
For the consultants and tech providers, working with manufacturers through the AMI was just like working with any other client. The subsidies didn’t affect their rates, because the manufacturers still had an interest in negotiating the best price they could get. If anything, the manufacturers that worked with the AMI were better informed of both their needs and their available options. The AMI was required to offer at least two possible providers for all implementation projects.
All told, the operating model of the AMI created strong incentives for the contractor to advance our goals, while minimally interfering with the incentives of other actors.
Government involvement. The AMI was designed to act as a mediator between the government, manufacturers, consultants, and tech providers. Clients of the AMI receive government subsidies without having to fill in a single government form or speak to a bureaucrat—they only talk to AMI staff. On the back end, the AMI handles all of the government bureaucracy. It is the AMI’s responsibility to make sure that its clients meet the requirements to be included in the program, and to collect all the data required. It is also the AMI’s job to make sure the consultants and tech providers meet the relevant professional standards, and that they report their work properly so that the government can deliver its share of the costs. If there is an issue, it is the AMI whose payments get delayed. This gives the AMI’s clients a very different experience as compared to other government programs, which in turn means that manufacturers who would not usually work with the government chose to opt in. Moreover, consultants and tech providers who would not usually work with government programs, especially the top-tier ones, were willing to work with clients through the AMI.
On the government side, working through the AMI, and not directly with the clients, allowed us to reduce much of the bureaucratic burden. We asked for a lot of data on the AMI’s clients, but were able to be flexible with the formats and methods through which the AMI gave us this data. After a brief learning period, we were able to get consistently reliable submissions from the AMI, and they would get consistently reliable feedback and green lights from us.
Our main concerns regarding this model of government involvement were that all of the professional knowledge would remain in the hands of the AMI, and that we would not be able to effectively supervise their actions. To counter these concerns, the model allowed us to be as involved in the AMI’s operation as we desired and access all of the information we wanted. In practice, we had to sign off on every client that opted in, we examined and signed off on every roadmap, and we signed off on every implementation project, among other tasks. We also had full access to the AMI’s bank accounts, and all of their information systems. While this created a bureaucratic burden for the AMI—but not for its clients who remained blissfully unaware—it allowed us to effectively supervise its activities, and to gain a significant level of professional knowledge.
Professionalism. Another benefit of the incentive model was that, in order to be profitable, the AMI had to be a highly professional and respected actor in the Israeli manufacturing ecosystem. This meant that the AMI had to have the drive to go above and beyond what the government specification required.
For example, the AMI hired as CEO one of the most respected managers in the Israeli manufacturing ecosystem. The AMI staff are all manufacturing experts—experienced individuals with many years of involvement in manufacturing and consulting. These people are more qualified than the minimum standards we set.
Another example is how the AMI created an academic training program for “transformation leaders”—employees of manufacturers who had a roadmap and were starting implementation. The AMI used the resources of its parent engineering college, along with a regional NGO, to create this training program. Its goal was to ensure that the manufacturers had the internal resources to lead successful implementations. This training program turned out to be such a success that the engineering college made a similar program that manufacturers pay for their employees to attend.
Flexibility. We designed this program with the express intent of making it as flexible as possible. Most government programs required manufacturers to fit themselves to what the program was offering. The AMI was the complete opposite—first, it would learn the needs of the client, and then tailor a specific aid package for that client. Some clients needed digitization solutions, while others needed management consulting. Some clients were willing to pay more for a top-tier consultant, while others would not. The program allowed for this flexibility.
Such flexibility came in very handy during Covid—the AMI officially started operating in late February of 2020, just two weeks before Covid hit Israel. The AMI needed to adapt very quickly to this crisis, because its potential clients’ needs and priorities suddenly changed dramatically. The model’s flexibility, along with the AMI’s incentive to obtain clients, allowed the AMI to seize the opportunity Covid presented. While other government programs were scaling down because of the pandemic, the AMI actively recruited Israeli manufacturers, convincing them that now was the time to think long term. The AMI handled the Covid crisis so well that it became part of the Israeli government’s national Covid response.
Results
The AMI has been operating since 2020. In that time, it has achieved far more than we had hoped when the project started.
In its first three years, more than 350 manufacturers worked with the AMI, with demand far outpacing our initial projections. The AMI was, and is, clearly providing a service that Israeli manufacturers want. It also shows that the AMI’s financial incentive to recruit clients is working. In fact, the only thing that prevents the AMI from taking on more clients is that the government hasn’t been able to appropriate more money to fund its part of the subsidy.
Manufacturers who work with the AMI have shown a 22 percent increase in labor productivity—twice the average rate of the Israeli manufacturing sector as a whole. This indicator is the principal metric we sought to improve—it shows that manufacturers who use the AMI’s services become more efficient. We expect this number to improve even further in the coming years, because the effects of adopting advanced manufacturing technologies take time to translate into labor productivity.
Manufacturers working with the AMI also achieved 16 percent growth in revenue—more than five times the growth for the average Israeli manufacturer. In addition, specific KPIs, such as overall equipment effectiveness (OEE), lead time, and on-time delivery, were selected and measured for every manufacturer. On average, these KPIs improved by more than 30 percent. Surveys conducted by the manufacturers’ association show that the AMI’s activity is one of the main reasons Israeli manufacturers invest in advanced manufacturing, even those who aren’t clients of the AMI. This shows that the AMI was successful in becoming a key player in Israel’s advanced manufacturing ecosystem.
The data collected by the AMI also helped the MoEI learn a great deal about Israel’s manufacturing sector. The AMI’s data is more recent and more comprehensive than other sources of data, such as the National Statistics Bureau or the Tax Authority. This allows us to make better, more informed policy decisions across the board.
Lessons for Export
The AMI is an Israeli project, which was designed for Israeli manufacturers; it relied on data and analysis of the Israeli ecosystem. As such, some of the lessons learned from creating and running this project are inherently local. Nevertheless, I believe that there are several key insights that have a broader relevance.
First, good policy can happen only when it is clear what problems a program should be designed to address, and what problems are outside its scope. It is possible to have good policy that aims to create more industry jobs, and good policy that aims to increase labor productivity. A policy intended to achieve both at the same time, however, would likely not achieve either.
When considering a specific government intervention, it is crucial to understand how it is expected to affect the targeted problem—what is the precise mechanism that leads from this specific government intervention to the desired outcome? What are the exact steps? The more detailed this account is, the more likely it is that government intervention will indeed lead to the desired results. Data is necessary to define the problem, to understand the effects of the intervention, and to create precise incentives. Without data, it is impossible to assess a policy’s success.
Harnessing private sector knowledge, practices, and abilities can improve public policy immensely. In order to do that, it is crucial to align the relevant incentives in such a way that both private sector actors and the government want the same thing. Private sector actors should not be incentivized merely to make their government counterparts happy, but rather to make a measurable change in the issue that policymakers want to address.
I’ll end with the story of one AMI client—an electronics manufacturer that specializes in printed circuit board assembly. It is a relatively small company with seventy employees that operates in a very competitive market. Before the company started working with the AMI, it had experienced several years of growth in revenue, though not in profitability, and management was thinking of investing in new machinery to keep up with growing demand. The AMI worked very hard to convince the owner to join the program. After all, the company was doing well; it did not seem necessary.
The initial roadmap changed that perception. The owner assumed that the company was relatively efficient and estimated the factory’s OEE to be around 70 percent. (OEE refers to the percentage of time the machines are fully functional, relative to the time they are supposed to be.) The owner had to estimate this indicator because no data was being collected. AMI staff visited and observed the factory floor and found out the actual OEE was under 20 percent. That means that for every hour the machines were supposed to be working, they only worked for twelve minutes.
If this factory hadn’t worked with the AMI, it would have invested in new machinery which would have operated at the same 20 percent efficiency. Instead, using the AMI’s roadmap, the owner decided to invest in an MES—a system that collects data from the factory floor—as the first step toward improving efficiency. The AMI found a small provider with specific expertise that was very useful for the factory, and within two months the MES was up and running. After three more months, the factory had collected enough data to clearly identify the inefficiencies in the manufacturing process and started an additional project to address them. In sum, within eighteen months of starting work with the AMI, the factory’s OEE went from 20 percent to 65 percent, and labor productivity went up by 41 percent. The government subsidy for this client was just under $25,000.